System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting a plurality of data, such as user behavior data, purchase history, search history, and the like; means for analyzing the data to identify a user's interests and preferences; means for feeding data into a generative AI model based on the analysis to generate optimized text and images for the user; and means for automatically creating a landing page by placing the generated text and images.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Conventional landing page creation systems require users to manually create content tailored to their preferences, which takes time and effort and is prone to human error. For this reason, there is a need for a system that can quickly and accurately generate and place content optimized for each user, and provide effective information to each individual user. [Means for solving the problem]

[0005] The present invention is a system that includes a means for collecting multiple data such as user behavior data, purchase history, and search history, a means for analyzing the data and identifying the user's interests and preferences, a means for feeding the data to a generative AI model based on the analysis results and generating text and images optimized for the user, and a means for arranging the generated text and images and automatically creating a landing page. This makes it possible to automate the collection and analysis of user data and use generative AI to quickly and accurately generate and arrange content.

[0006] "User behavioral data" refers to data related to a user's online and offline activities, such as website browsing history, app usage history, and click data.

[0007] "Purchase history" refers to information about products and services a user has purchased in the past, and is data that includes product names, purchase dates, prices, store names, etc.

[0008] "Search history" is data about the search queries and results a user performs through search engines and in-app searches.

[0009] "Means of collecting data" refers to functions and processes for automatically discharging and storing user behavioral data, purchase history, search history, etc.

[0010] "Means of analyzing data" refers to algorithms and technologies used to analyze collected data and derive specific information, such as a user's hobbies and preferences.

[0011] "Means for identifying user preferences" refers to a process that uses analytical techniques or models to identify user interests and concerns.

[0012] A "clustering algorithm" is a mathematical method for grouping similar data points based on data.

[0013] "Generative AI models" refer to artificial intelligence models and techniques for generating new text and images based on large amounts of data.

[0014] "Feeding data" refers to the process of providing input data to a particular algorithm or model in order to operate that algorithm or model.

[0015] "Means for generating text and images" refers to the ability to automatically create new text and visuals using generative AI models.

[0016] A "landing page" is the first web page a user sees when they click on a specific link, and is designed to suit a specific purpose or theme.

[0017] A "template" is a template with a specific design and layout that is pre-set, and is used to arrange content such as text and images. [Brief explanation of the drawings]

[0018] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0019] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0020] First, the terms used in the following description will be explained.

[0021] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0022] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0023] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0024] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0025] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0026] [First embodiment]

[0027] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0028] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0029] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0030] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0031] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0032] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0033] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0034] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0036] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0037] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0038] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0039] The system that embodies this invention mainly consists of multiple servers, terminals, and users. The server analyzes various user data, and based on the results, it uses a generation AI to automatically generate the landing page (LP) that is most suitable for the user. This system quickly and efficiently provides optimal content based on the user's interests and preferences.

[0040] Program processing

[0041] User data collection

[0042] The server collects data such as user behavior, purchase history, and search history. Specifically, information such as the pages the user visited on the Internet, the products they purchased, and the search queries they performed is obtained via API. This allows the server to store detailed user preference information in a local database.

[0043] Data analysis

[0044] The server analyzes the collected data. This analysis includes preprocessing steps such as data cleansing and feature extraction. For example, categories of products frequently purchased by a user are extracted from their purchase history, and this is then passed through a clustering algorithm to classify users into multiple clusters. This makes it clear what interests the user has.

[0045] Preparing for generative AI

[0046] The server prepares the analysis results to be fed into a generative AI model. For example, a natural language processing model is used for text generation, and an image generation model is used for image generation. The generative AI model receives input such as user features and cluster information.

[0047] Content Generation

[0048] The server uses a generative AI model to generate text and images optimized for each user. For example, to generate a travel guide article for "User A," the server generates text including a title such as "Top 10 Adventure Travel Destinations" and detailed information about the adventure trip. In addition, related beautiful landscape images are also generated.

[0049] Creating a Landing Page

[0050] The server creates a landing page based on the generated content. Specifically, it inserts the generated text and images into a pre-defined template, builds the page structure using HTML and CSS, and finally deploys the page on a web server so that users can access it through their browsers.

[0051] Specific examples

[0052] User data collection

[0053] User B frequently purchases outdoor equipment from a shopping site and frequently reads articles about hiking using a certain device. The server collects this data via an API and stores it in a local database.

[0054] Data analysis

[0055] The server analyzes User B's data and classifies him / her into a cluster called "outdoor enthusiasts" based on the categories of outdoor equipment he / she has purchased and the content of articles he / she has viewed. It then identifies the areas and activities in which User B is particularly interested.

[0056] Preparing for generative AI

[0057] The server feeds the analysis results into the generative AI model and prepares to generate optimal content for User B. Specifically, it inputs data indicating that User B is interested in hiking into the generative AI.

[0058] Content Generation

[0059] The server uses generative AI to generate text content such as "Top 10 hiking trails recommended for User B." It also generates beautiful landscape images related to each hiking trail.

[0060] Creating a Landing Page

[0061] The server places the generated text and images into a template to create a landing page for the hiking guide, and finally deploys this page to a web server, making it accessible to User B.

[0062] In this way, the system can efficiently provide optimized information to User B.

[0063] The processing flow will be explained below.

[0064] Step 1:

[0065] The server collects data such as user behavior, purchase history, and search history. For example, information about the pages a user visits on a shopping site, the products they purchase, and the queries they search for is obtained via API and stored in a local database. This allows detailed information about the user's preferences to be gathered.

[0066] Step 2:

[0067] The server performs preprocessing, cleaning the data. Specifically, it fills in missing values ​​and removes outliers. It also standardizes the data format and standardizes each data point. For example, it converts all purchase history prices into the same currency unit.

[0068] Step 3:

[0069] The server analyzes the preprocessed data and extracts features to identify the user's interests and preferences. For example, features could include the product categories the user frequently purchases or the themes of the articles they view. Text analysis and category classification techniques are used to extract these features.

[0070] Step 4:

[0071] The server uses a clustering algorithm to group users, for example, using the K-means algorithm to classify users with similar interests into the same cluster. This clustering helps clarify the areas of interest and concern of users.

[0072] Step 5:

[0073] The server then feeds the data to a generative AI model based on the analysis results. For example, if a user belongs to a cluster called "Outdoors Lovers," that information is input to the generative AI. These generative AI models include text generation models (e.g., GPT-4) and image generation models (e.g., DALL-E).

[0074] Step 6:

[0075] The server uses the generative AI model to generate text and images optimized for each user. For example, for "User B," it generates text content such as "The 10 Best Hiking Trails" and related beautiful scenery images.

[0076] Step 7:

[0077] The server places the generated text and images into a landing page template, building the page layout using HTML and CSS and inserting the generated content in the appropriate places, resulting in a landing page with a consistent design.

[0078] Step 8:

[0079] The server deploys the completed landing page to the web server, allowing users to access the landing page through their browsers. User B then views the "10 Best Hiking Trails" page.

[0080] In this way, the system collects data, analyzes it, generates AI, places content, and deploys it step by step, quickly and efficiently providing users with optimized landing pages.

[0081] Example 1

[0082] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0083] With current technology, efficiently generating landing pages tailored to user preferences requires a lot of manual work, making it difficult to quickly provide optimal content.In addition, there are limited methods for effectively utilizing user behavior data, purchase history, search history, etc., making it difficult to generate personalized content for users.

[0084] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0085] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, and search history, means for analyzing the data and identifying the user's interests and preferences, means for generating prompt sentences based on the analysis results and feeding the data to a generative AI model, means for generating text and images optimized for the user using the generative AI model, and means for arranging the generated text and images to automatically create a landing page. This makes it possible to quickly and automatically generate an optimal landing page based on the user's preferences and efficiently provide content.

[0086] "User" means an end user of the Internet service or website.

[0087] "Behavioral data" refers to data such as page browsing history and click history when a user visits a website.

[0088] "Purchase history" refers to historical data about products a user has previously purchased from online shops, etc.

[0089] "Search history" refers to historical data about queries and search results a user has performed on the Internet.

[0090] "Data analysis" refers to a series of processes that aggregate, classify, and evaluate collected data.

[0091] "Hobbies and preferences" refers to data that indicates a user's interests, concerns, and preferences.

[0092] A "prompt" is an instruction given to a generative AI model.

[0093] A "generative AI model" is an artificial intelligence model that generates text or images from specified input data.

[0094] "Text" refers to sentences or strings of characters generated by a generative AI model.

[0095] "Image" refers to the visual data generated by a generative AI model.

[0096] A "landing page" is a web page created for a user to access in a web browser.

[0097] A "template" is a predefined layout or design framework used to automatically create a landing page.

[0098] "Deployment" is the process of placing a completed landing page on a web server so that it can be accessed by users.

[0099] A system embodying this invention efficiently collects and analyzes multiple data sets, such as user behavior data, purchase history, and search history, and automatically generates content optimized for each user using a generative AI model. Specific embodiments of this system are described below.

[0100] First, the server collects information such as user behavior data, purchase history, and search history. This data is obtained from the user's browser history, purchase history, search queries, etc. recorded by the device. The server obtains this data via API and stores it in a local database. For example, if a user purchases outdoor equipment on a shopping site or reads an article about hiking, that information will be collected.

[0101] The server then analyzes the collected data. At this stage, data cleansing is performed to remove noise and missing data. Feature extraction is also performed to identify users' interests based on their purchase history and browser history. For example, a clustering algorithm is used to separate the data into multiple clusters, categorizing users into clusters such as "outdoor enthusiasts" or "travel enthusiasts."

[0102] Based on the analysis results, the server generates a prompt sentence and prepares to feed it to the generative AI model. For example, if the analysis finds that the user has specific interests, it generates a prompt sentence such as, "Please tell me some recommended hiking trails for User A." The server then creates an instance of a generative AI model (e.g., GPT-3 or StyleGAN) and sets the prompt sentence as input.

[0103] The server inputs prompts into the generative AI model to generate text and images optimized for the user. For example, the generative AI model generates text content such as "Top 10 recommended hiking trails for user A" and also creates related beautiful landscape images. This automatically generates content that is individually optimized for each user.

[0104] Finally, the server creates a landing page based on the generated text and images. At this stage, the generated content is inserted according to a pre-defined template, and the page structure is built using HTML and CSS. The completed landing page is deployed to a web server, where users can access it through their browsers.

[0105] To illustrate how this system works, let's use a practical example. If User B is buying outdoor gear on a shopping site and reading an article about hiking, the server will collect this behavioral data and identify the user's interest as "outdoors enthusiast." As a result, a prompt message will be created that generates "10 recommended hiking trails for User B," and specific content will be generated by the generative AI model. This generated content will be placed on a landing page based on a template and deployed for User B's easy access.

[0106] An example of a prompt sentence is, "Please generate content that introduces hiking trails that User B is interested in. User B has recently purchased a lot of outdoor equipment and particularly enjoys hiking."

[0107] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0108] Step 1: Data collection

[0109] The server collects information such as user behavior data, purchase history, and search history. Specifically, the device sends the user's browser history, purchase records, and search queries to the server via API. The server stores this data in a local database. For example, if a user purchases outdoor equipment on a shopping site or reads an article about hiking, that information will be collected.

[0110] Input: User behavior data, purchase history, search history

[0111] Output: User data stored in a local database

[0112] Step 2: Data analysis

[0113] The server retrieves collected user data from a local database and first cleanses the data, removing missing data and noise. Next, it performs feature extraction to identify users' interests based on their purchase history and browser history. It then uses a clustering algorithm to categorize users into clusters such as "outdoor enthusiasts" or "travel enthusiasts."

[0114] Input: User data from a local database

[0115] Output: User features and cluster information

[0116] Step 3: Prompt generation

[0117] The server generates a prompt based on the results of the data analysis. For example, if the server finds that the user has a specific interest, it creates a prompt such as, "Please recommend some hiking trails for User A." This prepares the instruction sentence to be given to the generative AI model.

[0118] Input: Data analysis results (user features and cluster information)

[0119] Output: Generated prompt statement

[0120] Step 4: Input to the generative AI model

[0121] The server creates an instance of a generative AI model (e.g., GPT-3 or StyleGAN) and sets the prompt sentence generated in the previous step as input, so the model is ready to generate optimal content.

[0122] Input: Generated prompt text

[0123] Output: Input data to a generative AI model

[0124] Step 5: Content Generation

[0125] The server inputs the prompt into the generative AI model, which generates text and images optimized for the user. The generative AI model then generates specific content based on the prompt. For example, it generates text content such as "Top 10 recommended hiking trails for user A" and related beautiful landscape images.

[0126] Input: Prompt sentence for generative AI model

[0127] Output: Generated text and images

[0128] Step 6: Create a Landing Page

[0129] The server creates a landing page by arranging the generated text and images according to a pre-defined template, building the page structure using HTML and CSS, and inserting the generated content. The completed landing page is then deployed to a web server, where users can access it through their browsers.

[0130] Input: Generated text and images

[0131] Output: The generated landing page

[0132] This process flow allows the system to quickly and efficiently generate optimal landing pages based on user preferences.

[0133] (Application example 1)

[0134] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0135] Conventional internet advertising and landing page creation systems have not adequately provided optimal content based on users' hobbies and preferences, and have had difficulty providing content in real time. As a result, advertisements and content that do not match the user's interests are displayed, making effective marketing impossible. Furthermore, even when displaying advertisements using wearable devices such as smart glasses, there is a lack of a means to provide users with content optimized for them in real time. There is a need to solve these issues.

[0136] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0137] In this invention, the server includes a means for collecting multiple data such as user behavior data, purchase history, and search history, a means for analyzing the data to identify the user's interests and preferences, and a means for feeding the data to a generative AI model based on the analysis results to generate text and images optimized for the user, thereby arranging the generated text and images to automatically create landing pages or advertising content and display them on the smart glasses.

[0138] "User behavior data" refers to information such as the pages a user visits on the Internet, the links they click, and the time they spend browsing.

[0139] "Purchase history" refers to information such as the products and services a user has purchased in the past, the date and time of those purchases, and frequency of those purchases.

[0140] "Search history" refers to the search queries a user has made on a search engine and the history of the related search result pages they have visited.

[0141] A "generative AI model" is an artificial intelligence model trained on a large dataset, and refers to a technique for generating text or images based on specific input data.

[0142] A "clustering algorithm" refers to a statistical method for analyzing large amounts of data and grouping data that are highly similar.

[0143] A "landing page" is a web page that appears first when a user clicks on an advertisement or link, and contains content designed to encourage a specific action (such as purchasing a product or entering information).

[0144] "Smart glasses" are a type of wearable device that has the shape of glasses but has the ability to display visual information as augmented reality (AR).

[0145] "Template" means a document or web page structure with a specific format or style predefined for efficient layout and display of content.

[0146] The system for implementing this invention consists of a server, a terminal, and a user. The server collects and analyzes information such as user behavior data, purchase history, and search history, and uses AI to display optimal landing pages and advertising content on the smart glasses based on the results.

[0147] Program Overview

[0148] The system includes the following procedures:

[0149] 1. User data collection:

[0150] The server collects data such as user behavior, purchase history, and search history via API, and stores detailed user preference information in a local database.

[0151] 2. Data Analysis:

[0152] The server cleanses the collected data, extracts and analyzes features, and uses a clustering algorithm to classify users into different clusters based on their purchase history and the pages they have viewed.

[0153] 3. Prepare the generative AI:

[0154] The server creates a prompt sentence to feed the analysis results to a generative AI model (e.g., GPT-3), which includes the user's cluster information and interests.

[0155] 4. Advertising content generation:

[0156] A generative AI model generates optimized text and images based on the prompt, and the generated ad content is tailored to what the user should see.

[0157] 5. Display on smart glasses:

[0158] The server displays the generated content on the smart glasses in real time, allowing users to visually view the most suitable advertising content through the glasses they are wearing.

[0159] Specific examples

[0160] Let's say a user is walking through a particular shopping mall. The server determines from past behavioral data and purchase history that the user is interested in outdoor gear. Based on this analysis, it inputs the following prompt into the generative AI model:

[0161] plaintext

[0162] "Generate ads that are best suited to users in Cluster 1, whose primary interest is outdoor gear. Include current sales and popular items in the ad content."

[0163] Based on this prompt, the generative AI model generates advertising text, such as "The latest outdoor gear is on sale!", along with images of the products on sale. The server then displays this generated advertising content on the smart glasses in real time, providing users with the most appropriate information.

[0164] Hardware and software used

[0165] Hardware: Smart glasses (e.g. HoloLens, Google Glass)

[0166] software:

[0167] Data collection via API: RESTful API

[0168] Data analysis: NumPy, scikit-learn

[0169] Generative AI model: GPT-3 (OpenAI)

[0170] This makes it possible to provide advertising content that matches the user's preferences in real time.

[0171] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0172] Step 1:

[0173] The server collects user behavioral data, purchase history, and search history via API. Specifically, the server accesses each API endpoint and retrieves data. For example, behavioral data including the user's visited pages and browsing time, history of past purchases, and search engine query history are collected. This data is stored in a local database.

[0174] Input: API endpoint

[0175] Output: User data stored in a local database

[0176] Step 2:

[0177] The server cleanses the collected data and extracts features. First, it removes duplicate data and missing values ​​to clean the data. Next, it extracts the categories of products that users frequently purchase and the features of their search queries. For example, if a user frequently purchases outdoor equipment, the features of that category are extracted.

[0178] Input: Raw data in a local database

[0179] Output: Cleansed feature data

[0180] Step 3:

[0181] The server runs a clustering algorithm on the cleansed data to classify users into different clusters. For example, it uses KMeans clustering to classify users into clusters such as "outdoor enthusiasts" and "tech gadget enthusiasts." This identifies the user's interests and preferences.

[0182] Input: Cleansed feature data

[0183] Output: User data categorized into clusters

[0184] Step 4:

[0185] The server then creates prompts to be input into the generative AI model based on the clustered user data. For example, for a user who loves the outdoors, the server might generate a prompt like, "Generate ads that are optimal for users in cluster 1. Their primary interest is outdoor gear. Include current sales information and popular products in the ad content."

[0186] Input: User data categorized into clusters

[0187] Output: A prompt to be input to the generative AI model

[0188] Step 5:

[0189] The server inputs a prompt into a generative AI model (e.g., GPT-3) to generate text and images. Based on the prompt, the generative AI model generates optimal ad text and related images. For example, ad text such as "Latest outdoor gear on sale!" and images of outdoor gear on sale are generated.

[0190] Input: prompt statement

[0191] Output: Generated ad text and images

[0192] Step 6:

[0193] The server transmits the generated advertising text and images to the smart glasses for display in real time, and the smart glasses visually display the received advertising content, allowing the user to easily check the information.

[0194] Input: Generated ad text and image

[0195] Output: Advertising content displayed on smart glasses

[0196] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0197] The system for implementing this invention comprises multiple servers, terminals, and users. This system includes an emotion engine that analyzes various user data and uses generative AI to automatically generate the optimal landing page (LP) for the user. This emotion engine analyzes the user's text messages, voice input, images, and videos, identifies the user's emotions, and generates optimized content. This allows for efficient provision of content that takes user emotions into consideration.

[0198] Program processing

[0199] User data collection

[0200] The server collects multiple data sets, including user behavior data, purchase history, and search history. Specifically, information about the pages the user visited, the products they purchased, and the queries they searched is obtained via API and stored in a local database. It also uses an emotion engine to collect the emotions expressed by the user during interactions.

[0201] Data analysis

[0202] The server analyzes the collected data. This analysis includes data cleaning and feature extraction. For example, it extracts the categories of products frequently purchased by users from their purchase history and classifies users into different clusters using a clustering algorithm. Furthermore, it uses an emotion engine to analyze emotions from users' text messages and voice inputs to identify what content is more appropriate.

