system

Abstract

Provide a system. The present invention comprises: a means for acquiring web page content in response to a user's web access; means for analyzing web page content in real time and extracting key topics and keywords; A means for generating native advertisements using a generative AI based on the extracted topics and keywords; means for inserting the generated native advertisement into an appropriate position on a web page; A means of sending updated web pages to the device and redisplaying them. A system including:

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] Traditional online advertising often has the drawback of being inconsistent with the content users are viewing, causing discomfort to users. Furthermore, ads often appear unnaturally and do not blend in with the content, which can detract from the user experience. Furthermore, for advertisers, this can reduce the effectiveness of their ads and lower their profitability. [Means for solving the problem]

[0005] To solve this problem, the present invention provides the following means. First, when a user accesses a webpage, a means is used to obtain the content of the webpage. Next, a means is provided to analyze the content of the webpage in real time and extract key topics and keywords. Based on the extracted topics and keywords, a means is introduced to generate native advertisements using generative AI. The generated native advertisements are adjusted to match the design and style of the webpage and inserted at appropriate positions. Finally, a means is provided to send the updated webpage to the terminal and redisplay it. Such a system allows advertisements to be displayed that blend into the content without causing a sense of incongruity to the user, improving the user experience.

[0006] "User" refers to any individual or entity that accesses, views, interacts with, or uses online content, such as web pages or applications.

[0007] "Web access" refers to a user's act of connecting to a specific web page or online service via the Internet and obtaining or transmitting data.

[0008] "Web page content" refers to the collection of information, such as text, images, video, audio, design, and layout, that appears within a particular web page.

[0009] "Real-time analysis" refers to a processing method in which data is processed immediately as it is generated or acquired, and the results are reflected instantly.

[0010] A "topic" refers to a subject or theme on which particular content or information is concentrated.

[0011] "Keywords" refer to words or phrases that are particularly important in describing specific information or content.

[0012] "Generative AI" refers to systems or algorithms that use artificial intelligence technology to automatically generate data such as text, images, and audio.

[0013] "Native advertising" refers to an advertising format that blends naturally with the content of a web page or application and is displayed in a way that does not cause discomfort to the user.

[0014] "Appropriate placement" refers to where the ad should be inserted in a way that does not disrupt the user experience and is consistent with the flow and intent of the content.

[0015] "Terminal" refers to a device or equipment that allows a user to access the web and use online services, including PCs, smartphones, tablets, etc.

[0016] "Natural language processing (NLP)" refers to the technology that allows computers to process, understand, and generate human language. [Brief explanation of the drawings]

[0017] [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

[0018] 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.

[0019] First, the terms used in the following description will be explained.

[0020] 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).

[0021] 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.

[0022] 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.

[0023] 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.

[0024] 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."

[0025] [First embodiment]

[0026] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0027] 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.

[0028] 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).

[0029] 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.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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."

[0038] The present invention relates to a real-time native advertising generation system using generation AI. Specifically, the system analyzes the content of web pages accessed by users, dynamically generates advertisements that match that content, and inserts them into the web pages, thereby enhancing advertising effectiveness without causing discomfort to users.

[0039] Specific Embodiments of the System

[0040] System Configuration

[0041] This system mainly consists of three main components: the user's device, the server, and the generative AI model.

[0042] 1. User's device

[0043] A device that allows users to access the web. The device has the ability to send requests to web servers and display the retrieved content. Devices include PCs, smartphones, tablets, etc.

[0044] 2. Server

[0045] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also generates advertisements using AI and processes them to insert them into web pages. It also has the function of sending updated web pages to users' devices.

[0046] 3. Generative AI Models

[0047] The generative AI model runs on the server and generates text ads based on the analyzed topics and keywords. The generative AI model uses natural language processing (NLP) techniques.

[0048] Program processing

[0049] When a user accesses a web page, the following sequence of events occurs:

[0050] 1. User Web Access

[0051] A user accesses a specific web page. For example, the user clicks on a blog page with the theme "Autumn Kyoto Sightseeing."

[0052] The terminal sends a request to the server requesting the HTML content of the web page.

[0053] 2. Getting HTML data

[0054] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[0055] 3. Content Analysis

[0056] The server analyzes the HTML content and extracts key topics and keywords. For example, the analysis extracts keywords such as "autumn leaves," "famous places," and "guided tours."