[0203] Preparing for generative AI

[0204] The server feeds data to a generative AI model based on the analysis results. For example, if a user has a cluster of "loves the outdoors" and an emotion of "joy," that information is input into the generative AI. These generative AI models include text generation models (e.g., GPT-4) and image generation models (e.g., DALL-E).

[0205] Content Generation

[0206] The server uses a generative AI model to generate text and images optimized for each user. For example, for "User B," it generates text content such as "The 10 Best Hiking Trails," along with related images of beautiful scenery that evoke a sense of enjoyment. Furthermore, if the user expresses the emotion of "joy," it selects positive words and images that further enhance this emotion.

[0207] Creating a Landing Page

[0208] The server places the generated text and images into a landing page template, building the page layout using HTML and CSS and inserting the generated content in the appropriate places, resulting in a landing page with a consistent design.

[0209] Deploy

[0210] The server deploys the completed landing page to the web server, which allows users to access the landing page through their browsers. User B views the "10 Best Hiking Trails" page, which is designed to provide emotional satisfaction.

[0211] Specific examples

[0212] User data collection

[0213] User B frequently purchases outdoor equipment on a shopping site and reads articles about hiking using a certain device. User B expresses emotions such as "joy" and "excitement" while interacting with the site. The server collects this information via an API and stores it in a local database.

[0214] Data analysis

[0215] The server analyzes User B's data and identifies a cluster called "outdoor enthusiasts" based on the categories of outdoor equipment purchased and the content of articles viewed. It also analyzes User B's emotional data using an emotion engine to identify "joy" and "excitement."

[0216] Preparing for generative AI

[0217] The server feeds the analysis results into a generative AI model. For example, it inputs the cluster "loves the outdoors" and the emotion information "joy." The generative AI generates optimal text and images based on a large amount of data.

[0218] Content Generation

[0219] The server uses the generative AI model to generate text content such as "10 recommended hiking trails for User B." It also generates beautiful scenic images of hiking trails that User B is interested in. It uses positive words and colors that reinforce User B's emotion of "joy."

[0220] Creating a Landing Page

[0221] The server places the generated text and images into a template, inserting the generated content into a travel guide template and building the page layout.

[0222] Deploy

[0223] The server deploys the completed landing page to the web server, and User B accesses this page through a browser, which also provides emotional satisfaction.

[0224] In this way, the system collects data, analyzes it, analyzes emotions, generates AI, places content, and deploys it at each step, enabling it to quickly and efficiently provide optimal landing pages that also take user emotions into consideration.

[0225] The processing flow will be explained below.

[0226] Step 1:

[0227] The server collects multiple data such as user behavior data, purchase history, search history, etc. For example, information about the pages a user visits on a shopping site, the products they purchase, and the queries they search for is obtained via API and stored in a local database. Additionally, the emotion engine analyzes users' text messages and voice inputs to collect emotion data.

[0228] Step 2:

[0229] The server preprocesses the collected data by cleaning it (imputing missing values, removing outliers), standardizing it, and encoding categorical data (for example, standardizing currency units for purchase history and standardizing search queries).

[0230] Step 3:

[0231] The server analyzes the preprocessed data to identify users' interests and preferences. Specifically, it extracts the categories of products frequently purchased by users and the themes of articles they read, and then uses a clustering algorithm to classify users into different clusters. It also takes into account the user's emotional data analyzed by the emotion engine.

[0232] Step 4:

[0233] The server prepares the analysis results to feed into the generative AI model. For example, if the user has the emotion of "joy" in addition to the cluster of "loves the outdoors," that information is input into the generative AI model.

[0234] Step 5:

[0235] The server generates text content and images using a generative AI model. For example, it creates the text "10 Best Hiking Trails" for User B and generates related beautiful landscape images. It also selects positive phrases and vividly colored images to reinforce the user's emotion of "joy."

[0236] Step 6:

[0237] The server places the generated text and images into a landing page template, constructs the page layout using HTML and CSS, and inserts the generated content in the appropriate places, such as placing the title "10 Best Hiking Trails" in the page header, followed by the following text and images, one paragraph at a time.

[0238] Step 7:

[0239] The server deploys the completed landing page to a web server, allowing users to access the landing page through their browsers. For example, User B can enter the URL in his internet browser and view the generated hiking trail page.

[0240] Specific examples

[0241] User data collection

[0242] User B frequently browses articles about buying outdoor gear and hiking using his device. The server collects this information via API and stores his purchase history, visited pages, and search queries in a local database. The emotion engine analyzes text and voice data expressing User B's emotions such as "joy" and "excitement" and stores this emotion data as well.

[0243] Data analysis

[0244] The server cleans and standardizes User B's data before analyzing it. Based on the categories of outdoor equipment purchased and the content of articles viewed, User B is classified into a cluster called "Outdoor Lovers." User B's emotional data, "Joy," is also taken into account.

[0245] Preparing for generative AI

[0246] The server prepares to feed the data to the generative AI model based on the analysis results. The cluster of "outdoor lover" and the emotion information of "joy" are input into the generative AI model.

[0247] Content Generation

[0248] The server uses a generative AI model to generate images related to the text "10 recommended hiking trails for User B." It also selects positive phrases and brightly colored images to amplify User B's emotion of "joy."

[0249] Creating a Landing Page

[0250] The server places the generated text and images into a travel guide template, using HTML and CSS to place "10 Best Hiking Trails" in the page header and insert the generated content for each paragraph.

[0251] Deploy

[0252] The server deploys the completed landing page to the web server. User B can then access the page in a browser, view the generated content, and experience emotional satisfaction. This system makes it possible to provide information optimized for User B's interests and emotions.

[0253] Example 2

[0254] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0255] Current web content generation systems have difficulty efficiently providing personalized content that fully considers each user's emotions and behavioral history. In particular, there is a need for a method that can collect and analyze a variety of user data in real time, automatically generate content that reflects the user's emotions, and provide it as an optimal landing page.

[0256] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0257] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, search history, text messages, and voice data, means for cleaning the collected data, extracting features, and analyzing them, means for analyzing user sentiment from the analyzed data, means for feeding data to a generative AI model based on the analysis results and generating text and images optimized for the user, and means for arranging the generated text and images and automatically creating a landing page, thereby making it possible to efficiently provide an optimal landing page based on the user's sentiments and preferences.

[0258] "Behavioral data" is information about specific actions a user takes on a website or application, such as clicks, views, or purchases.

[0259] "Purchase History" refers to the products and services a User has previously purchased and related details.

[0260] "Search history" refers to the history of queries a user has made on search engines and websites, and is data that indicates the user's interests and concerns.

[0261] "Text Message" means any written communication sent or received by a User on a Digital Platform.

[0262] "Voice Data" refers to voice information input by a user or recorded by a system.

[0263] "Means of collection" refers to the methods and technologies used to acquire and store user behavioral data, purchase history, search history, text messages, voice data, etc. in a system.

[0264] "Cleaning" refers to the process of removing unnecessary information from collected data or filling in gaps in the data.

[0265] "Means for extracting features" refers to techniques and methods for extracting important patterns and information from collected data and using them for analysis.

[0266] "Analysis methods" refers to the algorithms and technologies used to process collected data and identify user preferences and behavioral patterns.

[0267] "Means for analyzing emotions" refers to technology for reading and classifying emotions from users' text messages and voice data.

[0268] A "generative AI model" refers to artificial intelligence technology that generates text and images based on large amounts of data.

[0269] "Landing page" means the web page a user first arrives at after clicking on an internet advertisement, email link, etc.

[0270] "Positioning means" refers to a technique for appropriately positioning the generated text and images on the landing page.

[0271] The system embodying this invention collects and analyzes user behavior data, purchase history, search history, text messages, voice data, etc. to identify the user's hobbies, preferences, and emotions, and automatically generates an optimal landing page based on these. It is particularly characterized by the use of an emotion engine for analyzing user emotions and a generative AI model.

[0272] The server collects data about the various actions users take on websites and applications. Specifically, it periodically retrieves behavioral data, purchase history, and search history via APIs and stores them in a local database. It also collects text messages and voice data sent and received by users within the site in real time.

[0273] The collected data is cleaned by the server, with unnecessary information removed and missing values ​​filled in. Next, features are extracted from the data, such as frequently purchased product categories determined from a user's purchase history. Furthermore, a clustering algorithm is used to classify users into different clusters. This process allows the user's behavioral patterns and preferences to be identified.

[0274] The server then performs sentiment analysis using the collected text messages and voice data. For sentiment analysis, a sentiment analysis model using natural language processing technology is used to identify emotions from the user's text messages and voice data and assign emotion labels.

[0275] Based on the analysis results, the server feeds the data to the generative AI model. This generative AI model includes a text generation model and an image generation model. For example, if a user "likes the outdoors" and has "emotions of joy," that information is input as a prompt to the generative AI model. An example of a prompt is shown below:

[0276] "Users love outdoor activities and feel joy. Can you recommend some hiking trails?"

[0277] The server receives the text and images returned by the generative AI model and uses them to generate content optimized for the user. For example, it generates text content such as "10 Best Hiking Trails" and related images of beautiful scenery. This generated content is then placed on a landing page based on a template pre-configured by the server. Specifically, the page layout is constructed using HTML and CSS, and the generated text and images are inserted in the appropriate locations.

[0278] Finally, the server deploys the completed landing page to the web server, allowing users to access it through their browsers, allowing users to view the optimal landing page based on their preferences and emotions and achieve emotional satisfaction.

[0279] This system is able to efficiently provide optimal landing pages that also take user emotions into consideration through a series of steps, from data collection and analysis, sentiment analysis, AI generation, content placement, and deployment.

[0280] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0281] Step 1: Collect user data

[0282] The server collects data on user behavior on the website, purchase history, search history, text messages, voice data, etc. Specifically, it may use an API to send requests such as the following to obtain data:

[0283] GET / api / user_behavior?user_id={user ID}&event_type=purchase

[0284] This allows details of the products a user purchases, search history, text messages, voice data, etc. to be collected and stored in a local database.

[0285] Input: User behavior data, purchase history, search history, text messages, voice data

[0286] Output: Collected data stored in a local database

[0287] Step 2: Data cleaning and feature extraction

[0288] The server cleans the collected data, removing unnecessary information and formatting it. Specifically, it removes duplicate data and fills in missing values. Next, it extracts features from purchase and search histories. For example, it identifies the categories of products frequently purchased by users.

[0289] python

[0290] import pandas as pd

[0291] data = pd.read_csv('user_data.csv')

[0292] data.drop_duplicates(inplace=True)

[0293] data.fillna(method='ffill', inplace=True)

[0294] popular_categories = data['category'].value_counts().head(5)

[0295] Input: Collected user data

[0296] Output: Cleaned data and extracted features

[0297] Step 3: Cluster the data

[0298] The server classifies users into different clusters based on their features using a clustering algorithm (e.g., K-means), which groups users with similar behavioral patterns.

[0299] python

[0300] from sklearn.cluster import KMeans

[0301] kmeans = KMeans(n_clusters=5)

[0302] data['cluster'] = kmeans.fit_predict(data[['feature1', 'feature2']])

[0303] Input: Cleaned data and extracted features

[0304] Output: User data sorted into clusters

[0305] Step 4: Sentiment Analysis

[0306] The server performs sentiment analysis using text messages and voice data. For sentiment analysis, a sentiment analysis model using natural language processing technology is used. For example, sentiment labels are assigned from text in the following format:

[0307] python

[0308] from transformers import pipeline

[0309] sentiment_analysis = pipeline('sentiment-analysis')

[0310] results = sentiment_analysis("This hike was amazing!")

[0311] Input: Text messages and voice data

[0312] Output: Data with emotion labels

[0313] Step 5: Prepare the generative AI model and generate prompts

[0314] The server feeds data to the generative AI model based on the analysis results, and prepares prompt sentences to input to the generative AI model.

[0315] python

[0316] prompt = "User enjoys outdoor activities and has the emotion joy. What hiking trails do you recommend?"

[0317] ai_model_input = {

[0318] "prompt": prompt,

[0319] "user_cluster": user_cluster,

[0320] "emotion_label": emotion_label

[0321] }

[0322] Input: Cluster labels and emotion labels

[0323] Output: Prompt and data to be fed into the generative AI model

[0324] Step 6: Content Generation

[0325] The server generates text and images using generative AI models, for example, it gets the text content "10 Best Hiking Trails" from a text generation model and related scenic images from an image generation model.

[0326] python

[0327] response = ai_model.generate(prompt)

[0328] Image generation

[0329] image_prompt = "Beautiful hiking trail views"

[0330] generated_image = image_model.generate(image_prompt)

[0331] Input: Prompt statement (text and image)

[0332] Output: Generated text and images

[0333] Step 7: Create a Landing Page

[0334] The server places the generated text and images into a landing page based on a pre-defined template, and the page layout is built using HTML and CSS.

[0335] html

[0336]

[0337]

[0338] <title> Recommended hiking trails< / title>

[0339]

[0340]

[0341] <h1> 10 Great Hiking Trails< / h1>

[0342] Text content...

[0343]

[0344]

[0345]

[0346] Input: Generated text and images

[0347] Output: The created landing page

[0348] Step 8: Deploy

[0349] The server deploys the completed landing page to a web server, which allows users to access the landing page through their browsers.

[0350] shell

[0351] scp landing_page.html user@webserver: / var / www / html /

[0352] Input: The created landing page

[0353] Output: Deployed landing page

[0354] Through these steps, the optimal landing page based on the user's emotions and preferences is efficiently provided.

[0355] (Application example 2)

[0356] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0357] Conventional landing page generation systems and product recommendation systems perform analysis based on user behavior data and purchase history, but there is a need to achieve more accurate individual optimization by incorporating the user's emotional state. Specifically, there is a problem in that it is difficult to provide optimal content based on the user's physical emotional state. This has resulted in many cases where these systems are ineffective in improving the user experience or stimulating purchasing motivation.

[0358] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0359] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, and search history, means for analyzing the data to identify the user's hobbies, preferences, and emotional state, means for feeding the data to a generative AI model based on the analysis results to generate text and images optimized for the user, and means for arranging the generated text and images to automatically create a landing page or a recommended product list, thereby enabling the provision of individually optimized content that also takes into account the user's emotional state.

[0360] "User behavior data" refers to a record of the actions and behavior a user performs online, including, specifically, information such as the web pages visited, the links clicked, and the duration of viewing.

[0361] "Purchase history" refers to a record of products purchased by a user in the past, and includes information such as product name, purchase date and time, quantity, total amount, etc.

[0362] "Search history" refers to a record of search queries a user has made on the Internet, including the keywords searched, the date and time of the search, and click information for search results.

[0363] "Means of collection" refers to the methods and technologies used to obtain user behavioral data, purchase history, search history, etc., and includes data acquisition using APIs and tracking using cookies.

[0364] "Means of analysis" refers to methods and techniques for internally processing collected data and extracting useful information, including techniques such as data cleaning, feature extraction, and clustering.

[0365] "Hobbies and preferences" refers to areas or categories in which a user has particular interests or concerns, including preferences for specific products and interest in specific activities.

[0366] An "emotional state" refers to the emotional state a user feels at a particular time or in a particular situation, and includes emotions such as joy, excitement, sadness, and anger.

[0367] "Generative AI model" refers to a model that uses artificial intelligence to generate new content from specific input data, and includes technologies used for text generation and image generation (e.g., GPT and DALL-E).

[0368] "Means of feeding" refers to the methods and technologies for supplying analysis results and other necessary data to the generative AI model, including data preprocessing and input format adjustment.

[0369] "Text and Images" means the part of the content provided to users, including written and visual information.

[0370] A "landing page" refers to a web page accessed when an advertisement or link is clicked, and includes pages that provide specific information about a particular product or service.

[0371] A "recommended product list" is a series of products recommended to users based on their interests and is optimized based on their past purchasing history, behavioral data, emotional state, etc.

[0372] "Means of automatic creation" refers to methods and technologies that allow a system to automatically generate and arrange content without requiring analog manual work, including template-based arrangement and automatic layout generation.

[0373] This invention is a system that collects and analyzes data such as user behavior data, purchase history, and search history on the Internet, and provides individually optimized content that also takes into account emotional state. Specific implementation methods for realizing this system are described below.

[0374] Collection Stage

[0375] The server uses APIs to collect multiple data such as user behavior, purchase history, and search history. This data includes detailed information such as which web pages users visited, which links they clicked, and which product reviews they read. The collected data is stored in a local database for further analysis.

[0376] Analysis stage

[0377] The server cleans the collected data and extracts features. This includes removing noise and standardizing the format. It then uses a clustering algorithm and a sentiment analysis engine to classify users into different clusters and identify their hobbies, preferences, and interests. At the same time, it analyzes the user's text messages and voice inputs to understand their emotional state. Once this analysis is complete, it can be determined that the user is currently in a specific emotional state, such as "joy" or "excitement."

[0378] Generation stage

[0379] Based on the analysis results, the server feeds data to a generative AI model. For example, the user may input information such as "I love the outdoors" and being in a "joy" emotional state. This generative AI uses large-scale language models and image generation models (e.g., GPT-4 and DALL-E). The generative AI model generates text and images optimized for the user. For example, it generates text content such as "Top 10 recommended hiking trails" and images depicting the beautiful scenery of those hiking trails.

[0380] Placement and display stage

[0381] The generated text and images are placed in the appropriate positions based on a pre-defined template. The server uses these artifacts to automatically create a landing page and a list of recommended products. The template uses HTML and CSS to maintain a consistent design. The final landing page is deployed to a web server, where it can be accessed by users through a browser.

[0382] Specific examples

[0383] User A frequently purchases outdoor equipment from a specific shopping site and also frequently reads articles about hiking. After collecting this user's data and conducting clustering and sentiment analysis, the user was classified into a cluster called "Outdoors Lover" and identified as being in a "Joy" emotional state. The server fed this information into a generative AI model, which generated images of beautiful scenery associated with the text "10 Recommended Hiking Trails." This information was then arranged as a landing page based on a template and deployed.

[0384] Prompt Sentence Examples

[0385] An example of a prompt sentence to be input to the generative AI model is, "Create the optimal landing page for a user who is interested in outdoor gear and is currently in the emotional state of 'joy'."

[0386] In this way, this invention takes into consideration the emotional state of the user and generates and provides individually optimized content, thereby significantly improving the user experience.

[0387] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0388] Step 1: Collect user data

[0389] The server uses an API to collect multiple data such as user behavior data, purchase history, and search history. Specific operations include obtaining web page log data, clickstreams, and detailed data on viewed products. The input is data related to the user's online operations, and the output is an organized set of user behavior data.

[0390] Step 2: Data cleaning and feature extraction

[0391] The server cleans the collected data and extracts features. Specifically, it removes noise and standardizes the data format. The input is raw user data, and the output is clean data and feature-extracted data.

[0392] Step 3: Clustering and sentiment analysis

[0393] The server performs clustering and sentiment analysis on the cleaned data. Using a clustering algorithm, users are classified into different clusters. It also uses a sentiment analysis engine to analyze users' text messages and voice data to identify their emotional state. The inputs are the cleaned data and feature-extracted data, and the output is cluster information and emotional state information.

[0394] Step 4: Feed data into the generative AI model

[0395] The server feeds cluster information and emotional state information to the generative AI model. For example, input information for a user who is an "outdoor lover" and has the emotion of "joy." The input is the cluster information and emotional state information, and the output is the input data for the generative AI model.

[0396] Step 5: Generate optimized content

[0397] The server uses a generative AI model to generate text and images optimized for the user. For example, it generates text content such as "Top 10 recommended hiking trails" and images of beautiful scenery. The input is the input data to the generative AI model, and the output is the generated text and images. For generation, a text generation model such as GPT-4 and an image generation model such as DALL-E are used.