[0057] 4. Ad generation

[0058] The server inputs the extracted keywords into a generative AI model to generate ad text related to the content, such as an ad for a "Kyoto Autumn Foliage Tour Special."

[0059] 5. Ad Styling

[0060] The server adjusts the style and layout of the generated advertising text to match the design of the web page.

[0061] 6. Ad Insertion

[0062] The server then regenerates the HTML for the web page, inserting the ad in the appropriate place, for example after the second paragraph.

[0063] 7. Submitting Updated Content

[0064] The server sends the updated HTML content to the user's device, which redisplays it.

[0065] Specific examples

[0066] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following happens:

[0067] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[0068] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[0069] The server inputs these keywords into the generation AI, which generates advertising text such as "Special feature on Kyoto autumn foliage tours."

[0070] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[0071] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[0072] In this way, the system realizes native advertising that does not cause discomfort to users, maximizes advertising effectiveness, and improves the user experience.

[0073] The processing flow will be explained below.

[0074] Step 1:

[0075] A user accesses a web page of interest. For example, a user accesses a blog post about "Autumn Kyoto Sightseeing." The device sends a request to that URL, requesting the HTML content of the web page from the server.

[0076] Step 2:

[0077] The server receives the user's request and retrieves the appropriate HTML content from a database or file system. The web server then prepares a response containing all the data for the web page (text, images, style sheets, etc.).

[0078] Step 3:

[0079] The server analyzes the HTML retrieved. Specifically, it uses an HTML parser to analyze the structure of the web page and extract the text content. For example, it extracts the entire body of an article titled "Sightseeing in Kyoto in Autumn."

[0080] Step 4:

[0081] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the extracted text. For example, keywords such as "autumn leaves," "famous places," and "guided tours" might be found. This process may involve the use of NLP libraries or APIs (e.g., SpaCy or NLTK).

[0082] Step 5:

[0083] The server initializes the generative AI model and provides the extracted keywords as input data. The generative AI model (e.g., GPT) is used to generate text for native ads based on the keywords. For example, an ad copy such as "Special feature on Kyoto autumn foliage tours" is generated. Here, a pre-trained model is used for the generative AI.

[0084] Step 6:

[0085] The server adjusts the layout and style of the generated advertisement to match the design of the web page, specifically by applying a CSS stylesheet so that the advertisement blends into the existing design of the web page. This adjustment may include font size, color, margins, etc.

[0086] Step 7:

[0087] The server updates the HTML content, inserting the generated ad in the appropriate place (e.g., adding an ad after the third paragraph of a blog post), and restructures the entire HTML document to ensure consistency.

[0088] Step 8:

[0089] The server sends the updated HTML content to the device, and the device reads the new HTML data received from the server and redraws the web page.

[0090] Step 9:

[0091] Users can browse the updated web page and continue reading the original content without noticing the generated native ad, making the user experience natural and seamless.

[0092] Example 1

[0093] 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."

[0094] Conventional ad delivery methods often lack a proper blend of web page content and ads, creating a sense of discomfort for users. Furthermore, to maximize advertising effectiveness, it is necessary to generate and display ads in real time that are closely related to the content being viewed by users. However, the technology to accomplish this efficiently has not yet been fully developed.

[0095] 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.

[0096] In this invention, the server includes means for acquiring web page information from user web access, means for analyzing the web page information in real time and extracting main topics and keywords, means for generating advertisements using a generative AI model based on the extracted topics and keywords, means for adjusting the layout and style of the generated advertisement to match the design of the web page, means for inserting the generated advertisement into an appropriate position on the web page, and means for transmitting the updated web page to the terminal and redisplaying it, thereby enabling advertisements to be displayed in a way that does not cause discomfort to the user and maximizing advertising effectiveness.

[0097] "Web access" refers to the act of a user connecting to a web page via the Internet and obtaining information.

[0098] "Web page information" is a collection of data such as text, images, videos, and links displayed on a website.

[0099] "Real-time" refers to processing occurring immediately without delay.

[0100] "Analysis" refers to examining the content of data in detail, especially using natural language processing to extract specific information.

[0101] A "topic" refers to the main topic or subject matter in text data.

[0102] "Keywords" are important words or phrases in a document that help identify the topic.

[0103] A "generative AI model" refers to an algorithm or system that uses artificial intelligence to generate new content or data.

[0104] "Advertising" is information used to promote products and services to consumers.

[0105] "Layout" refers to the position and structure of each element arranged within a web page.

[0106] "Style" refers to the appearance and design of text and images on a web page, including color, font, size, etc.

[0107] "Insertion" means adding new information or elements into an existing Web page.

[0108] A "terminal" is a device that a user uses to access the Internet. Examples include personal computers, smartphones, and tablets.

[0109] "Redisplay" means displaying updated content or data again on the user's terminal.

[0110] This invention relates to a real-time native advertising generation system using generative AI. Specifically, it provides a system that analyzes the content of web pages accessed by users, dynamically generates advertisements that fit the content, and inserts them into the web pages, thereby enhancing advertising effectiveness without causing discomfort to users.

[0111] System Configuration

[0112] The system mainly consists of three main components: the user's device, the server, and the generative AI model.

[0113] 1. User's device

[0114] A device that allows users to access the web. The device has the ability to send requests to web servers and display the retrieved content. Devices include PCs, smartphones, tablets, etc.

[0115] 2. Server

[0116] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also generates advertisements using AI and processes them to insert them into web pages. It also has the function of sending updated web pages to users' devices.

[0117] 3. Generative AI Models

[0118] The generative AI model runs on a server and generates text ads based on the analyzed topics and keywords. The generative AI model uses natural language processing (NLP) technology. Specific examples of such technology include generative models such as GPT-3 (registered trademark).

[0119] Server Processing

[0120] When a user accesses a specific web page, the device requests the HTML content of that web page from the server. The server receives the request and retrieves the relevant HTML data. The server then analyzes the retrieved HTML content to extract key topics and keywords. For example, NLTK and spaCy are used to extract keywords such as "autumn leaves," "famous places," and "guided tours."

[0121] Next, the server inputs the extracted keywords into a generative AI model to generate advertising text related to the content. For example, an advertisement such as "Special feature on Kyoto autumn foliage tours" may be generated. An example of a prompt used in this case is "Please generate advertising text on the themes of 'Kyoto autumn foliage,' 'famous places,' and 'sightseeing.'"

[0122] The server then adjusts the generated ad text to match the style and layout of the web page. Specifically, it uses CSS and JavaScript® to make the ad's appearance blend with the web page. The server then inserts the ad in the appropriate position and regenerates the web page's HTML. For example, the server might insert the ad after the second paragraph.

[0123] Finally, the server sends the updated HTML content to the user's device, which redisplays it, allowing the user to seamlessly browse the page.

[0124] Specific examples

[0125] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following will happen:

[0126] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[0127] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[0128] The server inputs these keywords into the generation AI, which generates advertising text such as "Special feature on Kyoto autumn foliage tours."

[0129] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[0130] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[0131] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0132] Step 1:

[0133] A user accesses a specific web page. Specifically, the user enters a URL in the browser's address bar or clicks a link. The URL is entered as input, and the device sends an HTTP request to the server requesting the HTML content corresponding to that URL. The server receives this request as output.

[0134] Step 2:

[0135] The server receives a user request and retrieves the relevant HTML content. Specifically, the server loads the HTML file corresponding to the requested URL from a database or file system. The input is the requested URL, and the output is the corresponding HTML content.

[0136] Step 3:

[0137] The server analyzes the retrieved HTML content and extracts key topics and keywords. Specifically, it uses a natural language processing (NLP) library to perform text analysis. For example, it uses "NLTK" or "spaCy." The input is the entire HTML content, and the output is extracted keywords such as "autumn leaves," "famous places," and "guided tours."

[0138] Step 4:

[0139] The server inputs the extracted keywords into the generative AI model and generates advertising text related to the content. Specifically, the server sends a prompt to the generative AI model. An example of this prompt is "Please generate advertising text on the themes of 'Kyoto autumn leaves,' 'famous places,' and 'sightseeing.'" The input is the extracted keywords and the prompt, and the output is advertising text such as "Special feature on Kyoto autumn leaves tours."

[0140] Step 5:

[0141] The server adjusts the style and layout of the generated ad text to match the web page design. Specifically, it uses CSS and JavaScript to adjust the font, color, size, etc. of the ad text to match the web page design. The input is the generated ad text, and the output is the adjusted ad text.

[0142] Step 6:

[0143] The server inserts the ad in the appropriate position and regenerates the HTML for the webpage, using JavaScript DOM manipulation and a server-side template engine. For example, inserting the ad after the second paragraph. The input is the adjusted ad text and the original HTML content, and the output is the updated HTML content with the ad inserted.