[0398] Step 6: Arranging Content

[0399] The server places the generated text and images in the appropriate locations based on a pre-defined template. Specifically, it uses HTML and CSS to build a landing page or recommended product list. The input is the generated text and images, and the output is the HTML code for the completed web page.

[0400] Step 7: Deploy your landing page

[0401] The server deploys the completed landing page and recommended product list to a web server, where users can access these pages through their browsers. The input is the HTML code for the completed web page, and the output is a landing page accessible on the web.

[0402] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0403] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0404] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0405] [Second embodiment]

[0406] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0407] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0408] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0409] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0410] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0411] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0412] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0413] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0414] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0415] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0416] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0417] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0418] The system that embodies this invention mainly consists of multiple servers, terminals, and users. The server analyzes various user data, and based on the results, it uses a generation AI to automatically generate the landing page (LP) that is most suitable for the user. This system quickly and efficiently provides optimal content based on the user's interests and preferences.

[0419] Program processing

[0420] User data collection

[0421] The server collects data such as user behavior, purchase history, and search history. Specifically, information such as the pages the user visited on the Internet, the products they purchased, and the search queries they performed is obtained via API. This allows the server to store detailed user preference information in a local database.

[0422] Data analysis

[0423] The server analyzes the collected data. This analysis includes preprocessing steps such as data cleansing and feature extraction. For example, categories of products frequently purchased by a user are extracted from their purchase history, and this is then passed through a clustering algorithm to classify users into multiple clusters. This makes it clear what interests the user has.

[0424] Preparing for generative AI

[0425] The server prepares the analysis results to be fed into a generative AI model. For example, a natural language processing model is used for text generation, and an image generation model is used for image generation. The generative AI model receives input such as user features and cluster information.

[0426] Content Generation

[0427] The server uses a generative AI model to generate text and images optimized for each user. For example, to generate a travel guide article for "User A," the server generates text including a title such as "Top 10 Adventure Travel Destinations" and detailed information about the adventure trip. In addition, related beautiful landscape images are also generated.

[0428] Creating a Landing Page

[0429] The server creates a landing page based on the generated content. Specifically, it inserts the generated text and images into a pre-defined template, builds the page structure using HTML and CSS, and finally deploys the page on a web server so that users can access it through their browsers.

[0430] Specific examples

[0431] User data collection

[0432] User B frequently purchases outdoor equipment from a shopping site and frequently reads articles about hiking using a certain device. The server collects this data via an API and stores it in a local database.

[0433] Data analysis

[0434] The server analyzes User B's data and classifies him / her into a cluster called "outdoor enthusiasts" based on the categories of outdoor equipment he / she has purchased and the content of articles he / she has viewed. It then identifies the areas and activities in which User B is particularly interested.

[0435] Preparing for generative AI

[0436] The server feeds the analysis results into the generative AI model and prepares to generate optimal content for User B. Specifically, it inputs data indicating that User B is interested in hiking into the generative AI.

[0437] Content Generation

[0438] The server uses generative AI to generate text content such as "Top 10 hiking trails recommended for User B." It also generates beautiful landscape images related to each hiking trail.

[0439] Creating a Landing Page

[0440] The server places the generated text and images into a template to create a landing page for the hiking guide, and finally deploys this page to a web server, making it accessible to User B.

[0441] In this way, the system can efficiently provide optimized information to User B.

[0442] The processing flow will be explained below.

[0443] Step 1:

[0444] The server collects data such as user behavior, purchase history, and search history. For example, information about the pages a user visits on a shopping site, the products they purchase, and the queries they search for is obtained via API and stored in a local database. This allows detailed information about the user's preferences to be gathered.

[0445] Step 2:

[0446] The server performs preprocessing, cleaning the data. Specifically, it fills in missing values ​​and removes outliers. It also standardizes the data format and standardizes each data point. For example, it converts all purchase history prices into the same currency unit.

[0447] Step 3:

[0448] The server analyzes the preprocessed data and extracts features to identify the user's interests and preferences. For example, features could include the product categories the user frequently purchases or the themes of the articles they view. Text analysis and category classification techniques are used to extract these features.

[0449] Step 4:

[0450] The server uses a clustering algorithm to group users, for example, using the K-means algorithm to classify users with similar interests into the same cluster. This clustering helps clarify the areas of interest and concern of users.

[0451] Step 5:

[0452] The server then feeds the data to a generative AI model based on the analysis results. For example, if a user belongs to a cluster called "Outdoors Lovers," that information is input to the generative AI. These generative AI models include text generation models (e.g., GPT-4) and image generation models (e.g., DALL-E).

[0453] Step 6:

[0454] The server uses the generative AI model to generate text and images optimized for each user. For example, for "User B," it generates text content such as "The 10 Best Hiking Trails" and related beautiful scenery images.

[0455] Step 7:

[0456] The server places the generated text and images into a landing page template, building the page layout using HTML and CSS and inserting the generated content in the appropriate places, resulting in a landing page with a consistent design.

[0457] Step 8:

[0458] The server deploys the completed landing page to the web server, allowing users to access the landing page through their browsers. User B then views the "10 Best Hiking Trails" page.

[0459] In this way, the system collects data, analyzes it, generates AI, places content, and deploys it step by step, quickly and efficiently providing users with optimized landing pages.

[0460] Example 1

[0461] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0462] With current technology, efficiently generating landing pages tailored to user preferences requires a lot of manual work, making it difficult to quickly provide optimal content.In addition, there are limited methods for effectively utilizing user behavior data, purchase history, search history, etc., making it difficult to generate personalized content for users.

[0463] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0464] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, and search history, means for analyzing the data and identifying the user's interests and preferences, means for generating prompt sentences based on the analysis results and feeding the data to a generative AI model, means for generating text and images optimized for the user using the generative AI model, and means for arranging the generated text and images to automatically create a landing page. This makes it possible to quickly and automatically generate an optimal landing page based on the user's preferences and efficiently provide content.

[0465] "User" means an end user of the Internet service or website.

[0466] "Behavioral data" refers to data such as page browsing history and click history when a user visits a website.

[0467] "Purchase history" refers to historical data about products a user has previously purchased from online shops, etc.

[0468] "Search history" refers to historical data about queries and search results a user has performed on the Internet.

[0469] "Data analysis" refers to a series of processes that aggregate, classify, and evaluate collected data.

[0470] "Hobbies and preferences" refers to data that indicates a user's interests, concerns, and preferences.

[0471] A "prompt" is an instruction given to a generative AI model.

[0472] A "generative AI model" is an artificial intelligence model that generates text or images from specified input data.

[0473] "Text" refers to sentences or strings of characters generated by a generative AI model.

[0474] "Image" refers to the visual data generated by a generative AI model.

[0475] A "landing page" is a web page created for a user to access in a web browser.

[0476] A "template" is a predefined layout or design framework used to automatically create a landing page.

[0477] "Deployment" is the process of placing a completed landing page on a web server so that it can be accessed by users.

[0478] A system embodying this invention efficiently collects and analyzes multiple data sets, such as user behavior data, purchase history, and search history, and automatically generates content optimized for each user using a generative AI model. Specific embodiments of this system are described below.

[0479] First, the server collects information such as user behavior data, purchase history, and search history. This data is obtained from the user's browser history, purchase history, search queries, etc. recorded by the device. The server obtains this data via API and stores it in a local database. For example, if a user purchases outdoor equipment on a shopping site or reads an article about hiking, that information will be collected.

[0480] The server then analyzes the collected data. At this stage, data cleansing is performed to remove noise and missing data. Feature extraction is also performed to identify users' interests based on their purchase history and browser history. For example, a clustering algorithm is used to separate the data into multiple clusters, categorizing users into clusters such as "outdoor enthusiasts" or "travel enthusiasts."

[0481] Based on the analysis results, the server generates a prompt sentence and prepares to feed it to the generative AI model. For example, if the analysis finds that the user has specific interests, it generates a prompt sentence such as, "Please tell me some recommended hiking trails for User A." The server then creates an instance of a generative AI model (e.g., GPT-3 or StyleGAN) and sets the prompt sentence as input.

[0482] The server inputs prompts into the generative AI model to generate text and images optimized for the user. For example, the generative AI model generates text content such as "Top 10 recommended hiking trails for user A" and also creates related beautiful landscape images. This automatically generates content that is individually optimized for each user.

[0483] Finally, the server creates a landing page based on the generated text and images. At this stage, the generated content is inserted according to a pre-defined template, and the page structure is built using HTML and CSS. The completed landing page is deployed to a web server, where users can access it through their browsers.

[0484] To illustrate how this system works, let's use a practical example. If User B is buying outdoor gear on a shopping site and reading an article about hiking, the server will collect this behavioral data and identify the user's interest as "outdoors enthusiast." As a result, a prompt message will be created that generates "10 recommended hiking trails for User B," and specific content will be generated by the generative AI model. This generated content will be placed on a landing page based on a template and deployed for User B's easy access.

[0485] An example of a prompt sentence is, "Please generate content that introduces hiking trails that User B is interested in. User B has recently purchased a lot of outdoor equipment and particularly enjoys hiking."

[0486] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0487] Step 1: Data collection

[0488] The server collects information such as user behavior data, purchase history, and search history. Specifically, the device sends the user's browser history, purchase records, and search queries to the server via API. The server stores this data in a local database. For example, if a user purchases outdoor equipment on a shopping site or reads an article about hiking, that information will be collected.

[0489] Input: User behavior data, purchase history, search history

[0490] Output: User data stored in a local database

[0491] Step 2: Data analysis

[0492] The server retrieves collected user data from a local database and first cleanses the data, removing missing data and noise. Next, it performs feature extraction to identify users' interests based on their purchase history and browser history. It then uses a clustering algorithm to categorize users into clusters such as "outdoor enthusiasts" or "travel enthusiasts."

[0493] Input: User data from a local database

[0494] Output: User features and cluster information

[0495] Step 3: Prompt generation

[0496] The server generates a prompt based on the results of the data analysis. For example, if the server finds that the user has a specific interest, it creates a prompt such as, "Please recommend some hiking trails for User A." This prepares the instruction sentence to be given to the generative AI model.

[0497] Input: Data analysis results (user features and cluster information)

[0498] Output: Generated prompt statement

[0499] Step 4: Input to the generative AI model

[0500] The server creates an instance of a generative AI model (e.g., GPT-3 or StyleGAN) and sets the prompt sentence generated in the previous step as input, so the model is ready to generate optimal content.

[0501] Input: Generated prompt text

[0502] Output: Input data to a generative AI model

[0503] Step 5: Content Generation

[0504] The server inputs the prompt into the generative AI model, which generates text and images optimized for the user. The generative AI model then generates specific content based on the prompt. For example, it generates text content such as "Top 10 recommended hiking trails for user A" and related beautiful landscape images.

[0505] Input: Prompt sentence for generative AI model

[0506] Output: Generated text and images

[0507] Step 6: Create a Landing Page

[0508] The server creates a landing page by arranging the generated text and images according to a pre-defined template, building the page structure using HTML and CSS, and inserting the generated content. The completed landing page is then deployed to a web server, where users can access it through their browsers.

[0509] Input: Generated text and images

[0510] Output: The generated landing page

[0511] This process flow allows the system to quickly and efficiently generate optimal landing pages based on user preferences.

[0512] (Application example 1)

[0513] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0514] Conventional internet advertising and landing page creation systems have not adequately provided optimal content based on users' hobbies and preferences, and have had difficulty providing content in real time. As a result, advertisements and content that do not match the user's interests are displayed, making effective marketing impossible. Furthermore, even when displaying advertisements using wearable devices such as smart glasses, there is a lack of a means to provide users with content optimized for them in real time. There is a need to solve these issues.

[0515] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0516] In this invention, the server includes a means for collecting multiple data such as user behavior data, purchase history, and search history, a means for analyzing the data to identify the user's interests and preferences, and a means for feeding the data to a generative AI model based on the analysis results to generate text and images optimized for the user, thereby arranging the generated text and images to automatically create landing pages or advertising content and display them on the smart glasses.

[0517] "User behavior data" refers to information such as the pages a user visits on the Internet, the links they click, and the time they spend browsing.

[0518] "Purchase history" refers to information such as the products and services a user has purchased in the past, the date and time of those purchases, and frequency of those purchases.

[0519] "Search history" refers to the search queries a user has made on a search engine and the history of the related search result pages they have visited.

[0520] A "generative AI model" is an artificial intelligence model trained on a large dataset, and refers to a technique for generating text or images based on specific input data.

[0521] A "clustering algorithm" refers to a statistical method for analyzing large amounts of data and grouping data that are highly similar.

[0522] A "landing page" is a web page that appears first when a user clicks on an advertisement or link, and contains content designed to encourage a specific action (such as purchasing a product or entering information).

[0523] "Smart glasses" are a type of wearable device that has the shape of glasses but has the ability to display visual information as augmented reality (AR).

[0524] "Template" means a document or web page structure with a specific format or style predefined for efficient layout and display of content.

[0525] The system for implementing this invention consists of a server, a terminal, and a user. The server collects and analyzes information such as user behavior data, purchase history, and search history, and uses AI to display optimal landing pages and advertising content on the smart glasses based on the results.

[0526] Program Overview

[0527] The system includes the following procedures:

[0528] 1. User data collection:

[0529] The server collects data such as user behavior, purchase history, and search history via API, and stores detailed user preference information in a local database.

[0530] 2. Data Analysis:

[0531] The server cleanses the collected data, extracts and analyzes features, and uses a clustering algorithm to classify users into different clusters based on their purchase history and the pages they have viewed.

[0532] 3. Prepare the generative AI:

[0533] The server creates a prompt sentence to feed the analysis results to a generative AI model (e.g., GPT-3), which includes the user's cluster information and interests.

[0534] 4. Advertising content generation:

[0535] A generative AI model generates optimized text and images based on the prompt, and the generated ad content is tailored to what the user should see.

[0536] 5. Display on smart glasses:

[0537] The server displays the generated content on the smart glasses in real time, allowing users to visually view the most suitable advertising content through the glasses they are wearing.

[0538] Specific examples

[0539] Let's say a user is walking through a particular shopping mall. The server determines from past behavioral data and purchase history that the user is interested in outdoor gear. Based on this analysis, it inputs the following prompt into the generative AI model:

[0540] plaintext

[0541] "Generate ads that are best suited to users in Cluster 1, whose primary interest is outdoor gear. Include current sales and popular items in the ad content."

[0542] Based on this prompt, the generative AI model generates advertising text, such as "The latest outdoor gear is on sale!", along with images of the products on sale. The server then displays this generated advertising content on the smart glasses in real time, providing users with the most appropriate information.

[0543] Hardware and software used

[0544] Hardware: Smart glasses (e.g. HoloLens, Google Glass)

[0545] software:

[0546] Data collection via API: RESTful API

[0547] Data analysis: NumPy, scikit-learn

[0548] Generative AI model: GPT-3 (OpenAI)

[0549] This makes it possible to provide advertising content that matches the user's preferences in real time.

[0550] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0551] Step 1:

[0552] The server collects user behavioral data, purchase history, and search history via API. Specifically, the server accesses each API endpoint and retrieves data. For example, behavioral data including the user's visited pages and browsing time, history of past purchases, and search engine query history are collected. This data is stored in a local database.

[0553] Input: API endpoint

[0554] Output: User data stored in a local database

[0555] Step 2:

[0556] The server cleanses the collected data and extracts features. First, it removes duplicate data and missing values ​​to clean the data. Next, it extracts the categories of products that users frequently purchase and the features of their search queries. For example, if a user frequently purchases outdoor equipment, the features of that category are extracted.

[0557] Input: Raw data in a local database

[0558] Output: Cleansed feature data

[0559] Step 3:

[0560] The server runs a clustering algorithm on the cleansed data to classify users into different clusters. For example, it uses KMeans clustering to classify users into clusters such as "outdoor enthusiasts" and "tech gadget enthusiasts." This identifies the user's interests and preferences.

[0561] Input: Cleansed feature data

[0562] Output: User data categorized into clusters

[0563] Step 4:

[0564] The server then creates prompts to be input into the generative AI model based on the clustered user data. For example, for a user who loves the outdoors, the server might generate a prompt like, "Generate ads that are optimal for users in cluster 1. Their primary interest is outdoor gear. Include current sales information and popular products in the ad content."

[0565] Input: User data categorized into clusters

[0566] Output: A prompt to be input to the generative AI model

[0567] Step 5:

[0568] The server inputs a prompt into a generative AI model (e.g., GPT-3) to generate text and images. Based on the prompt, the generative AI model generates optimal ad text and related images. For example, ad text such as "Latest outdoor gear on sale!" and images of outdoor gear on sale are generated.

[0569] Input: prompt statement

[0570] Output: Generated ad text and images

[0571] Step 6:

[0572] The server transmits the generated advertising text and images to the smart glasses for display in real time, and the smart glasses visually display the received advertising content, allowing the user to easily check the information.

[0573] Input: Generated ad text and image

[0574] Output: Advertising content displayed on smart glasses

[0575] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0576] The system for implementing this invention comprises multiple servers, terminals, and users. This system includes an emotion engine that analyzes various user data and uses generative AI to automatically generate the optimal landing page (LP) for the user. This emotion engine analyzes the user's text messages, voice input, images, and videos, identifies the user's emotions, and generates optimized content. This allows for efficient provision of content that takes user emotions into consideration.

[0577] Program processing

[0578] User data collection

[0579] The server collects multiple data sets, including user behavior data, purchase history, and search history. Specifically, information about the pages the user visited, the products they purchased, and the queries they searched is obtained via API and stored in a local database. It also uses an emotion engine to collect the emotions expressed by the user during interactions.

[0580] Data analysis

[0581] The server analyzes the collected data. This analysis includes data cleaning and feature extraction. For example, it extracts the categories of products frequently purchased by users from their purchase history and classifies users into different clusters using a clustering algorithm. Furthermore, it uses an emotion engine to analyze emotions from users' text messages and voice inputs to identify what content is more appropriate.

[0582] Preparing for generative AI

[0583] The server feeds data to a generative AI model based on the analysis results. For example, if a user has a cluster of "loves the outdoors" and an emotion of "joy," that information is input into the generative AI. These generative AI models include text generation models (e.g., GPT-4) and image generation models (e.g., DALL-E).

[0584] Content Generation

[0585] The server uses a generative AI model to generate text and images optimized for each user. For example, for "User B," it generates text content such as "The 10 Best Hiking Trails," along with related images of beautiful scenery that evoke a sense of enjoyment. Furthermore, if the user expresses the emotion of "joy," it selects positive words and images that further enhance this emotion.

[0586] Creating a Landing Page

[0587] The server places the generated text and images into a landing page template, building the page layout using HTML and CSS and inserting the generated content in the appropriate places, resulting in a landing page with a consistent design.

[0588] Deploy

[0589] The server deploys the completed landing page to the web server, which allows users to access the landing page through their browsers. User B views the "10 Best Hiking Trails" page, which is designed to provide emotional satisfaction.

[0590] Specific examples

[0591] User data collection

[0592] User B frequently purchases outdoor equipment on a shopping site and reads articles about hiking using a certain device. User B expresses emotions such as "joy" and "excitement" while interacting with the site. The server collects this information via an API and stores it in a local database.

[0593] Data analysis

[0594] The server analyzes User B's data and identifies a cluster called "outdoor enthusiasts" based on the categories of outdoor equipment purchased and the content of articles viewed. It also analyzes User B's emotional data using an emotion engine to identify "joy" and "excitement."

[0595] Preparing for generative AI

[0596] The server feeds the analysis results into a generative AI model. For example, it inputs the cluster "loves the outdoors" and the emotion information "joy." The generative AI generates optimal text and images based on a large amount of data.