[0144] Step 7:

[0145] The server sends the updated HTML content to the user's device, which then redisplays it. Specifically, the newly generated HTML is returned to the device as an HTTP response. The input is the updated HTML content with the ad inserted, and the output is the updated web page displayed on the device. This allows the user to view the page seamlessly.

[0146] (Application example 1)

[0147] 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."

[0148] Modern internet advertising tends to be intrusive to users, often detracting from the user experience. Furthermore, static advertisements tend to be less relevant to the page content, reducing advertising effectiveness. Conventional advertising systems are unable to dynamically generate advertisements that match the content of web pages, which can be unsettling for users. The present invention aims to solve these problems.

[0149] 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.

[0150] In this invention, the server includes means for acquiring web page content through user web access, means for analyzing the web page content in real time and extracting key topics and keywords, and means for generating native advertisements using a generation AI based on the extracted topics and keywords. This makes it possible to improve the user experience while increasing advertising effectiveness by inserting dynamically generated advertisements in appropriate positions based on the content viewed by the user.

[0151] "User web access" refers to the act of a user accessing a web page via the Internet and obtaining or viewing information.

[0152] "Web page content" refers collectively to text, images, video, and other media or information displayed on a web page.

[0153] "Real-time analysis" is the process of analyzing the content of a web page immediately after a user accesses it.

[0154] "Major topics and keywords" refers to the important subjects and keywords in the content of a web page, which serve as the basis for generating advertisements.

[0155] "Generative AI" is a system that uses artificial intelligence techniques to generate new text based on specific prompts.

[0156] "Native advertising" is advertising that is displayed in a way that is integrated with the content of a web page and is designed to not cause any discomfort to the user.

[0157] A "dynamically generated advertisement based on content" is an advertisement that is automatically generated in real time according to the content of the web page that the user is viewing.

[0158] "Inserting in an appropriate position on a web page" refers to placing advertisements in appropriate locations on a web page so as not to impair the user experience.

[0159] "User experience" refers to the overall experience and satisfaction a user has when using a product or service.

[0160] An "application" is a software program that runs on a device and provides a particular function or service.

[0161] This invention is a system that uses generative AI to generate native ads in real time and insert them into the content of web pages accessed by users. This article mainly explains an example of applying this system to a news application for smartphones.

[0162] System Configuration

[0163] The system consists of three main components: the user's device, a server, and a generative AI model.

[0164] 1. User's device

[0165] This includes devices such as smartphones and tablets.

[0166] It provides an interface for users to access articles through a news app.

[0167] 2. Server

[0168] It plays a central role in retrieving and parsing the HTML content of web pages.

[0169] Generative AI is used to generate advertisements, and processing is performed to properly insert the advertisements into web pages.

[0170] 3. Generative AI Models

[0171] Natural language processing (NLP) techniques are used to generate text ads based on analyzed topics and keywords.

[0172] Generative AI models used include, for example, GPT-4 (registered trademark) from OpenAI (registered trademark).

[0173] Program processing overview

[0174] When a user accesses a web page, the server performs the following process.

[0175] 1. User Web Access

[0176] A user clicks on a particular article in a news app, and the device requests the HTML content of that news article from the server.

[0177] 2. Getting HTML data

[0178] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[0179] 3. Content Analysis

[0180] The server analyzes the HTML content and extracts key topics and keywords, such as "climate change," "renewable energy," and "eco-activities."

[0181] 4. Ad generation

[0182] The server inputs the extracted keywords into a generative AI model to generate advertising text related to the content. An example of a prompt sentence to be used is as follows:

[0183] Generate ad text based on the following keywords: climate change, renewable energy, eco-activism

[0184] 5. Ad Styling

[0185] The server adjusts the style and layout of the generated advertising text to match the design of the news article.

[0186] 6. Ad Insertion

[0187] The server inserts the advertisement at the appropriate position in the news article, for example after the second paragraph of the article.

[0188] 7. Submitting Updated Content

[0189] The server sends the updated HTML content to the user's device, which redisplays it.

[0190] Specific examples

[0191] For example, if a user accesses an article on the topic of "Thinking about Environmental Issues" in a news app, the following steps are taken:

[0192] The user clicks on the URL of a news article, and the device requests the HTML content from the server.

[0193] The server retrieves the HTML content, analyzes it, and extracts the keywords "climate change," "renewable energy," and "eco-activities."