[0597] Content Generation

[0598] The server uses the generative AI model to generate text content such as "10 recommended hiking trails for User B." It also generates beautiful scenic images of hiking trails that User B is interested in. It uses positive words and colors that reinforce User B's emotion of "joy."

[0599] Creating a Landing Page

[0600] The server places the generated text and images into a template, inserting the generated content into a travel guide template and building the page layout.

[0601] Deploy

[0602] The server deploys the completed landing page to the web server, and User B accesses this page through a browser, which also provides emotional satisfaction.

[0603] In this way, the system collects data, analyzes it, analyzes emotions, generates AI, places content, and deploys it at each step, enabling it to quickly and efficiently provide optimal landing pages that also take user emotions into consideration.

[0604] The processing flow will be explained below.

[0605] Step 1:

[0606] The server collects multiple data such as user behavior data, purchase history, search history, etc. For example, information about the pages a user visits on a shopping site, the products they purchase, and the queries they search for is obtained via API and stored in a local database. Additionally, the emotion engine analyzes users' text messages and voice inputs to collect emotion data.

[0607] Step 2:

[0608] The server preprocesses the collected data by cleaning it (imputing missing values, removing outliers), standardizing it, and encoding categorical data (for example, standardizing currency units for purchase history and standardizing search queries).

[0609] Step 3:

[0610] The server analyzes the preprocessed data to identify users' interests and preferences. Specifically, it extracts the categories of products frequently purchased by users and the themes of articles they read, and then uses a clustering algorithm to classify users into different clusters. It also takes into account the user's emotional data analyzed by the emotion engine.

[0611] Step 4:

[0612] The server prepares the analysis results to feed into the generative AI model. For example, if the user has the emotion of "joy" in addition to the cluster of "loves the outdoors," that information is input into the generative AI model.

[0613] Step 5:

[0614] The server generates text content and images using a generative AI model. For example, it creates the text "10 Best Hiking Trails" for User B and generates related beautiful landscape images. It also selects positive phrases and vividly colored images to reinforce the user's emotion of "joy."

[0615] Step 6:

[0616] The server places the generated text and images into a landing page template, constructs the page layout using HTML and CSS, and inserts the generated content in the appropriate places, such as placing the title "10 Best Hiking Trails" in the page header, followed by the following text and images, one paragraph at a time.

[0617] Step 7:

[0618] The server deploys the completed landing page to a web server, allowing users to access the landing page through their browsers. For example, User B can enter the URL in his internet browser and view the generated hiking trail page.

[0619] Specific examples

[0620] User data collection

[0621] User B frequently browses articles about buying outdoor gear and hiking using his device. The server collects this information via API and stores his purchase history, visited pages, and search queries in a local database. The emotion engine analyzes text and voice data expressing User B's emotions such as "joy" and "excitement" and stores this emotion data as well.

[0622] Data analysis

[0623] The server cleans and standardizes User B's data before analyzing it. Based on the categories of outdoor equipment purchased and the content of articles viewed, User B is classified into a cluster called "Outdoor Lovers." User B's emotional data, "Joy," is also taken into account.

[0624] Preparing for generative AI

[0625] The server prepares to feed the data to the generative AI model based on the analysis results. The cluster of "outdoor lover" and the emotion information of "joy" are input into the generative AI model.

[0626] Content Generation

[0627] The server uses a generative AI model to generate images related to the text "10 recommended hiking trails for User B." It also selects positive phrases and brightly colored images to amplify User B's emotion of "joy."

[0628] Creating a Landing Page

[0629] The server places the generated text and images into a travel guide template, using HTML and CSS to place "10 Best Hiking Trails" in the page header and insert the generated content for each paragraph.

[0630] Deploy

[0631] The server deploys the completed landing page to the web server. User B can then access the page in a browser, view the generated content, and experience emotional satisfaction. This system makes it possible to provide information optimized for User B's interests and emotions.

[0632] Example 2

[0633] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0634] Current web content generation systems have difficulty efficiently providing personalized content that fully considers each user's emotions and behavioral history. In particular, there is a need for a method that can collect and analyze a variety of user data in real time, automatically generate content that reflects the user's emotions, and provide it as an optimal landing page.

[0635] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0636] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, search history, text messages, and voice data, means for cleaning the collected data, extracting features, and analyzing them, means for analyzing user sentiment from the analyzed data, means for feeding data to a generative AI model based on the analysis results and generating text and images optimized for the user, and means for arranging the generated text and images and automatically creating a landing page, thereby making it possible to efficiently provide an optimal landing page based on the user's sentiments and preferences.

[0637] "Behavioral data" is information about specific actions a user takes on a website or application, such as clicks, views, or purchases.

[0638] "Purchase History" refers to the products and services a User has previously purchased and related details.

[0639] "Search history" refers to the history of queries a user has made on search engines and websites, and is data that indicates the user's interests and concerns.

[0640] "Text Message" means any written communication sent or received by a User on a Digital Platform.

[0641] "Voice Data" refers to voice information input by a user or recorded by a system.

[0642] "Means of collection" refers to the methods and technologies used to acquire and store user behavioral data, purchase history, search history, text messages, voice data, etc. in a system.

[0643] "Cleaning" refers to the process of removing unnecessary information from collected data or filling in gaps in the data.

[0644] "Means for extracting features" refers to techniques and methods for extracting important patterns and information from collected data and using them for analysis.

[0645] "Analysis methods" refers to the algorithms and technologies used to process collected data and identify user preferences and behavioral patterns.

[0646] "Means for analyzing emotions" refers to technology for reading and classifying emotions from users' text messages and voice data.

[0647] A "generative AI model" refers to artificial intelligence technology that generates text and images based on large amounts of data.

[0648] "Landing page" means the web page a user first arrives at after clicking on an internet advertisement, email link, etc.

[0649] "Positioning means" refers to a technique for appropriately positioning the generated text and images on the landing page.

[0650] The system embodying this invention collects and analyzes user behavior data, purchase history, search history, text messages, voice data, etc. to identify the user's hobbies, preferences, and emotions, and automatically generates an optimal landing page based on these. It is particularly characterized by the use of an emotion engine for analyzing user emotions and a generative AI model.

[0651] The server collects data about the various actions users take on websites and applications. Specifically, it periodically retrieves behavioral data, purchase history, and search history via APIs and stores them in a local database. It also collects text messages and voice data sent and received by users within the site in real time.

[0652] The collected data is cleaned by the server, with unnecessary information removed and missing values ​​filled in. Next, features are extracted from the data, such as frequently purchased product categories determined from a user's purchase history. Furthermore, a clustering algorithm is used to classify users into different clusters. This process allows the user's behavioral patterns and preferences to be identified.

[0653] The server then performs sentiment analysis using the collected text messages and voice data. For sentiment analysis, a sentiment analysis model using natural language processing technology is used to identify emotions from the user's text messages and voice data and assign emotion labels.

[0654] Based on the analysis results, the server feeds the data to the generative AI model. This generative AI model includes a text generation model and an image generation model. For example, if a user "likes the outdoors" and has "emotions of joy," that information is input as a prompt to the generative AI model. An example of a prompt is shown below:

[0655] "Users love outdoor activities and feel joy. Can you recommend some hiking trails?"

[0656] The server receives the text and images returned by the generative AI model and uses them to generate content optimized for the user. For example, it generates text content such as "10 Best Hiking Trails" and related images of beautiful scenery. This generated content is then placed on a landing page based on a template pre-configured by the server. Specifically, the page layout is constructed using HTML and CSS, and the generated text and images are inserted in the appropriate locations.

[0657] Finally, the server deploys the completed landing page to the web server, allowing users to access it through their browsers, allowing users to view the optimal landing page based on their preferences and emotions and achieve emotional satisfaction.

[0658] This system is able to efficiently provide optimal landing pages that also take user emotions into consideration through a series of steps, from data collection and analysis, sentiment analysis, AI generation, content placement, and deployment.

[0659] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0660] Step 1: Collect user data

[0661] The server collects data on user behavior on the website, purchase history, search history, text messages, voice data, etc. Specifically, it may use an API to send requests such as the following to obtain data:

[0662] GET / api / user_behavior?user_id={user ID}&event_type=purchase

[0663] This allows details of the products a user purchases, search history, text messages, voice data, etc. to be collected and stored in a local database.

[0664] Input: User behavior data, purchase history, search history, text messages, voice data

[0665] Output: Collected data stored in a local database

[0666] Step 2: Data cleaning and feature extraction

[0667] The server cleans the collected data, removing unnecessary information and formatting it. Specifically, it removes duplicate data and fills in missing values. Next, it extracts features from purchase and search histories. For example, it identifies the categories of products frequently purchased by users.

[0668] python

[0669] import pandas as pd

[0670] data = pd.read_csv('user_data.csv')

[0671] data.drop_duplicates(inplace=True)

[0672] data.fillna(method='ffill', inplace=True)

[0673] popular_categories = data['category'].value_counts().head(5)

[0674] Input: Collected user data

[0675] Output: Cleaned data and extracted features

[0676] Step 3: Cluster the data

[0677] The server classifies users into different clusters based on their features using a clustering algorithm (e.g., K-means), which groups users with similar behavioral patterns.

[0678] python

[0679] from sklearn.cluster import KMeans

[0680] kmeans = KMeans(n_clusters=5)

[0681] data['cluster'] = kmeans.fit_predict(data[['feature1', 'feature2']])

[0682] Input: Cleaned data and extracted features

[0683] Output: User data sorted into clusters

[0684] Step 4: Sentiment Analysis

[0685] The server performs sentiment analysis using text messages and voice data. For sentiment analysis, a sentiment analysis model using natural language processing technology is used. For example, sentiment labels are assigned from text in the following format:

[0686] python

[0687] from transformers import pipeline

[0688] sentiment_analysis = pipeline('sentiment-analysis')

[0689] results = sentiment_analysis("This hike was amazing!")

[0690] Input: Text messages and voice data

[0691] Output: Data with emotion labels

[0692] Step 5: Prepare the generative AI model and generate prompts

[0693] The server feeds data to the generative AI model based on the analysis results, and prepares prompt sentences to input to the generative AI model.

[0694] python

[0695] prompt = "User enjoys outdoor activities and has the emotion joy. What hiking trails do you recommend?"

[0696] ai_model_input = {

[0697] "prompt": prompt,

[0698] "user_cluster": user_cluster,

[0699] "emotion_label": emotion_label

[0700] }

[0701] Input: Cluster labels and emotion labels

[0702] Output: Prompt and data to be fed into the generative AI model

[0703] Step 6: Content Generation

[0704] The server generates text and images using generative AI models, for example, it gets the text content "10 Best Hiking Trails" from a text generation model and related scenic images from an image generation model.

[0705] python

[0706] response = ai_model.generate(prompt)

[0707] Image generation

[0708] image_prompt = "Beautiful hiking trail views"

[0709] generated_image = image_model.generate(image_prompt)

[0710] Input: Prompt statement (text and image)

[0711] Output: Generated text and images

[0712] Step 7: Create a Landing Page

[0713] The server places the generated text and images into a landing page based on a pre-defined template, and the page layout is built using HTML and CSS.

[0714] html

[0715]

[0716]

[0717] <title> Recommended hiking trails< / title>

[0718]

[0719]

[0720] <h1> 10 Great Hiking Trails< / h1>

[0721] Text content...

[0722]

[0723]

[0724]

[0725] Input: Generated text and images

[0726] Output: The created landing page

[0727] Step 8: Deploy

[0728] The server deploys the completed landing page to a web server, which allows users to access the landing page through their browsers.

[0729] shell

[0730] scp landing_page.html user@webserver: / var / www / html /

[0731] Input: The created landing page

[0732] Output: Deployed landing page

[0733] Through these steps, the optimal landing page based on the user's emotions and preferences is efficiently provided.

[0734] (Application example 2)

[0735] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0736] Conventional landing page generation systems and product recommendation systems perform analysis based on user behavior data and purchase history, but there is a need to achieve more accurate individual optimization by incorporating the user's emotional state. Specifically, there is a problem in that it is difficult to provide optimal content based on the user's physical emotional state. This has resulted in many cases where these systems are ineffective in improving the user experience or stimulating purchasing motivation.

[0737] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0738] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, and search history, means for analyzing the data to identify the user's hobbies, preferences, and emotional state, means for feeding the data to a generative AI model based on the analysis results to generate text and images optimized for the user, and means for arranging the generated text and images to automatically create a landing page or a recommended product list, thereby enabling the provision of individually optimized content that also takes into account the user's emotional state.

[0739] "User behavior data" refers to a record of the actions and behavior a user performs online, including, specifically, information such as the web pages visited, the links clicked, and the duration of viewing.

[0740] "Purchase history" refers to a record of products purchased by a user in the past, and includes information such as product name, purchase date and time, quantity, total amount, etc.

[0741] "Search history" refers to a record of search queries a user has made on the Internet, including the keywords searched, the date and time of the search, and click information for search results.

[0742] "Means of collection" refers to the methods and technologies used to obtain user behavioral data, purchase history, search history, etc., and includes data acquisition using APIs and tracking using cookies.

[0743] "Means of analysis" refers to methods and techniques for internally processing collected data and extracting useful information, including techniques such as data cleaning, feature extraction, and clustering.

[0744] "Hobbies and preferences" refers to areas or categories in which a user has particular interests or concerns, including preferences for specific products and interest in specific activities.

[0745] An "emotional state" refers to the emotional state a user feels at a particular time or in a particular situation, and includes emotions such as joy, excitement, sadness, and anger.

[0746] "Generative AI model" refers to a model that uses artificial intelligence to generate new content from specific input data, and includes technologies used for text generation and image generation (e.g., GPT and DALL-E).

[0747] "Means of feeding" refers to the methods and technologies for supplying analysis results and other necessary data to the generative AI model, including data preprocessing and input format adjustment.

[0748] "Text and Images" means the part of the content provided to users, including written and visual information.

[0749] A "landing page" refers to a web page accessed when an advertisement or link is clicked, and includes pages that provide specific information about a particular product or service.

[0750] A "recommended product list" is a series of products recommended to users based on their interests and is optimized based on their past purchasing history, behavioral data, emotional state, etc.

[0751] "Means of automatic creation" refers to methods and technologies that allow a system to automatically generate and arrange content without requiring analog manual work, including template-based arrangement and automatic layout generation.

[0752] This invention is a system that collects and analyzes data such as user behavior data, purchase history, and search history on the Internet, and provides individually optimized content that also takes into account emotional state. Specific implementation methods for realizing this system are described below.

[0753] Collection Stage

[0754] The server uses APIs to collect multiple data such as user behavior, purchase history, and search history. This data includes detailed information such as which web pages users visited, which links they clicked, and which product reviews they read. The collected data is stored in a local database for further analysis.

[0755] Analysis stage

[0756] The server cleans the collected data and extracts features. This includes removing noise and standardizing the format. It then uses a clustering algorithm and a sentiment analysis engine to classify users into different clusters and identify their hobbies, preferences, and interests. At the same time, it analyzes the user's text messages and voice inputs to understand their emotional state. Once this analysis is complete, it can be determined that the user is currently in a specific emotional state, such as "joy" or "excitement."

[0757] Generation stage

[0758] Based on the analysis results, the server feeds data to a generative AI model. For example, the user may input information such as "I love the outdoors" and being in a "joy" emotional state. This generative AI uses large-scale language models and image generation models (e.g., GPT-4 and DALL-E). The generative AI model generates text and images optimized for the user. For example, it generates text content such as "Top 10 recommended hiking trails" and images depicting the beautiful scenery of those hiking trails.

[0759] Placement and display stage

[0760] The generated text and images are placed in the appropriate positions based on a pre-defined template. The server uses these artifacts to automatically create a landing page and a list of recommended products. The template uses HTML and CSS to maintain a consistent design. The final landing page is deployed to a web server, where it can be accessed by users through a browser.

[0761] Specific examples

[0762] User A frequently purchases outdoor equipment from a specific shopping site and also frequently reads articles about hiking. After collecting this user's data and conducting clustering and sentiment analysis, the user was classified into a cluster called "Outdoors Lover" and identified as being in a "Joy" emotional state. The server fed this information into a generative AI model, which generated images of beautiful scenery associated with the text "10 Recommended Hiking Trails." This information was then arranged as a landing page based on a template and deployed.

[0763] Prompt Sentence Examples

[0764] An example of a prompt sentence to be input to the generative AI model is, "Create the optimal landing page for a user who is interested in outdoor gear and is currently in the emotional state of 'joy'."

[0765] In this way, this invention takes into consideration the emotional state of the user and generates and provides individually optimized content, thereby significantly improving the user experience.

[0766] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0767] Step 1: Collect user data

[0768] The server uses an API to collect multiple data such as user behavior data, purchase history, and search history. Specific operations include obtaining web page log data, clickstreams, and detailed data on viewed products. The input is data related to the user's online operations, and the output is an organized set of user behavior data.

[0769] Step 2: Data cleaning and feature extraction

[0770] The server cleans the collected data and extracts features. Specifically, it removes noise and standardizes the data format. The input is raw user data, and the output is clean data and feature-extracted data.

[0771] Step 3: Clustering and sentiment analysis

[0772] The server performs clustering and sentiment analysis on the cleaned data. Using a clustering algorithm, users are classified into different clusters. It also uses a sentiment analysis engine to analyze users' text messages and voice data to identify their emotional state. The inputs are the cleaned data and feature-extracted data, and the output is cluster information and emotional state information.

[0773] Step 4: Feed data into the generative AI model

[0774] The server feeds cluster information and emotional state information to the generative AI model. For example, input information for a user who is an "outdoor lover" and has the emotion of "joy." The input is the cluster information and emotional state information, and the output is the input data for the generative AI model.

[0775] Step 5: Generate optimized content

[0776] The server uses a generative AI model to generate text and images optimized for the user. For example, it generates text content such as "Top 10 recommended hiking trails" and images of beautiful scenery. The input is the input data to the generative AI model, and the output is the generated text and images. For generation, a text generation model such as GPT-4 and an image generation model such as DALL-E are used.

[0777] Step 6: Arranging Content

[0778] The server places the generated text and images in the appropriate locations based on a pre-defined template. Specifically, it uses HTML and CSS to build a landing page or recommended product list. The input is the generated text and images, and the output is the HTML code for the completed web page.

[0779] Step 7: Deploy your landing page

[0780] The server deploys the completed landing page and recommended product list to a web server, where users can access these pages through their browsers. The input is the HTML code for the completed web page, and the output is a landing page accessible on the web.

[0781] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0782] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0783] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0784] [Third embodiment]

[0785] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0786] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0787] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0788] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0789] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0790] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0791] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0792] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0793] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0794] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0795] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0796] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0797] The system that embodies this invention mainly consists of multiple servers, terminals, and users. The server analyzes various user data, and based on the results, it uses a generation AI to automatically generate the landing page (LP) that is most suitable for the user. This system quickly and efficiently provides optimal content based on the user's interests and preferences.

[0798] Program processing

[0799] User data collection

[0800] The server collects data such as user behavior, purchase history, and search history. Specifically, information such as the pages the user visited on the Internet, the products they purchased, and the search queries they performed is obtained via API. This allows the server to store detailed user preference information in a local database.

[0801] Data analysis

[0802] The server analyzes the collected data. This analysis includes preprocessing steps such as data cleansing and feature extraction. For example, categories of products frequently purchased by a user are extracted from their purchase history, and this is then passed through a clustering algorithm to classify users into multiple clusters. This makes it clear what interests the user has.

[0803] Preparing for generative AI

[0804] The server prepares the analysis results to be fed into a generative AI model. For example, a natural language processing model is used for text generation, and an image generation model is used for image generation. The generative AI model receives input such as user features and cluster information.