[0194] The server inputs these keywords into a generation AI, which generates ad text such as "Introducing a new electric vehicle."

[0195] The server adapts the generated advertisements to the style of the news article and inserts them at the appropriate locations.

[0196] The updated page will be redisplayed on the user's device, allowing the user to continue reading the article without any discomfort.

[0197] This invention makes it possible to maximize advertising effectiveness by dynamically generating and inserting advertisements that match the content without impairing the user experience.

[0198] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0199] Step 1:

[0200] A user clicks on a specific article in a news app.

[0201] Input: The URL of the news article selected by the user.

[0202] Specific operation: The device sends a request to the server for the HTML content of the news article.

[0203] Output: The server receives the request.

[0204] Step 2:

[0205] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[0206] Input: A news article URL request.

[0207] Specific operation: The server searches for and retrieves the corresponding HTML content from the database.

[0208] Output: The retrieved HTML content.

[0209] Step 3:

[0210] The server parses the retrieved HTML content and extracts key topics and keywords.

[0211] Input: The retrieved HTML content.

[0212] What it does: The server uses natural language processing (NLP) techniques to parse the HTML content and extract key topics and keywords (e.g., "climate change," "renewable energy," "green activities").

[0213] Output: Extracted main topics and keywords.

[0214] Step 4:

[0215] The server inputs the extracted keywords into a generative AI model to generate advertising text relevant to the content.

[0216] Input: Extracted main topics and keywords.

[0217] Specific operation: The server sends the following prompt to the generated AI model:

[0218] Generate ad text based on the following keywords: climate change, renewable energy, eco-activism

[0219] A generative AI model generates ad text.

[0220] Output: The generated ad text.

[0221] Step 5:

[0222] The server adjusts the style and layout of the generated advertising text to match the design of the news article.

[0223] Input: The generated ad text.

[0224] What it does: The server uses CSS and HTML templates to adjust the ad design and integrate it into the style of the news article.

[0225] Output: The adjusted ad text.

[0226] Step 6:

[0227] The server inserts the advertisements into the news articles at appropriate locations.

[0228] Input: Your adjusted ad text.

[0229] Specific behavior: The server regenerates the HTML content of the news article, for example inserting an advertisement after the second paragraph of the article.

[0230] Output: The updated HTML content.

[0231] Step 7:

[0232] The server sends the updated HTML content to the user's device, and the user redisplays it through the news app.

[0233] Input: The updated HTML content.

[0234] What happens: The server sends HTML content to the device, and the news app receives and displays it.

[0235] Output: The user can view the updated news article.

[0236] These steps ensure that users are presented with ads that are tailored to their content and that fit their content seamlessly, improving the user experience.

[0237] 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.

[0238] This invention relates to a system that recognizes user emotions and automatically inserts native ads generated based on those emotions into web pages. Specifically, the system analyzes the content of web pages retrieved by users' web access in real time and extracts key topics and keywords. It also uses an emotion engine to recognize user emotions and generates native ads based on this emotion using a generation AI. The generated native ads are inserted in appropriate positions on the web page, and the updated web page is sent to the terminal. This system provides ads with optimal timing and content to users, achieving higher advertising effectiveness.

[0239] Specific Embodiments of the System

[0240] System Configuration

[0241] This system mainly consists of four main components: the user's device, the server, the generative AI model, and the emotion engine.

[0242] 1. User's device

[0243] A device that allows users to access the web. The terminal has the function of sending requests to a web server and displaying the retrieved content. It also has the function of collecting data such as the user's facial expressions, voice tone, and text input, and sending it to the server.

[0244] 2. Server

[0245] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also analyzes user emotion data, generates advertisements based on the emotions using generative AI, and processes the advertisements to be inserted into web pages. It also has the function of sending updated web pages to users' devices.

[0246] 3. Generative AI Models

[0247] The generative AI model runs on a server and generates text ads based on the analyzed topics and keywords, and also takes into account user sentiment data to generate ads with the appropriate tone and content.

[0248] 4. Emotion Engine

[0249] The emotion engine analyzes users' facial expressions, voice tone, and text input to recognize their emotional state in real time. This emotional data is fed into a generative AI model and used to generate ads.

[0250] Program processing

[0251] When a user accesses a web page, the following sequence of events occurs:

[0252] 1. User Web Access

[0253] When a user accesses a specific web page, for example, a blog post about "Autumn Kyoto Sightseeing," the device sends a request to the URL, requesting the HTML content of the web page from the server.