[0805] Content Generation

[0806] The server uses a generative AI model to generate text and images optimized for each user. For example, to generate a travel guide article for "User A," the server generates text including a title such as "Top 10 Adventure Travel Destinations" and detailed information about the adventure trip. In addition, related beautiful landscape images are also generated.

[0807] Creating a Landing Page

[0808] The server creates a landing page based on the generated content. Specifically, it inserts the generated text and images into a pre-defined template, builds the page structure using HTML and CSS, and finally deploys the page on a web server so that users can access it through their browsers.

[0809] Specific examples

[0810] User data collection

[0811] User B frequently purchases outdoor equipment from a shopping site and frequently reads articles about hiking using a certain device. The server collects this data via an API and stores it in a local database.

[0812] Data analysis

[0813] The server analyzes User B's data and classifies him / her into a cluster called "outdoor enthusiasts" based on the categories of outdoor equipment he / she has purchased and the content of articles he / she has viewed. It then identifies the areas and activities in which User B is particularly interested.

[0814] Preparing for generative AI

[0815] The server feeds the analysis results into the generative AI model and prepares to generate optimal content for User B. Specifically, it inputs data indicating that User B is interested in hiking into the generative AI.

[0816] Content Generation

[0817] The server uses generative AI to generate text content such as "Top 10 hiking trails recommended for User B." It also generates beautiful landscape images related to each hiking trail.

[0818] Creating a Landing Page

[0819] The server places the generated text and images into a template to create a landing page for the hiking guide, and finally deploys this page to a web server, making it accessible to User B.

[0820] In this way, the system can efficiently provide optimized information to User B.

[0821] The processing flow will be explained below.

[0822] Step 1:

[0823] The server collects data such as user behavior, purchase history, and search history. For example, information about the pages a user visits on a shopping site, the products they purchase, and the queries they search for is obtained via API and stored in a local database. This allows detailed information about the user's preferences to be gathered.

[0824] Step 2:

[0825] The server performs preprocessing, cleaning the data. Specifically, it fills in missing values ​​and removes outliers. It also standardizes the data format and standardizes each data point. For example, it converts all purchase history prices into the same currency unit.

[0826] Step 3:

[0827] The server analyzes the preprocessed data and extracts features to identify the user's interests and preferences. For example, features could include the product categories the user frequently purchases or the themes of the articles they view. Text analysis and category classification techniques are used to extract these features.

[0828] Step 4:

[0829] The server uses a clustering algorithm to group users, for example, using the K-means algorithm to classify users with similar interests into the same cluster. This clustering helps clarify the areas of interest and concern of users.

[0830] Step 5:

[0831] The server then feeds the data to a generative AI model based on the analysis results. For example, if a user belongs to a cluster called "Outdoors Lovers," that information is input to the generative AI. These generative AI models include text generation models (e.g., GPT-4) and image generation models (e.g., DALL-E).

[0832] Step 6:

[0833] The server uses the generative AI model to generate text and images optimized for each user. For example, for "User B," it generates text content such as "The 10 Best Hiking Trails" and related beautiful scenery images.

[0834] Step 7:

[0835] The server places the generated text and images into a landing page template, building the page layout using HTML and CSS and inserting the generated content in the appropriate places, resulting in a landing page with a consistent design.

[0836] Step 8:

[0837] The server deploys the completed landing page to the web server, allowing users to access the landing page through their browsers. User B then views the "10 Best Hiking Trails" page.

[0838] In this way, the system collects data, analyzes it, generates AI, places content, and deploys it step by step, quickly and efficiently providing users with optimized landing pages.

[0839] Example 1

[0840] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0841] With current technology, efficiently generating landing pages tailored to user preferences requires a lot of manual work, making it difficult to quickly provide optimal content.In addition, there are limited methods for effectively utilizing user behavior data, purchase history, search history, etc., making it difficult to generate personalized content for users.

[0842] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0843] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, and search history, means for analyzing the data and identifying the user's interests and preferences, means for generating prompt sentences based on the analysis results and feeding the data to a generative AI model, means for generating text and images optimized for the user using the generative AI model, and means for arranging the generated text and images to automatically create a landing page. This makes it possible to quickly and automatically generate an optimal landing page based on the user's preferences and efficiently provide content.

[0844] "User" means an end user of the Internet service or website.

[0845] "Behavioral data" refers to data such as page browsing history and click history when a user visits a website.

[0846] "Purchase history" refers to historical data about products a user has previously purchased from online shops, etc.

[0847] "Search history" refers to historical data about queries and search results a user has performed on the Internet.

[0848] "Data analysis" refers to a series of processes that aggregate, classify, and evaluate collected data.

[0849] "Hobbies and preferences" refers to data that indicates a user's interests, concerns, and preferences.

[0850] A "prompt" is an instruction given to a generative AI model.

[0851] A "generative AI model" is an artificial intelligence model that generates text or images from specified input data.

[0852] "Text" refers to sentences or strings of characters generated by a generative AI model.

[0853] "Image" refers to the visual data generated by a generative AI model.

[0854] A "landing page" is a web page created for a user to access in a web browser.

[0855] A "template" is a predefined layout or design framework used to automatically create a landing page.

[0856] "Deployment" is the process of placing a completed landing page on a web server so that it can be accessed by users.

[0857] A system embodying this invention efficiently collects and analyzes multiple data sets, such as user behavior data, purchase history, and search history, and automatically generates content optimized for each user using a generative AI model. Specific embodiments of this system are described below.

[0858] First, the server collects information such as user behavior data, purchase history, and search history. This data is obtained from the user's browser history, purchase history, search queries, etc. recorded by the device. The server obtains this data via API and stores it in a local database. For example, if a user purchases outdoor equipment on a shopping site or reads an article about hiking, that information will be collected.

[0859] The server then analyzes the collected data. At this stage, data cleansing is performed to remove noise and missing data. Feature extraction is also performed to identify users' interests based on their purchase history and browser history. For example, a clustering algorithm is used to separate the data into multiple clusters, categorizing users into clusters such as "outdoor enthusiasts" or "travel enthusiasts."

[0860] Based on the analysis results, the server generates a prompt sentence and prepares to feed it to the generative AI model. For example, if the analysis finds that the user has specific interests, it generates a prompt sentence such as, "Please tell me some recommended hiking trails for User A." The server then creates an instance of a generative AI model (e.g., GPT-3 or StyleGAN) and sets the prompt sentence as input.

[0861] The server inputs prompts into the generative AI model to generate text and images optimized for the user. For example, the generative AI model generates text content such as "Top 10 recommended hiking trails for user A" and also creates related beautiful landscape images. This automatically generates content that is individually optimized for each user.

[0862] Finally, the server creates a landing page based on the generated text and images. At this stage, the generated content is inserted according to a pre-defined template, and the page structure is built using HTML and CSS. The completed landing page is deployed to a web server, where users can access it through their browsers.

[0863] To illustrate how this system works, let's use a practical example. If User B is buying outdoor gear on a shopping site and reading an article about hiking, the server will collect this behavioral data and identify the user's interest as "outdoors enthusiast." As a result, a prompt message will be created that generates "10 recommended hiking trails for User B," and specific content will be generated by the generative AI model. This generated content will be placed on a landing page based on a template and deployed for User B's easy access.

[0864] An example of a prompt sentence is, "Please generate content that introduces hiking trails that User B is interested in. User B has recently purchased a lot of outdoor equipment and particularly enjoys hiking."

[0865] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0866] Step 1: Data collection

[0867] The server collects information such as user behavior data, purchase history, and search history. Specifically, the device sends the user's browser history, purchase records, and search queries to the server via API. The server stores this data in a local database. For example, if a user purchases outdoor equipment on a shopping site or reads an article about hiking, that information will be collected.

[0868] Input: User behavior data, purchase history, search history

[0869] Output: User data stored in a local database

[0870] Step 2: Data analysis

[0871] The server retrieves collected user data from a local database and first cleanses the data, removing missing data and noise. Next, it performs feature extraction to identify users' interests based on their purchase history and browser history. It then uses a clustering algorithm to categorize users into clusters such as "outdoor enthusiasts" or "travel enthusiasts."

[0872] Input: User data from a local database

[0873] Output: User features and cluster information

[0874] Step 3: Prompt generation

[0875] The server generates a prompt based on the results of the data analysis. For example, if the server finds that the user has a specific interest, it creates a prompt such as, "Please recommend some hiking trails for User A." This prepares the instruction sentence to be given to the generative AI model.

[0876] Input: Data analysis results (user features and cluster information)

[0877] Output: Generated prompt statement

[0878] Step 4: Input to the generative AI model

[0879] The server creates an instance of a generative AI model (e.g., GPT-3 or StyleGAN) and sets the prompt sentence generated in the previous step as input, so the model is ready to generate optimal content.

[0880] Input: Generated prompt text

[0881] Output: Input data to a generative AI model

[0882] Step 5: Content Generation

[0883] The server inputs the prompt into the generative AI model, which generates text and images optimized for the user. The generative AI model then generates specific content based on the prompt. For example, it generates text content such as "Top 10 recommended hiking trails for user A" and related beautiful landscape images.

[0884] Input: Prompt sentence for generative AI model

[0885] Output: Generated text and images

[0886] Step 6: Create a Landing Page

[0887] The server creates a landing page by arranging the generated text and images according to a pre-defined template, building the page structure using HTML and CSS, and inserting the generated content. The completed landing page is then deployed to a web server, where users can access it through their browsers.

[0888] Input: Generated text and images

[0889] Output: The generated landing page

[0890] This process flow allows the system to quickly and efficiently generate optimal landing pages based on user preferences.

[0891] (Application example 1)

[0892] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0893] Conventional internet advertising and landing page creation systems have not adequately provided optimal content based on users' hobbies and preferences, and have had difficulty providing content in real time. As a result, advertisements and content that do not match the user's interests are displayed, making effective marketing impossible. Furthermore, even when displaying advertisements using wearable devices such as smart glasses, there is a lack of a means to provide users with content optimized for them in real time. There is a need to solve these issues.

[0894] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0895] In this invention, the server includes a means for collecting multiple data such as user behavior data, purchase history, and search history, a means for analyzing the data to identify the user's interests and preferences, and a means for feeding the data to a generative AI model based on the analysis results to generate text and images optimized for the user, thereby arranging the generated text and images to automatically create landing pages or advertising content and display them on the smart glasses.

[0896] "User behavior data" refers to information such as the pages a user visits on the Internet, the links they click, and the time they spend browsing.

[0897] "Purchase history" refers to information such as the products and services a user has purchased in the past, the date and time of those purchases, and frequency of those purchases.

[0898] "Search history" refers to the search queries a user has made on a search engine and the history of the related search result pages they have visited.

[0899] A "generative AI model" is an artificial intelligence model trained on a large dataset, and refers to a technique for generating text or images based on specific input data.

[0900] A "clustering algorithm" refers to a statistical method for analyzing large amounts of data and grouping data that are highly similar.

[0901] A "landing page" is a web page that appears first when a user clicks on an advertisement or link, and contains content designed to encourage a specific action (such as purchasing a product or entering information).

[0902] "Smart glasses" are a type of wearable device that has the shape of glasses but has the ability to display visual information as augmented reality (AR).

[0903] "Template" means a document or web page structure with a specific format or style predefined for efficient layout and display of content.

[0904] The system for implementing this invention consists of a server, a terminal, and a user. The server collects and analyzes information such as user behavior data, purchase history, and search history, and uses AI to display optimal landing pages and advertising content on the smart glasses based on the results.

[0905] Program Overview

[0906] The system includes the following procedures:

[0907] 1. User data collection:

[0908] The server collects data such as user behavior, purchase history, and search history via API, and stores detailed user preference information in a local database.

[0909] 2. Data Analysis:

[0910] The server cleanses the collected data, extracts and analyzes features, and uses a clustering algorithm to classify users into different clusters based on their purchase history and the pages they have viewed.

[0911] 3. Prepare the generative AI:

[0912] The server creates a prompt sentence to feed the analysis results to a generative AI model (e.g., GPT-3), which includes the user's cluster information and interests.

[0913] 4. Advertising content generation:

[0914] A generative AI model generates optimized text and images based on the prompt, and the generated ad content is tailored to what the user should see.

[0915] 5. Display on smart glasses:

[0916] The server displays the generated content on the smart glasses in real time, allowing users to visually view the most suitable advertising content through the glasses they are wearing.

[0917] Specific examples

[0918] Let's say a user is walking through a particular shopping mall. The server determines from past behavioral data and purchase history that the user is interested in outdoor gear. Based on this analysis, it inputs the following prompt into the generative AI model:

[0919] plaintext

[0920] "Generate ads that are best suited to users in Cluster 1, whose primary interest is outdoor gear. Include current sales and popular items in the ad content."

[0921] Based on this prompt, the generative AI model generates advertising text, such as "The latest outdoor gear is on sale!", along with images of the products on sale. The server then displays this generated advertising content on the smart glasses in real time, providing users with the most appropriate information.

[0922] Hardware and software used

[0923] Hardware: Smart glasses (e.g. HoloLens, Google Glass)

[0924] software:

[0925] Data collection via API: RESTful API

[0926] Data analysis: NumPy, scikit-learn

[0927] Generative AI model: GPT-3 (OpenAI)

[0928] This makes it possible to provide advertising content that matches the user's preferences in real time.

[0929] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0930] Step 1:

[0931] The server collects user behavioral data, purchase history, and search history via API. Specifically, the server accesses each API endpoint and retrieves data. For example, behavioral data including the user's visited pages and browsing time, history of past purchases, and search engine query history are collected. This data is stored in a local database.

[0932] Input: API endpoint

[0933] Output: User data stored in a local database

[0934] Step 2:

[0935] The server cleanses the collected data and extracts features. First, it removes duplicate data and missing values ​​to clean the data. Next, it extracts the categories of products that users frequently purchase and the features of their search queries. For example, if a user frequently purchases outdoor equipment, the features of that category are extracted.

[0936] Input: Raw data in a local database

[0937] Output: Cleansed feature data

[0938] Step 3:

[0939] The server runs a clustering algorithm on the cleansed data to classify users into different clusters. For example, it uses KMeans clustering to classify users into clusters such as "outdoor enthusiasts" and "tech gadget enthusiasts." This identifies the user's interests and preferences.

[0940] Input: Cleansed feature data

[0941] Output: User data categorized into clusters

[0942] Step 4:

[0943] The server then creates prompts to be input into the generative AI model based on the clustered user data. For example, for a user who loves the outdoors, the server might generate a prompt like, "Generate ads that are optimal for users in cluster 1. Their primary interest is outdoor gear. Include current sales information and popular products in the ad content."

[0944] Input: User data categorized into clusters

[0945] Output: A prompt to be input to the generative AI model

[0946] Step 5:

[0947] The server inputs a prompt into a generative AI model (e.g., GPT-3) to generate text and images. Based on the prompt, the generative AI model generates optimal ad text and related images. For example, ad text such as "Latest outdoor gear on sale!" and images of outdoor gear on sale are generated.

[0948] Input: prompt statement

[0949] Output: Generated ad text and images

[0950] Step 6:

[0951] The server transmits the generated advertising text and images to the smart glasses for display in real time, and the smart glasses visually display the received advertising content, allowing the user to easily check the information.

[0952] Input: Generated ad text and image

[0953] Output: Advertising content displayed on smart glasses

[0954] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0955] The system for implementing this invention comprises multiple servers, terminals, and users. This system includes an emotion engine that analyzes various user data and uses generative AI to automatically generate the optimal landing page (LP) for the user. This emotion engine analyzes the user's text messages, voice input, images, and videos, identifies the user's emotions, and generates optimized content. This allows for efficient provision of content that takes user emotions into consideration.

[0956] Program processing

[0957] User data collection

[0958] The server collects multiple data sets, including user behavior data, purchase history, and search history. Specifically, information about the pages the user visited, the products they purchased, and the queries they searched is obtained via API and stored in a local database. It also uses an emotion engine to collect the emotions expressed by the user during interactions.

[0959] Data analysis

[0960] The server analyzes the collected data. This analysis includes data cleaning and feature extraction. For example, it extracts the categories of products frequently purchased by users from their purchase history and classifies users into different clusters using a clustering algorithm. Furthermore, it uses an emotion engine to analyze emotions from users' text messages and voice inputs to identify what content is more appropriate.

[0961] Preparing for generative AI

[0962] The server feeds data to a generative AI model based on the analysis results. For example, if a user has a cluster of "loves the outdoors" and an emotion of "joy," that information is input into the generative AI. These generative AI models include text generation models (e.g., GPT-4) and image generation models (e.g., DALL-E).

[0963] Content Generation

[0964] The server uses a generative AI model to generate text and images optimized for each user. For example, for "User B," it generates text content such as "The 10 Best Hiking Trails," along with related images of beautiful scenery that evoke a sense of enjoyment. Furthermore, if the user expresses the emotion of "joy," it selects positive words and images that further enhance this emotion.

[0965] Creating a Landing Page

[0966] The server places the generated text and images into a landing page template, building the page layout using HTML and CSS and inserting the generated content in the appropriate places, resulting in a landing page with a consistent design.

[0967] Deploy

[0968] The server deploys the completed landing page to the web server, which allows users to access the landing page through their browsers. User B views the "10 Best Hiking Trails" page, which is designed to provide emotional satisfaction.

[0969] Specific examples

[0970] User data collection

[0971] User B frequently purchases outdoor equipment on a shopping site and reads articles about hiking using a certain device. User B expresses emotions such as "joy" and "excitement" while interacting with the site. The server collects this information via an API and stores it in a local database.

[0972] Data analysis

[0973] The server analyzes User B's data and identifies a cluster called "outdoor enthusiasts" based on the categories of outdoor equipment purchased and the content of articles viewed. It also analyzes User B's emotional data using an emotion engine to identify "joy" and "excitement."

[0974] Preparing for generative AI

[0975] The server feeds the analysis results into a generative AI model. For example, it inputs the cluster "loves the outdoors" and the emotion information "joy." The generative AI generates optimal text and images based on a large amount of data.

[0976] Content Generation

[0977] The server uses the generative AI model to generate text content such as "10 recommended hiking trails for User B." It also generates beautiful scenic images of hiking trails that User B is interested in. It uses positive words and colors that reinforce User B's emotion of "joy."

[0978] Creating a Landing Page

[0979] The server places the generated text and images into a template, inserting the generated content into a travel guide template and building the page layout.

[0980] Deploy

[0981] The server deploys the completed landing page to the web server, and User B accesses this page through a browser, which also provides emotional satisfaction.

[0982] In this way, the system collects data, analyzes it, analyzes emotions, generates AI, places content, and deploys it at each step, enabling it to quickly and efficiently provide optimal landing pages that also take user emotions into consideration.

[0983] The processing flow will be explained below.

[0984] Step 1:

[0985] The server collects multiple data such as user behavior data, purchase history, search history, etc. For example, information about the pages a user visits on a shopping site, the products they purchase, and the queries they search for is obtained via API and stored in a local database. Additionally, the emotion engine analyzes users' text messages and voice inputs to collect emotion data.

[0986] Step 2:

[0987] The server preprocesses the collected data by cleaning it (imputing missing values, removing outliers), standardizing it, and encoding categorical data (for example, standardizing currency units for purchase history and standardizing search queries).

[0988] Step 3:

[0989] The server analyzes the preprocessed data to identify users' interests and preferences. Specifically, it extracts the categories of products frequently purchased by users and the themes of articles they read, and then uses a clustering algorithm to classify users into different clusters. It also takes into account the user's emotional data analyzed by the emotion engine.

[0990] Step 4:

[0991] The server prepares the analysis results to feed into the generative AI model. For example, if the user has the emotion of "joy" in addition to the cluster of "loves the outdoors," that information is input into the generative AI model.

[0992] Step 5:

[0993] The server generates text content and images using a generative AI model. For example, it creates the text "10 Best Hiking Trails" for User B and generates related beautiful landscape images. It also selects positive phrases and vividly colored images to reinforce the user's emotion of "joy."

[0994] Step 6:

[0995] The server places the generated text and images into a landing page template, constructs the page layout using HTML and CSS, and inserts the generated content in the appropriate places, such as placing the title "10 Best Hiking Trails" in the page header, followed by the following text and images, one paragraph at a time.

[0996] Step 7:

[0997] The server deploys the completed landing page to a web server, allowing users to access the landing page through their browsers. For example, User B can enter the URL in his internet browser and view the generated hiking trail page.

[0998] Specific examples

[0999] User data collection

[1000] User B frequently browses articles about buying outdoor gear and hiking using his device. The server collects this information via API and stores his purchase history, visited pages, and search queries in a local database. The emotion engine analyzes text and voice data expressing User B's emotions such as "joy" and "excitement" and stores this emotion data as well.

[1001] Data analysis

[1002] The server cleans and standardizes User B's data before analyzing it. Based on the categories of outdoor equipment purchased and the content of articles viewed, User B is classified into a cluster called "Outdoor Lovers." User B's emotional data, "Joy," is also taken into account.

[1003] Preparing for generative AI

[1004] The server prepares to feed the data to the generative AI model based on the analysis results. The cluster of "outdoor lover" and the emotion information of "joy" are input into the generative AI model.

[1005] Content Generation

[1006] The server uses a generative AI model to generate images related to the text "10 recommended hiking trails for User B." It also selects positive phrases and brightly colored images to amplify User B's emotion of "joy."

[1007] Creating a Landing Page

[1008] The server places the generated text and images into a travel guide template, using HTML and CSS to place "10 Best Hiking Trails" in the page header and insert the generated content for each paragraph.

[1009] Deploy

[1010] The server deploys the completed landing page to the web server. User B can then access the page in a browser, view the generated content, and experience emotional satisfaction. This system makes it possible to provide information optimized for User B's interests and emotions.

[1011] Example 2

[1012] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1013] Current web content generation systems have difficulty efficiently providing personalized content that fully considers each user's emotions and behavioral history. In particular, there is a need for a method that can collect and analyze a variety of user data in real time, automatically generate content that reflects the user's emotions, and provide it as an optimal landing page.

[1014] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1015] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, search history, text messages, and voice data, means for cleaning the collected data, extracting features, and analyzing them, means for analyzing user sentiment from the analyzed data, means for feeding data to a generative AI model based on the analysis results and generating text and images optimized for the user, and means for arranging the generated text and images and automatically creating a landing page, thereby making it possible to efficiently provide an optimal landing page based on the user's sentiments and preferences.

[1016] "Behavioral data" is information about specific actions a user takes on a website or application, such as clicks, views, or purchases.

[1017] "Purchase History" refers to the products and services a User has previously purchased and related details.

[1018] "Search history" refers to the history of queries a user has made on search engines and websites, and is data that indicates the user's interests and concerns.

[1019] "Text Message" means any written communication sent or received by a User on a Digital Platform.

[1020] "Voice Data" refers to voice information input by a user or recorded by a system.

[1021] "Means of collection" refers to the methods and technologies used to acquire and store user behavioral data, purchase history, search history, text messages, voice data, etc. in a system.

[1022] "Cleaning" refers to the process of removing unnecessary information from collected data or filling in gaps in the data.

[1023] "Means for extracting features" refers to techniques and methods for extracting important patterns and information from collected data and using them for analysis.

[1024] "Analysis methods" refers to the algorithms and technologies used to process collected data and identify user preferences and behavioral patterns.

[1025] "Means for analyzing emotions" refers to technology for reading and classifying emotions from users' text messages and voice data.

[1026] A "generative AI model" refers to artificial intelligence technology that generates text and images based on large amounts of data.

[1027] "Landing page" means the web page a user first arrives at after clicking on an internet advertisement, email link, etc.

[1028] "Positioning means" refers to a technique for appropriately positioning the generated text and images on the landing page.

[1029] The system embodying this invention collects and analyzes user behavior data, purchase history, search history, text messages, voice data, etc. to identify the user's hobbies, preferences, and emotions, and automatically generates an optimal landing page based on these. It is particularly characterized by the use of an emotion engine for analyzing user emotions and a generative AI model.

[1030] The server collects data about the various actions users take on websites and applications. Specifically, it periodically retrieves behavioral data, purchase history, and search history via APIs and stores them in a local database. It also collects text messages and voice data sent and received by users within the site in real time.

[1031] The collected data is cleaned by the server, with unnecessary information removed and missing values ​​filled in. Next, features are extracted from the data, such as frequently purchased product categories determined from a user's purchase history. Furthermore, a clustering algorithm is used to classify users into different clusters. This process allows the user's behavioral patterns and preferences to be identified.

[1032] The server then performs sentiment analysis using the collected text messages and voice data. For sentiment analysis, a sentiment analysis model using natural language processing technology is used to identify emotions from the user's text messages and voice data and assign emotion labels.

[1033] Based on the analysis results, the server feeds the data to the generative AI model. This generative AI model includes a text generation model and an image generation model. For example, if a user "likes the outdoors" and has "emotions of joy," that information is input as a prompt to the generative AI model. An example of a prompt is shown below:

[1034] "Users love outdoor activities and feel joy. Can you recommend some hiking trails?"

[1035] The server receives the text and images returned by the generative AI model and uses them to generate content optimized for the user. For example, it generates text content such as "10 Best Hiking Trails" and related images of beautiful scenery. This generated content is then placed on a landing page based on a template pre-configured by the server. Specifically, the page layout is constructed using HTML and CSS, and the generated text and images are inserted in the appropriate locations.

[1036] Finally, the server deploys the completed landing page to the web server, allowing users to access it through their browsers, allowing users to view the optimal landing page based on their preferences and emotions and achieve emotional satisfaction.

[1037] This system is able to efficiently provide optimal landing pages that also take user emotions into consideration through a series of steps, from data collection and analysis, sentiment analysis, AI generation, content placement, and deployment.

[1038] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1039] Step 1: Collect user data

[1040] The server collects data on user behavior on the website, purchase history, search history, text messages, voice data, etc. Specifically, it may use an API to send requests such as the following to obtain data:

[1041] GET / api / user_behavior?user_id={user ID}&event_type=purchase

[1042] This allows details of the products a user purchases, search history, text messages, voice data, etc. to be collected and stored in a local database.

[1043] Input: User behavior data, purchase history, search history, text messages, voice data

[1044] Output: Collected data stored in a local database

[1045] Step 2: Data cleaning and feature extraction

[1046] The server cleans the collected data, removing unnecessary information and formatting it. Specifically, it removes duplicate data and fills in missing values. Next, it extracts features from purchase and search histories. For example, it identifies the categories of products frequently purchased by users.

[1047] python

[1048] import pandas as pd

[1049] data = pd.read_csv('user_data.csv')

[1050] data.drop_duplicates(inplace=True)

[1051] data.fillna(method='ffill', inplace=True)

[1052] popular_categories = data['category'].value_counts().head(5)

[1053] Input: Collected user data

[1054] Output: Cleaned data and extracted features

[1055] Step 3: Cluster the data

[1056] The server classifies users into different clusters based on their features using a clustering algorithm (e.g., K-means), which groups users with similar behavioral patterns.

[1057] python

[1058] from sklearn.cluster import KMeans

[1059] kmeans = KMeans(n_clusters=5)

[1060] data['cluster'] = kmeans.fit_predict(data[['feature1', 'feature2']])

[1061] Input: Cleaned data and extracted features

[1062] Output: User data sorted into clusters

[1063] Step 4: Sentiment Analysis

[1064] The server performs sentiment analysis using text messages and voice data. For sentiment analysis, a sentiment analysis model using natural language processing technology is used. For example, sentiment labels are assigned from text in the following format:

[1065] python

[1066] from transformers import pipeline

[1067] sentiment_analysis = pipeline('sentiment-analysis')

[1068] results = sentiment_analysis("This hike was amazing!")

[1069] Input: Text messages and voice data

[1070] Output: Data with emotion labels

[1071] Step 5: Prepare the generative AI model and generate prompts

[1072] The server feeds data to the generative AI model based on the analysis results, and prepares prompt sentences to input to the generative AI model.

[1073] python

[1074] prompt = "User enjoys outdoor activities and has the emotion joy. What hiking trails do you recommend?"

[1075] ai_model_input = {

[1076] "prompt": prompt,

[1077] "user_cluster": user_cluster,

[1078] "emotion_label": emotion_label

[1079] }

[1080] Input: Cluster labels and emotion labels

[1081] Output: Prompt and data to be fed into the generative AI model

[1082] Step 6: Content Generation

[1083] The server generates text and images using generative AI models, for example, it gets the text content "10 Best Hiking Trails" from a text generation model and related scenic images from an image generation model.

[1084] python

[1085] response = ai_model.generate(prompt)

[1086] Image generation

[1087] image_prompt = "Beautiful hiking trail views"

[1088] generated_image = image_model.generate(image_prompt)

[1089] Input: Prompt statement (text and image)

[1090] Output: Generated text and images

[1091] Step 7: Create a Landing Page

[1092] The server places the generated text and images into a landing page based on a pre-defined template, and the page layout is built using HTML and CSS.

[1093] html

[1094]

[1095]

[1096] <title> Recommended hiking trails< / title>

[1097]

[1098]

[1099] <h1> 10 Great Hiking Trails< / h1>

[1100] Text content...

[1101]

[1102]

[1103]

[1104] Input: Generated text and images

[1105] Output: The created landing page

[1106] Step 8: Deploy

[1107] The server deploys the completed landing page to a web server, which allows users to access the landing page through their browsers.

[1108] shell

[1109] scp landing_page.html user@webserver: / var / www / html /

[1110] Input: The created landing page

[1111] Output: Deployed landing page

[1112] Through these steps, the optimal landing page based on the user's emotions and preferences is efficiently provided.

[1113] (Application example 2)

[1114] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1115] Conventional landing page generation systems and product recommendation systems perform analysis based on user behavior data and purchase history, but there is a need to achieve more accurate individual optimization by incorporating the user's emotional state. Specifically, there is a problem in that it is difficult to provide optimal content based on the user's physical emotional state. This has resulted in many cases where these systems are ineffective in improving the user experience or stimulating purchasing motivation.

[1116] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1117] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, and search history, means for analyzing the data to identify the user's hobbies, preferences, and emotional state, means for feeding the data to a generative AI model based on the analysis results to generate text and images optimized for the user, and means for arranging the generated text and images to automatically create a landing page or a recommended product list, thereby enabling the provision of individually optimized content that also takes into account the user's emotional state.

[1118] "User behavior data" refers to a record of the actions and behavior a user performs online, including, specifically, information such as the web pages visited, the links clicked, and the duration of viewing.

[1119] "Purchase history" refers to a record of products purchased by a user in the past, and includes information such as product name, purchase date and time, quantity, total amount, etc.

[1120] "Search history" refers to a record of search queries a user has made on the Internet, including the keywords searched, the date and time of the search, and click information for search results.

[1121] "Means of collection" refers to the methods and technologies used to obtain user behavioral data, purchase history, search history, etc., and includes data acquisition using APIs and tracking using cookies.

[1122] "Means of analysis" refers to methods and techniques for internally processing collected data and extracting useful information, including techniques such as data cleaning, feature extraction, and clustering.

[1123] "Hobbies and preferences" refers to areas or categories in which a user has particular interests or concerns, including preferences for specific products and interest in specific activities.

[1124] An "emotional state" refers to the emotional state a user feels at a particular time or in a particular situation, and includes emotions such as joy, excitement, sadness, and anger.

[1125] "Generative AI model" refers to a model that uses artificial intelligence to generate new content from specific input data, and includes technologies used for text generation and image generation (e.g., GPT and DALL-E).

[1126] "Means of feeding" refers to the methods and technologies for supplying analysis results and other necessary data to the generative AI model, including data preprocessing and input format adjustment.

[1127] "Text and Images" means the part of the content provided to users, including written and visual information.

[1128] A "landing page" refers to a web page accessed when an advertisement or link is clicked, and includes pages that provide specific information about a particular product or service.

[1129] A "recommended product list" is a series of products recommended to users based on their interests and is optimized based on their past purchasing history, behavioral data, emotional state, etc.

[1130] "Means of automatic creation" refers to methods and technologies that allow a system to automatically generate and arrange content without requiring analog manual work, including template-based arrangement and automatic layout generation.

[1131] This invention is a system that collects and analyzes data such as user behavior data, purchase history, and search history on the Internet, and provides individually optimized content that also takes into account emotional state. Specific implementation methods for realizing this system are described below.

[1132] Collection Stage

[1133] The server uses APIs to collect multiple data such as user behavior, purchase history, and search history. This data includes detailed information such as which web pages users visited, which links they clicked, and which product reviews they read. The collected data is stored in a local database for further analysis.

[1134] Analysis stage

[1135] The server cleans the collected data and extracts features. This includes removing noise and standardizing the format. It then uses a clustering algorithm and a sentiment analysis engine to classify users into different clusters and identify their hobbies, preferences, and interests. At the same time, it analyzes the user's text messages and voice inputs to understand their emotional state. Once this analysis is complete, it can be determined that the user is currently in a specific emotional state, such as "joy" or "excitement."

[1136] Generation stage

[1137] Based on the analysis results, the server feeds data to a generative AI model. For example, the user may input information such as "I love the outdoors" and being in a "joy" emotional state. This generative AI uses large-scale language models and image generation models (e.g., GPT-4 and DALL-E). The generative AI model generates text and images optimized for the user. For example, it generates text content such as "Top 10 recommended hiking trails" and images depicting the beautiful scenery of those hiking trails.

[1138] Placement and display stage

[1139] The generated text and images are placed in the appropriate positions based on a pre-defined template. The server uses these artifacts to automatically create a landing page and a list of recommended products. The template uses HTML and CSS to maintain a consistent design. The final landing page is deployed to a web server, where it can be accessed by users through a browser.

[1140] Specific examples

[1141] User A frequently purchases outdoor equipment from a specific shopping site and also frequently reads articles about hiking. After collecting this user's data and conducting clustering and sentiment analysis, the user was classified into a cluster called "Outdoors Lover" and identified as being in a "Joy" emotional state. The server fed this information into a generative AI model, which generated images of beautiful scenery associated with the text "10 Recommended Hiking Trails." This information was then arranged as a landing page based on a template and deployed.

[1142] Prompt Sentence Examples

[1143] An example of a prompt sentence to be input to the generative AI model is, "Create the optimal landing page for a user who is interested in outdoor gear and is currently in the emotional state of 'joy'."

[1144] In this way, this invention takes into consideration the emotional state of the user and generates and provides individually optimized content, thereby significantly improving the user experience.

[1145] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1146] Step 1: Collect user data

[1147] The server uses an API to collect multiple data such as user behavior data, purchase history, and search history. Specific operations include obtaining web page log data, clickstreams, and detailed data on viewed products. The input is data related to the user's online operations, and the output is an organized set of user behavior data.

[1148] Step 2: Data cleaning and feature extraction

[1149] The server cleans the collected data and extracts features. Specifically, it removes noise and standardizes the data format. The input is raw user data, and the output is clean data and feature-extracted data.

[1150] Step 3: Clustering and sentiment analysis

[1151] The server performs clustering and sentiment analysis on the cleaned data. Using a clustering algorithm, users are classified into different clusters. It also uses a sentiment analysis engine to analyze users' text messages and voice data to identify their emotional state. The inputs are the cleaned data and feature-extracted data, and the output is cluster information and emotional state information.

[1152] Step 4: Feed data into the generative AI model

[1153] The server feeds cluster information and emotional state information to the generative AI model. For example, input information for a user who is an "outdoor lover" and has the emotion of "joy." The input is the cluster information and emotional state information, and the output is the input data for the generative AI model.

[1154] Step 5: Generate optimized content

[1155] The server uses a generative AI model to generate text and images optimized for the user. For example, it generates text content such as "Top 10 recommended hiking trails" and images of beautiful scenery. The input is the input data to the generative AI model, and the output is the generated text and images. For generation, a text generation model such as GPT-4 and an image generation model such as DALL-E are used.

[1156] Step 6: Arranging Content

[1157] The server places the generated text and images in the appropriate locations based on a pre-defined template. Specifically, it uses HTML and CSS to build a landing page or recommended product list. The input is the generated text and images, and the output is the HTML code for the completed web page.

[1158] Step 7: Deploy your landing page

[1159] The server deploys the completed landing page and recommended product list to a web server, where users can access these pages through their browsers. The input is the HTML code for the completed web page, and the output is a landing page accessible on the web.

[1160] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1161] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1162] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1163] [Fourth embodiment]

[1164] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1165] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1166] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1167] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1168] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1169] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1170] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1171] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1172] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1173] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1174] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1175] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1176] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1177] The system that embodies this invention mainly consists of multiple servers, terminals, and users. The server analyzes various user data, and based on the results, it uses a generation AI to automatically generate the landing page (LP) that is most suitable for the user. This system quickly and efficiently provides optimal content based on the user's interests and preferences.

[1178] Program processing

[1179] User data collection

[1180] The server collects data such as user behavior, purchase history, and search history. Specifically, information such as the pages the user visited on the Internet, the products they purchased, and the search queries they performed is obtained via API. This allows the server to store detailed user preference information in a local database.

[1181] Data analysis

[1182] The server analyzes the collected data. This analysis includes preprocessing steps such as data cleansing and feature extraction. For example, categories of products frequently purchased by a user are extracted from their purchase history, and this is then passed through a clustering algorithm to classify users into multiple clusters. This makes it clear what interests the user has.

[1183] Preparing for generative AI

[1184] The server prepares the analysis results to be fed into a generative AI model. For example, a natural language processing model is used for text generation, and an image generation model is used for image generation. The generative AI model receives input such as user features and cluster information.

[1185] Content Generation

[1186] The server uses a generative AI model to generate text and images optimized for each user. For example, to generate a travel guide article for "User A," the server generates text including a title such as "Top 10 Adventure Travel Destinations" and detailed information about the adventure trip. In addition, related beautiful landscape images are also generated.

[1187] Creating a Landing Page

[1188] The server creates a landing page based on the generated content. Specifically, it inserts the generated text and images into a pre-defined template, builds the page structure using HTML and CSS, and finally deploys the page on a web server so that users can access it through their browsers.

[1189] Specific examples

[1190] User data collection

[1191] User B frequently purchases outdoor equipment from a shopping site and frequently reads articles about hiking using a certain device. The server collects this data via an API and stores it in a local database.

[1192] Data analysis

[1193] The server analyzes User B's data and classifies him / her into a cluster called "outdoor enthusiasts" based on the categories of outdoor equipment he / she has purchased and the content of articles he / she has viewed. It then identifies the areas and activities in which User B is particularly interested.

[1194] Preparing for generative AI

[1195] The server feeds the analysis results into the generative AI model and prepares to generate optimal content for User B. Specifically, it inputs data indicating that User B is interested in hiking into the generative AI.

[1196] Content Generation

[1197] The server uses generative AI to generate text content such as "Top 10 hiking trails recommended for User B." It also generates beautiful landscape images related to each hiking trail.

[1198] Creating a Landing Page

[1199] The server places the generated text and images into a template to create a landing page for the hiking guide, and finally deploys this page to a web server, making it accessible to User B.

[1200] In this way, the system can efficiently provide optimized information to User B.

[1201] The processing flow will be explained below.

[1202] Step 1:

[1203] The server collects data such as user behavior, purchase history, and search history. For example, information about the pages a user visits on a shopping site, the products they purchase, and the queries they search for is obtained via API and stored in a local database. This allows detailed information about the user's preferences to be gathered.

[1204] Step 2:

[1205] The server performs preprocessing, cleaning the data. Specifically, it fills in missing values ​​and removes outliers. It also standardizes the data format and standardizes each data point. For example, it converts all purchase history prices into the same currency unit.

[1206] Step 3:

[1207] The server analyzes the preprocessed data and extracts features to identify the user's interests and preferences. For example, features could include the product categories the user frequently purchases or the themes of the articles they view. Text analysis and category classification techniques are used to extract these features.

[1208] Step 4:

[1209] The server uses a clustering algorithm to group users, for example, using the K-means algorithm to classify users with similar interests into the same cluster. This clustering helps clarify the areas of interest and concern of users.

[1210] Step 5:

[1211] The server then feeds the data to a generative AI model based on the analysis results. For example, if a user belongs to a cluster called "Outdoors Lovers," that information is input to the generative AI. These generative AI models include text generation models (e.g., GPT-4) and image generation models (e.g., DALL-E).

[1212] Step 6:

[1213] The server uses the generative AI model to generate text and images optimized for each user. For example, for "User B," it generates text content such as "The 10 Best Hiking Trails" and related beautiful scenery images.

[1214] Step 7:

[1215] The server places the generated text and images into a landing page template, building the page layout using HTML and CSS and inserting the generated content in the appropriate places, resulting in a landing page with a consistent design.

[1216] Step 8:

[1217] The server deploys the completed landing page to the web server, allowing users to access the landing page through their browsers. User B then views the "10 Best Hiking Trails" page.

[1218] In this way, the system collects data, analyzes it, generates AI, places content, and deploys it step by step, quickly and efficiently providing users with optimized landing pages.

[1219] Example 1

[1220] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1221] With current technology, efficiently generating landing pages tailored to user preferences requires a lot of manual work, making it difficult to quickly provide optimal content.In addition, there are limited methods for effectively utilizing user behavior data, purchase history, search history, etc., making it difficult to generate personalized content for users.

[1222] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1223] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, and search history, means for analyzing the data and identifying the user's interests and preferences, means for generating prompt sentences based on the analysis results and feeding the data to a generative AI model, means for generating text and images optimized for the user using the generative AI model, and means for arranging the generated text and images to automatically create a landing page. This makes it possible to quickly and automatically generate an optimal landing page based on the user's preferences and efficiently provide content.

[1224] "User" means an end user of the Internet service or website.

[1225] "Behavioral data" refers to data such as page browsing history and click history when a user visits a website.

[1226] "Purchase history" refers to historical data about products a user has previously purchased from online shops, etc.

[1227] "Search history" refers to historical data about queries and search results a user has performed on the Internet.

[1228] "Data analysis" refers to a series of processes that aggregate, classify, and evaluate collected data.

[1229] "Hobbies and preferences" refers to data that indicates a user's interests, concerns, and preferences.

[1230] A "prompt" is an instruction given to a generative AI model.

[1231] A "generative AI model" is an artificial intelligence model that generates text or images from specified input data.

[1232] "Text" refers to sentences or strings of characters generated by a generative AI model.

[1233] "Image" refers to the visual data generated by a generative AI model.

[1234] A "landing page" is a web page created for a user to access in a web browser.

[1235] A "template" is a predefined layout or design framework used to automatically create a landing page.

[1236] "Deployment" is the process of placing a completed landing page on a web server so that it can be accessed by users.

[1237] A system embodying this invention efficiently collects and analyzes multiple data sets, such as user behavior data, purchase history, and search history, and automatically generates content optimized for each user using a generative AI model. Specific embodiments of this system are described below.

[1238] First, the server collects information such as user behavior data, purchase history, and search history. This data is obtained from the user's browser history, purchase history, search queries, etc. recorded by the device. The server obtains this data via API and stores it in a local database. For example, if a user purchases outdoor equipment on a shopping site or reads an article about hiking, that information will be collected.

[1239] The server then analyzes the collected data. At this stage, data cleansing is performed to remove noise and missing data. Feature extraction is also performed to identify users' interests based on their purchase history and browser history. For example, a clustering algorithm is used to separate the data into multiple clusters, categorizing users into clusters such as "outdoor enthusiasts" or "travel enthusiasts."

[1240] Based on the analysis results, the server generates a prompt sentence and prepares to feed it to the generative AI model. For example, if the analysis finds that the user has specific interests, it generates a prompt sentence such as, "Please tell me some recommended hiking trails for User A." The server then creates an instance of a generative AI model (e.g., GPT-3 or StyleGAN) and sets the prompt sentence as input.

[1241] The server inputs prompts into the generative AI model to generate text and images optimized for the user. For example, the generative AI model generates text content such as "Top 10 recommended hiking trails for user A" and also creates related beautiful landscape images. This automatically generates content that is individually optimized for each user.

[1242] Finally, the server creates a landing page based on the generated text and images. At this stage, the generated content is inserted according to a pre-defined template, and the page structure is built using HTML and CSS. The completed landing page is deployed to a web server, where users can access it through their browsers.

[1243] To illustrate how this system works, let's use a practical example. If User B is buying outdoor gear on a shopping site and reading an article about hiking, the server will collect this behavioral data and identify the user's interest as "outdoors enthusiast." As a result, a prompt message will be created that generates "10 recommended hiking trails for User B," and specific content will be generated by the generative AI model. This generated content will be placed on a landing page based on a template and deployed for User B's easy access.

[1244] An example of a prompt sentence is, "Please generate content that introduces hiking trails that User B is interested in. User B has recently purchased a lot of outdoor equipment and particularly enjoys hiking."

[1245] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1246] Step 1: Data collection

[1247] The server collects information such as user behavior data, purchase history, and search history. Specifically, the device sends the user's browser history, purchase records, and search queries to the server via API. The server stores this data in a local database. For example, if a user purchases outdoor equipment on a shopping site or reads an article about hiking, that information will be collected.

[1248] Input: User behavior data, purchase history, search history

[1249] Output: User data stored in a local database

[1250] Step 2: Data analysis

[1251] The server retrieves collected user data from a local database and first cleanses the data, removing missing data and noise. Next, it performs feature extraction to identify users' interests based on their purchase history and browser history. It then uses a clustering algorithm to categorize users into clusters such as "outdoor enthusiasts" or "travel enthusiasts."

[1252] Input: User data from a local database

[1253] Output: User features and cluster information

[1254] Step 3: Prompt generation

[1255] The server generates a prompt based on the results of the data analysis. For example, if the server finds that the user has a specific interest, it creates a prompt such as, "Please recommend some hiking trails for User A." This prepares the instruction sentence to be given to the generative AI model.

[1256] Input: Data analysis results (user features and cluster information)

[1257] Output: Generated prompt statement

[1258] Step 4: Input to the generative AI model

[1259] The server creates an instance of a generative AI model (e.g., GPT-3 or StyleGAN) and sets the prompt sentence generated in the previous step as input, so the model is ready to generate optimal content.

[1260] Input: Generated prompt text

[1261] Output: Input data to a generative AI model

[1262] Step 5: Content Generation

[1263] The server inputs the prompt into the generative AI model, which generates text and images optimized for the user. The generative AI model then generates specific content based on the prompt. For example, it generates text content such as "Top 10 recommended hiking trails for user A" and related beautiful landscape images.

[1264] Input: Prompt sentence for generative AI model

[1265] Output: Generated text and images

[1266] Step 6: Create a Landing Page

[1267] The server creates a landing page by arranging the generated text and images according to a pre-defined template, building the page structure using HTML and CSS, and inserting the generated content. The completed landing page is then deployed to a web server, where users can access it through their browsers.

[1268] Input: Generated text and images

[1269] Output: The generated landing page

[1270] This process flow allows the system to quickly and efficiently generate optimal landing pages based on user preferences.

[1271] (Application example 1)

[1272] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1273] Conventional internet advertising and landing page creation systems have not adequately provided optimal content based on users' hobbies and preferences, and have had difficulty providing content in real time. As a result, advertisements and content that do not match the user's interests are displayed, making effective marketing impossible. Furthermore, even when displaying advertisements using wearable devices such as smart glasses, there is a lack of a means to provide users with content optimized for them in real time. There is a need to solve these issues.

[1274] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1275] In this invention, the server includes a means for collecting multiple data such as user behavior data, purchase history, and search history, a means for analyzing the data to identify the user's interests and preferences, and a means for feeding the data to a generative AI model based on the analysis results to generate text and images optimized for the user, thereby arranging the generated text and images to automatically create landing pages or advertising content and display them on the smart glasses.

[1276] "User behavior data" refers to information such as the pages a user visits on the Internet, the links they click, and the time they spend browsing.

[1277] "Purchase history" refers to information such as the products and services a user has purchased in the past, the date and time of those purchases, and frequency of those purchases.

[1278] "Search history" refers to the search queries a user has made on a search engine and the history of the related search result pages they have visited.

[1279] A "generative AI model" is an artificial intelligence model trained on a large dataset, and refers to a technique for generating text or images based on specific input data.

[1280] A "clustering algorithm" refers to a statistical method for analyzing large amounts of data and grouping data that are highly similar.

[1281] A "landing page" is a web page that appears first when a user clicks on an advertisement or link, and contains content designed to encourage a specific action (such as purchasing a product or entering information).

[1282] "Smart glasses" are a type of wearable device that has the shape of glasses but has the ability to display visual information as augmented reality (AR).

[1283] "Template" means a document or web page structure with a specific format or style predefined for efficient layout and display of content.

[1284] The system for implementing this invention consists of a server, a terminal, and a user. The server collects and analyzes information such as user behavior data, purchase history, and search history, and uses AI to display optimal landing pages and advertising content on the smart glasses based on the results.

[1285] Program Overview

[1286] The system includes the following procedures:

[1287] 1. User data collection:

[1288] The server collects data such as user behavior, purchase history, and search history via API, and stores detailed user preference information in a local database.

[1289] 2. Data Analysis:

[1290] The server cleanses the collected data, extracts and analyzes features, and uses a clustering algorithm to classify users into different clusters based on their purchase history and the pages they have viewed.

[1291] 3. Prepare the generative AI:

[1292] The server creates a prompt sentence to feed the analysis results to a generative AI model (e.g., GPT-3), which includes the user's cluster information and interests.

[1293] 4. Advertising content generation:

[1294] A generative AI model generates optimized text and images based on the prompt, and the generated ad content is tailored to what the user should see.

[1295] 5. Display on smart glasses:

[1296] The server displays the generated content on the smart glasses in real time, allowing users to visually view the most suitable advertising content through the glasses they are wearing.

[1297] Specific examples

[1298] Let's say a user is walking through a particular shopping mall. The server determines from past behavioral data and purchase history that the user is interested in outdoor gear. Based on this analysis, it inputs the following prompt into the generative AI model:

[1299] plaintext

[1300] "Generate ads that are best suited to users in Cluster 1, whose primary interest is outdoor gear. Include current sales and popular items in the ad content."

[1301] Based on this prompt, the generative AI model generates advertising text, such as "The latest outdoor gear is on sale!", along with images of the products on sale. The server then displays this generated advertising content on the smart glasses in real time, providing users with the most appropriate information.

[1302] Hardware and software used

[1303] Hardware: Smart glasses (e.g. HoloLens, Google Glass)

[1304] software:

[1305] Data collection via API: RESTful API

[1306] Data analysis: NumPy, scikit-learn

[1307] Generative AI model: GPT-3 (OpenAI)

[1308] This makes it possible to provide advertising content that matches the user's preferences in real time.

[1309] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1310] Step 1:

[1311] The server collects user behavioral data, purchase history, and search history via API. Specifically, the server accesses each API endpoint and retrieves data. For example, behavioral data including the user's visited pages and browsing time, history of past purchases, and search engine query history are collected. This data is stored in a local database.

[1312] Input: API endpoint

[1313] Output: User data stored in a local database

[1314] Step 2:

[1315] The server cleanses the collected data and extracts features. First, it removes duplicate data and missing values ​​to clean the data. Next, it extracts the categories of products that users frequently purchase and the features of their search queries. For example, if a user frequently purchases outdoor equipment, the features of that category are extracted.

[1316] Input: Raw data in a local database

[1317] Output: Cleansed feature data

[1318] Step 3:

[1319] The server runs a clustering algorithm on the cleansed data to classify users into different clusters. For example, it uses KMeans clustering to classify users into clusters such as "outdoor enthusiasts" and "tech gadget enthusiasts." This identifies the user's interests and preferences.

[1320] Input: Cleansed feature data

[1321] Output: User data categorized into clusters

[1322] Step 4:

[1323] The server then creates prompts to be input into the generative AI model based on the clustered user data. For example, for a user who loves the outdoors, the server might generate a prompt like, "Generate ads that are optimal for users in cluster 1. Their primary interest is outdoor gear. Include current sales information and popular products in the ad content."

[1324] Input: User data categorized into clusters

[1325] Output: A prompt to be input to the generative AI model

[1326] Step 5:

[1327] The server inputs a prompt into a generative AI model (e.g., GPT-3) to generate text and images. Based on the prompt, the generative AI model generates optimal ad text and related images. For example, ad text such as "Latest outdoor gear on sale!" and images of outdoor gear on sale are generated.

[1328] Input: prompt statement

[1329] Output: Generated ad text and images

[1330] Step 6:

[1331] The server transmits the generated advertising text and images to the smart glasses for display in real time, and the smart glasses visually display the received advertising content, allowing the user to easily check the information.

[1332] Input: Generated ad text and image

[1333] Output: Advertising content displayed on smart glasses

[1334] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1335] The system for implementing this invention comprises multiple servers, terminals, and users. This system includes an emotion engine that analyzes various user data and uses generative AI to automatically generate the optimal landing page (LP) for the user. This emotion engine analyzes the user's text messages, voice input, images, and videos, identifies the user's emotions, and generates optimized content. This allows for efficient provision of content that takes user emotions into consideration.

[1336] Program processing

[1337] User data collection

[1338] The server collects multiple data sets, including user behavior data, purchase history, and search history. Specifically, information about the pages the user visited, the products they purchased, and the queries they searched is obtained via API and stored in a local database. It also uses an emotion engine to collect the emotions expressed by the user during interactions.

[1339] Data analysis

[1340] The server analyzes the collected data. This analysis includes data cleaning and feature extraction. For example, it extracts the categories of products frequently purchased by users from their purchase history and classifies users into different clusters using a clustering algorithm. Furthermore, it uses an emotion engine to analyze emotions from users' text messages and voice inputs to identify what content is more appropriate.

[1341] Preparing for generative AI

[1342] The server feeds data to a generative AI model based on the analysis results. For example, if a user has a cluster of "loves the outdoors" and an emotion of "joy," that information is input into the generative AI. These generative AI models include text generation models (e.g., GPT-4) and image generation models (e.g., DALL-E).

[1343] Content Generation

[1344] The server uses a generative AI model to generate text and images optimized for each user. For example, for "User B," it generates text content such as "The 10 Best Hiking Trails," along with related images of beautiful scenery that evoke a sense of enjoyment. Furthermore, if the user expresses the emotion of "joy," it selects positive words and images that further enhance this emotion.

[1345] Creating a Landing Page

[1346] The server places the generated text and images into a landing page template, building the page layout using HTML and CSS and inserting the generated content in the appropriate places, resulting in a landing page with a consistent design.

[1347] Deploy

[1348] The server deploys the completed landing page to the web server, which allows users to access the landing page through their browsers. User B views the "10 Best Hiking Trails" page, which is designed to provide emotional satisfaction.

[1349] Specific examples

[1350] User data collection

[1351] User B frequently purchases outdoor equipment on a shopping site and reads articles about hiking using a certain device. User B expresses emotions such as "joy" and "excitement" while interacting with the site. The server collects this information via an API and stores it in a local database.

[1352] Data analysis

[1353] The server analyzes User B's data and identifies a cluster called "outdoor enthusiasts" based on the categories of outdoor equipment purchased and the content of articles viewed. It also analyzes User B's emotional data using an emotion engine to identify "joy" and "excitement."

[1354] Preparing for generative AI

[1355] The server feeds the analysis results into a generative AI model. For example, it inputs the cluster "loves the outdoors" and the emotion information "joy." The generative AI generates optimal text and images based on a large amount of data.

[1356] Content Generation

[1357] The server uses the generative AI model to generate text content such as "10 recommended hiking trails for User B." It also generates beautiful scenic images of hiking trails that User B is interested in. It uses positive words and colors that reinforce User B's emotion of "joy."

[1358] Creating a Landing Page

[1359] The server places the generated text and images into a template, inserting the generated content into a travel guide template and building the page layout.

[1360] Deploy

[1361] The server deploys the completed landing page to the web server, and User B accesses this page through a browser, which also provides emotional satisfaction.

[1362] In this way, the system collects data, analyzes it, analyzes emotions, generates AI, places content, and deploys it at each step, enabling it to quickly and efficiently provide optimal landing pages that also take user emotions into consideration.

[1363] The processing flow will be explained below.

[1364] Step 1:

[1365] The server collects multiple data such as user behavior data, purchase history, search history, etc. For example, information about the pages a user visits on a shopping site, the products they purchase, and the queries they search for is obtained via API and stored in a local database. Additionally, the emotion engine analyzes users' text messages and voice inputs to collect emotion data.

[1366] Step 2:

[1367] The server preprocesses the collected data by cleaning it (imputing missing values, removing outliers), standardizing it, and encoding categorical data (for example, standardizing currency units for purchase history and standardizing search queries).

[1368] Step 3:

[1369] The server analyzes the preprocessed data to identify users' interests and preferences. Specifically, it extracts the categories of products frequently purchased by users and the themes of articles they read, and then uses a clustering algorithm to classify users into different clusters. It also takes into account the user's emotional data analyzed by the emotion engine.

[1370] Step 4:

[1371] The server prepares the analysis results to feed into the generative AI model. For example, if the user has the emotion of "joy" in addition to the cluster of "loves the outdoors," that information is input into the generative AI model.

[1372] Step 5:

[1373] The server generates text content and images using a generative AI model. For example, it creates the text "10 Best Hiking Trails" for User B and generates related beautiful landscape images. It also selects positive phrases and vividly colored images to reinforce the user's emotion of "joy."

[1374] Step 6:

[1375] The server places the generated text and images into a landing page template, constructs the page layout using HTML and CSS, and inserts the generated content in the appropriate places, such as placing the title "10 Best Hiking Trails" in the page header, followed by the following text and images, one paragraph at a time.

[1376] Step 7:

[1377] The server deploys the completed landing page to a web server, allowing users to access the landing page through their browsers. For example, User B can enter the URL in his internet browser and view the generated hiking trail page.

[1378] Specific examples

[1379] User data collection

[1380] User B frequently browses articles about buying outdoor gear and hiking using his device. The server collects this information via API and stores his purchase history, visited pages, and search queries in a local database. The emotion engine analyzes text and voice data expressing User B's emotions such as "joy" and "excitement" and stores this emotion data as well.

[1381] Data analysis

[1382] The server cleans and standardizes User B's data before analyzing it. Based on the categories of outdoor equipment purchased and the content of articles viewed, User B is classified into a cluster called "Outdoor Lovers." User B's emotional data, "Joy," is also taken into account.

[1383] Preparing for generative AI

[1384] The server prepares to feed the data to the generative AI model based on the analysis results. The cluster of "outdoor lover" and the emotion information of "joy" are input into the generative AI model.

[1385] Content Generation

[1386] The server uses a generative AI model to generate images related to the text "10 recommended hiking trails for User B." It also selects positive phrases and brightly colored images to amplify User B's emotion of "joy."

[1387] Creating a Landing Page

[1388] The server places the generated text and images into a travel guide template, using HTML and CSS to place "10 Best Hiking Trails" in the page header and insert the generated content for each paragraph.

[1389] Deploy

[1390] The server deploys the completed landing page to the web server. User B can then access the page in a browser, view the generated content, and experience emotional satisfaction. This system makes it possible to provide information optimized for User B's interests and emotions.

[1391] Example 2

[1392] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1393] Current web content generation systems have difficulty efficiently providing personalized content that fully considers each user's emotions and behavioral history. In particular, there is a need for a method that can collect and analyze a variety of user data in real time, automatically generate content that reflects the user's emotions, and provide it as an optimal landing page.

[1394] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1395] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, search history, text messages, and voice data, means for cleaning the collected data, extracting features, and analyzing them, means for analyzing user sentiment from the analyzed data, means for feeding data to a generative AI model based on the analysis results and generating text and images optimized for the user, and means for arranging the generated text and images and automatically creating a landing page, thereby making it possible to efficiently provide an optimal landing page based on the user's sentiments and preferences.

[1396] "Behavioral data" is information about specific actions a user takes on a website or application, such as clicks, views, or purchases.

[1397] "Purchase History" refers to the products and services a User has previously purchased and related details.

[1398] "Search history" refers to the history of queries a user has made on search engines and websites, and is data that indicates the user's interests and concerns.

[1399] "Text Message" means any written communication sent or received by a User on a Digital Platform.

[1400] "Voice Data" refers to voice information input by a user or recorded by a system.

[1401] "Means of collection" refers to the methods and technologies used to acquire and store user behavioral data, purchase history, search history, text messages, voice data, etc. in a system.

[1402] "Cleaning" refers to the process of removing unnecessary information from collected data or filling in gaps in the data.

[1403] "Means for extracting features" refers to techniques and methods for extracting important patterns and information from collected data and using them for analysis.

[1404] "Analysis methods" refers to the algorithms and technologies used to process collected data and identify user preferences and behavioral patterns.

[1405] "Means for analyzing emotions" refers to technology for reading and classifying emotions from users' text messages and voice data.

[1406] A "generative AI model" refers to artificial intelligence technology that generates text and images based on large amounts of data.

[1407] "Landing page" means the web page a user first arrives at after clicking on an internet advertisement, email link, etc.

[1408] "Positioning means" refers to a technique for appropriately positioning the generated text and images on the landing page.

[1409] The system embodying this invention collects and analyzes user behavior data, purchase history, search history, text messages, voice data, etc. to identify the user's hobbies, preferences, and emotions, and automatically generates an optimal landing page based on these. It is particularly characterized by the use of an emotion engine for analyzing user emotions and a generative AI model.

[1410] The server collects data about the various actions users take on websites and applications. Specifically, it periodically retrieves behavioral data, purchase history, and search history via APIs and stores them in a local database. It also collects text messages and voice data sent and received by users within the site in real time.

[1411] The collected data is cleaned by the server, with unnecessary information removed and missing values ​​filled in. Next, features are extracted from the data, such as frequently purchased product categories determined from a user's purchase history. Furthermore, a clustering algorithm is used to classify users into different clusters. This process allows the user's behavioral patterns and preferences to be identified.

[1412] The server then performs sentiment analysis using the collected text messages and voice data. For sentiment analysis, a sentiment analysis model using natural language processing technology is used to identify emotions from the user's text messages and voice data and assign emotion labels.

[1413] Based on the analysis results, the server feeds the data to the generative AI model. This generative AI model includes a text generation model and an image generation model. For example, if a user "likes the outdoors" and has "emotions of joy," that information is input as a prompt to the generative AI model. An example of a prompt is shown below:

[1414] "Users love outdoor activities and feel joy. Can you recommend some hiking trails?"

[1415] The server receives the text and images returned by the generative AI model and uses them to generate content optimized for the user. For example, it generates text content such as "10 Best Hiking Trails" and related images of beautiful scenery. This generated content is then placed on a landing page based on a template pre-configured by the server. Specifically, the page layout is constructed using HTML and CSS, and the generated text and images are inserted in the appropriate locations.

[1416] Finally, the server deploys the completed landing page to the web server, allowing users to access it through their browsers, allowing users to view the optimal landing page based on their preferences and emotions and achieve emotional satisfaction.

[1417] This system is able to efficiently provide optimal landing pages that also take user emotions into consideration through a series of steps, from data collection and analysis, sentiment analysis, AI generation, content placement, and deployment.

[1418] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1419] Step 1: Collect user data

[1420] The server collects data on user behavior on the website, purchase history, search history, text messages, voice data, etc. Specifically, it may use an API to send requests such as the following to obtain data:

[1421] GET / api / user_behavior?user_id={user ID}&event_type=purchase

[1422] This allows details of the products a user purchases, search history, text messages, voice data, etc. to be collected and stored in a local database.

[1423] Input: User behavior data, purchase history, search history, text messages, voice data

[1424] Output: Collected data stored in a local database

[1425] Step 2: Data cleaning and feature extraction

[1426] The server cleans the collected data, removing unnecessary information and formatting it. Specifically, it removes duplicate data and fills in missing values. Next, it extracts features from purchase and search histories. For example, it identifies the categories of products frequently purchased by users.

[1427] python

[1428] import pandas as pd

[1429] data = pd.read_csv('user_data.csv')

[1430] data.drop_duplicates(inplace=True)

[1431] data.fillna(method='ffill', inplace=True)

[1432] popular_categories = data['category'].value_counts().head(5)

[1433] Input: Collected user data

[1434] Output: Cleaned data and extracted features

[1435] Step 3: Cluster the data

[1436] The server classifies users into different clusters based on their features using a clustering algorithm (e.g., K-means), which groups users with similar behavioral patterns.

[1437] python

[1438] from sklearn.cluster import KMeans

[1439] kmeans = KMeans(n_clusters=5)

[1440] data['cluster'] = kmeans.fit_predict(data[['feature1', 'feature2']])

[1441] Input: Cleaned data and extracted features

[1442] Output: User data sorted into clusters

[1443] Step 4: Sentiment Analysis

[1444] The server performs sentiment analysis using text messages and voice data. For sentiment analysis, a sentiment analysis model using natural language processing technology is used. For example, sentiment labels are assigned from text in the following format:

[1445] python

[1446] from transformers import pipeline

[1447] sentiment_analysis = pipeline('sentiment-analysis')

[1448] results = sentiment_analysis("This hike was amazing!")

[1449] Input: Text messages and voice data

[1450] Output: Data with emotion labels

[1451] Step 5: Prepare the generative AI model and generate prompts

[1452] The server feeds data to the generative AI model based on the analysis results, and prepares prompt sentences to input to the generative AI model.

[1453] python

[1454] prompt = "User enjoys outdoor activities and has the emotion joy. What hiking trails do you recommend?"

[1455] ai_model_input = {

[1456] "prompt": prompt,

[1457] "user_cluster": user_cluster,

[1458] "emotion_label": emotion_label

[1459] }

[1460] Input: Cluster labels and emotion labels

[1461] Output: Prompt and data to be fed into the generative AI model

[1462] Step 6: Content Generation

[1463] The server generates text and images using generative AI models, for example, it gets the text content "10 Best Hiking Trails" from a text generation model and related scenic images from an image generation model.

[1464] python

[1465] response = ai_model.generate(prompt)

[1466] Image generation

[1467] image_prompt = "Beautiful hiking trail views"

[1468] generated_image = image_model.generate(image_prompt)

[1469] Input: Prompt statement (text and image)

[1470] Output: Generated text and images

[1471] Step 7: Create a Landing Page

[1472] The server places the generated text and images into a landing page based on a pre-defined template, and the page layout is built using HTML and CSS.

[1473] html

[1474]

[1475]

[1476] <title> Recommended hiking trails< / title>

[1477]

[1478]

[1479] <h1> 10 Great Hiking Trails< / h1>

[1480] Text content...

[1481]

[1482]

[1483]

[1484] Input: Generated text and images

[1485] Output: The created landing page

[1486] Step 8: Deploy

[1487] The server deploys the completed landing page to a web server, which allows users to access the landing page through their browsers.

[1488] shell

[1489] scp landing_page.html user@webserver: / var / www / html /

[1490] Input: The created landing page

[1491] Output: Deployed landing page

[1492] Through these steps, the optimal landing page based on the user's emotions and preferences is efficiently provided.

[1493] (Application example 2)

[1494] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1495] Conventional landing page generation systems and product recommendation systems perform analysis based on user behavior data and purchase history, but there is a need to achieve more accurate individual optimization by incorporating the user's emotional state. Specifically, there is a problem in that it is difficult to provide optimal content based on the user's physical emotional state. This has resulted in many cases where these systems are ineffective in improving the user experience or stimulating purchasing motivation.

[1496] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1497] In this invention, the server includes means for collecting multiple data such as user behavior data, purchase history, and search history, means for analyzing the data to identify the user's hobbies, preferences, and emotional state, means for feeding the data to a generative AI model based on the analysis results to generate text and images optimized for the user, and means for arranging the generated text and images to automatically create a landing page or a recommended product list, thereby enabling the provision of individually optimized content that also takes into account the user's emotional state.

[1498] "User behavior data" refers to a record of the actions and behavior a user performs online, including, specifically, information such as the web pages visited, the links clicked, and the duration of viewing.

[1499] "Purchase history" refers to a record of products purchased by a user in the past, and includes information such as product name, purchase date and time, quantity, total amount, etc.

[1500] "Search history" refers to a record of search queries a user has made on the Internet, including the keywords searched, the date and time of the search, and click information for search results.

[1501] "Means of collection" refers to the methods and technologies used to obtain user behavioral data, purchase history, search history, etc., and includes data acquisition using APIs and tracking using cookies.

[1502] "Means of analysis" refers to methods and techniques for internally processing collected data and extracting useful information, including techniques such as data cleaning, feature extraction, and clustering.

[1503] "Hobbies and preferences" refers to areas or categories in which a user has particular interests or concerns, including preferences for specific products and interest in specific activities.

[1504] An "emotional state" refers to the emotional state a user feels at a particular time or in a particular situation, and includes emotions such as joy, excitement, sadness, and anger.

[1505] "Generative AI model" refers to a model that uses artificial intelligence to generate new content from specific input data, and includes technologies used for text generation and image generation (e.g., GPT and DALL-E).

[1506] "Means of feeding" refers to the methods and technologies for supplying analysis results and other necessary data to the generative AI model, including data preprocessing and input format adjustment.

[1507] "Text and Images" means the part of the content provided to users, including written and visual information.

[1508] A "landing page" refers to a web page accessed when an advertisement or link is clicked, and includes pages that provide specific information about a particular product or service.

[1509] A "recommended product list" is a series of products recommended to users based on their interests and is optimized based on their past purchasing history, behavioral data, emotional state, etc.

[1510] "Means of automatic creation" refers to methods and technologies that allow a system to automatically generate and arrange content without requiring analog manual work, including template-based arrangement and automatic layout generation.

[1511] This invention is a system that collects and analyzes data such as user behavior data, purchase history, and search history on the Internet, and provides individually optimized content that also takes into account emotional state. Specific implementation methods for realizing this system are described below.

[1512] Collection Stage

[1513] The server uses APIs to collect multiple data such as user behavior, purchase history, and search history. This data includes detailed information such as which web pages users visited, which links they clicked, and which product reviews they read. The collected data is stored in a local database for further analysis.

[1514] Analysis stage

[1515] The server cleans the collected data and extracts features. This includes removing noise and standardizing the format. It then uses a clustering algorithm and a sentiment analysis engine to classify users into different clusters and identify their hobbies, preferences, and interests. At the same time, it analyzes the user's text messages and voice inputs to understand their emotional state. Once this analysis is complete, it can be determined that the user is currently in a specific emotional state, such as "joy" or "excitement."

[1516] Generation stage

[1517] Based on the analysis results, the server feeds data to a generative AI model. For example, the user may input information such as "I love the outdoors" and being in a "joy" emotional state. This generative AI uses large-scale language models and image generation models (e.g., GPT-4 and DALL-E). The generative AI model generates text and images optimized for the user. For example, it generates text content such as "Top 10 recommended hiking trails" and images depicting the beautiful scenery of those hiking trails.

[1518] Placement and display stage

[1519] The generated text and images are placed in the appropriate positions based on a pre-defined template. The server uses these artifacts to automatically create a landing page and a list of recommended products. The template uses HTML and CSS to maintain a consistent design. The final landing page is deployed to a web server, where it can be accessed by users through a browser.

[1520] Specific examples

[1521] User A frequently purchases outdoor equipment from a specific shopping site and also frequently reads articles about hiking. After collecting this user's data and conducting clustering and sentiment analysis, the user was classified into a cluster called "Outdoors Lover" and identified as being in a "Joy" emotional state. The server fed this information into a generative AI model, which generated images of beautiful scenery associated with the text "10 Recommended Hiking Trails." This information was then arranged as a landing page based on a template and deployed.

[1522] Prompt Sentence Examples

[1523] An example of a prompt sentence to be input to the generative AI model is, "Create the optimal landing page for a user who is interested in outdoor gear and is currently in the emotional state of 'joy'."

[1524] In this way, this invention takes into consideration the emotional state of the user and generates and provides individually optimized content, thereby significantly improving the user experience.

[1525] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1526] Step 1: Collect user data

[1527] The server uses an API to collect multiple data such as user behavior data, purchase history, and search history. Specific operations include obtaining web page log data, clickstreams, and detailed data on viewed products. The input is data related to the user's online operations, and the output is an organized set of user behavior data.

[1528] Step 2: Data cleaning and feature extraction

[1529] The server cleans the collected data and extracts features. Specifically, it removes noise and standardizes the data format. The input is raw user data, and the output is clean data and feature-extracted data.

[1530] Step 3: Clustering and sentiment analysis

[1531] The server performs clustering and sentiment analysis on the cleaned data. Using a clustering algorithm, users are classified into different clusters. It also uses a sentiment analysis engine to analyze users' text messages and voice data to identify their emotional state. The inputs are the cleaned data and feature-extracted data, and the output is cluster information and emotional state information.

[1532] Step 4: Feed data into the generative AI model

[1533] The server feeds cluster information and emotional state information to the generative AI model. For example, input information for a user who is an "outdoor lover" and has the emotion of "joy." The input is the cluster information and emotional state information, and the output is the input data for the generative AI model.

[1534] Step 5: Generate optimized content

[1535] The server uses a generative AI model to generate text and images optimized for the user. For example, it generates text content such as "Top 10 recommended hiking trails" and images of beautiful scenery. The input is the input data to the generative AI model, and the output is the generated text and images. For generation, a text generation model such as GPT-4 and an image generation model such as DALL-E are used.

[1536] Step 6: Arranging Content

[1537] The server places the generated text and images in the appropriate locations based on a pre-defined template. Specifically, it uses HTML and CSS to build a landing page or recommended product list. The input is the generated text and images, and the output is the HTML code for the completed web page.

[1538] Step 7: Deploy your landing page

[1539] The server deploys the completed landing page and recommended product list to a web server, where users can access these pages through their browsers. The input is the HTML code for the completed web page, and the output is a landing page accessible on the web.

[1540] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1541] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1542] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1543] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1544] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1545] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1546] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1547] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1548] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1549] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1550] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1551] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1552] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1553] 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.

[1554] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1555] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1556] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1557] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1558] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1559] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1560] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1561] The following is further disclosed regarding the above embodiment.

[1562] (Claim 1)

[1563] A means of collecting multiple data such as user behavior data, purchase history, and search history,

[1564] A means for analyzing the data and identifying the user's interests and preferences;

[1565] means for feeding data to a generative AI model based on the analysis results to generate user-optimized text and images;

[1566] means for automatically creating a landing page by arranging the generated text and images;

[1567] A system including:

[1568] (Claim 2)

[1569] 2. The system of claim 1, wherein the means for identifying the user's interests and preferences uses a clustering algorithm to classify users into different clusters.

[1570] (Claim 3)

[1571] 10. The system of claim 1, wherein the means for automatically creating the landing page arranges generated text and images based on a pre-defined template.

[1572] "Example 1"

[1573] (Claim 1)

[1574] A means of collecting multiple data such as user behavior data, purchase history, and search history,

[1575] A means for analyzing the data and identifying the user's interests and preferences;

[1576] means for generating prompt sentences based on the analysis results and feeding the data to a generative AI model;

[1577] means for generating user-optimized text and images using the generative AI model;

[1578] means for automatically creating a landing page by arranging the generated text and images;

[1579] A system including:

[1580] (Claim 2)

[1581] 2. The system of claim 1, wherein the means for identifying the user's interests and preferences uses a clustering algorithm to classify users into different clusters.

[1582] (Claim 3)

[1583] 10. The system of claim 1, wherein the means for automatically creating the landing page arranges generated text and images based on a pre-defined template.

[1584] "Application Example 1"

[1585] (Claim 1)

[1586] A means of collecting multiple data such as user behavior data, purchase history, and search history,

[1587] A means for analyzing the data and identifying the user's interests and preferences;

[1588] means for feeding data to a generative AI model based on the analysis results to generate user-optimized text and images;

[1589] means for arranging the generated text and images to automatically create a landing page or advertising content and display it on the smart glasses;

[1590] A system including:

[1591] (Claim 2)

[1592] 2. The system of claim 1, wherein the means for identifying the user's interests and preferences uses a clustering algorithm to classify users into different clusters.

[1593] (Claim 3)

[1594] 10. The system of claim 1, wherein the means for automatically creating a landing page or advertising content arranges generated text and images based on a pre-defined template.

[1595] "Example 2: Combining Emotion Engines"

[1596] (Claim 1)

[1597] A means of collecting multiple data such as user behavior data, purchase history, search history, text messages, and voice data;

[1598] means for cleaning the collected data, extracting features, and analyzing the data;

[1599] A means for analyzing user emotions from the analyzed data;

[1600] means for feeding data to a generative AI model based on the analysis results to generate user-optimized text and images;

[1601] means for automatically creating a landing page by arranging the generated text and images;

[1602] A system including:

[1603] (Claim 2)

[1604] 2. The system of claim 1, wherein the means for identifying the user's interests and preferences uses a clustering algorithm to classify users into different clusters.

[1605] (Claim 3)

[1606] 10. The system of claim 1, wherein the means for automatically creating the landing page arranges generated text and images based on a pre-defined template.

[1607] "Application example 2 when combining emotion engines"

[1608] (Claim 1)

[1609] A means of collecting multiple data such as user behavior data, purchase history, and search history,

[1610] means for analyzing said data to identify the user's preferences and emotional state;

[1611] means for feeding data to a generative AI model based on the analysis results to generate user-optimized text and images;

[1612] means for arranging the generated text and images to automatically create a landing page or a recommended product list;

[1613] A system including:

[1614] (Claim 2)

[1615] 2. The system of claim 1, wherein the means for identifying the user's preferences and emotional state uses a clustering algorithm and a sentiment analysis engine to classify users into different clusters.

[1616] (Claim 3)

[1617] 2. The system of claim 1, wherein the means for automatically creating the landing page or recommended product list arranges generated text and images based on a pre-set template. [Explanation of symbols]

[1618] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of collecting multiple data such as user behavior data, purchase history, and search history, A means for analyzing the data and identifying the user's interests and preferences; means for feeding data to a generative AI model based on the analysis results to generate user-optimized text and images; means for automatically creating a landing page by arranging the generated text and images; A system including:

2. 2. The system of claim 1, wherein the means for identifying user preferences uses a clustering algorithm to classify users into different clusters.

3. The system of claim 1 , wherein the means for automatically creating the landing page arranges generated text and images based on a pre-defined template.

Images