[0254] 2. Getting HTML data

[0255] The server receives the user's request and retrieves the relevant HTML data from a database or file system. The web server then prepares a response containing all the data for the web page (text, images, style sheets, etc.).

[0256] 3. Content Analysis

[0257] The server analyzes the HTML retrieved. Specifically, it uses an HTML parser to analyze the structure of the web page and extract the text content. For example, it extracts the entire body of an article titled "Sightseeing in Kyoto in Autumn."

[0258] 4. Topic and Keyword Extraction

[0259] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the extracted text. For example, keywords such as "autumn leaves," "famous places," and "guided tours" might be found. This process may involve the use of NLP libraries or APIs (e.g., SpaCy or NLTK).

[0260] 5. Emotional Recognition

[0261] While a user is browsing a web page, the device collects data such as facial expressions and voice and sends it to the emotion engine. The emotion engine analyzes this data and identifies the user's emotional state. For example, if the user is smiling, it is judged to be "happy," and if they are frowning, it is judged to be "unhappy."

[0262] 6. Ad Generation

[0263] The server inputs the extracted keywords and recognized emotion data into a generative AI model to generate native ads that match the emotion. For example, if the user is in a "happy" state, it generates an ad with a positive tone, and if the user is in a "dissatisfied" state, it generates a problem-solving ad.

[0264] 7. Ad Styling

[0265] The server adjusts the layout and style of the generated advertisement to match the design of the web page, specifically by applying a CSS stylesheet so that the advertisement blends into the existing design of the web page. This adjustment may include font size, color, margins, etc.

[0266] 8. Ad Insertion

[0267] The server updates the HTML content, inserting the generated ad in the appropriate place (e.g., adding an ad after the third paragraph of a blog post), and restructures the entire HTML document to ensure consistency.

[0268] 9. Submitting Updated Content

[0269] The server sends the updated HTML content to the device, and the device reads the new HTML data received from the server and redraws the web page.

[0270] Specific examples

[0271] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following happens:

[0272] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[0273] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[0274] The server inputs these keywords into the AI ​​generator, which generates ad text such as "Kyoto Autumn Foliage Tour Special." It also generates ad text with a positive tone based on the user's emotional data (e.g., "happiness").

[0275] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[0276] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[0277] In this way, the system provides advertisements that best suit the user's emotional state, maximizing advertising effectiveness and improving the user experience.

[0278] The processing flow will be explained below.

[0279] Step 1:

[0280] A user clicks on a URL in their web browser to access a blog post about "Autumn Kyoto Sightseeing." The device sends a request to the server for that URL, requesting the HTML content of the web page.

[0281] Step 2:

[0282] The server receives the user's request and retrieves the appropriate HTML content from a database or file system.

[0283] Step 3:

[0284] The HTML content retrieved by the server is analyzed using an HTML parser, extracting each element of the web page, such as text content, images, and style sheets.

[0285] Step 4:

[0286] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the text content, such as "autumn leaves," "famous places," and "guided tours."

[0287] Step 5:

[0288] While a user is browsing a web page, the device uses a camera and microphone to transmit the user's facial expressions and voice tone to the emotion engine, which analyzes this data and recognizes the user's emotional state. For example, a smile on the user's face indicates "happiness," while a frown indicates "dissatisfaction."

[0289] Step 6:

[0290] The server inputs the extracted keywords and recognized emotion data into a generative AI model. The generative AI generates native ads based on the keywords while adjusting the tone and content of the ads based on the emotion data. For example, if the user is in a "happy" state, it generates a positive-toned ad for a "Kyoto Autumn Foliage Tour Special Campaign."

[0291] Step 7:

[0292] The server adjusts the layout and style of the generated ad to match the design of the web page, for example by using CSS to adjust the font size, color, and margins of the ad so that it blends seamlessly into the web page.

[0293] Step 8:

[0294] The server generates the updated HTML content and inserts the ad in the appropriate position (e.g. after the third paragraph of the article).

[0295] Step 9:

[0296] The server sends the updated web page to the device, which reads the new HTML data and redraws the web page.

[0297] Step 10:

[0298] When users browse the updated web page, they will notice the generated native ads and can continue reading the page content naturally. Users will receive ads that match their emotions, resulting in a more comfortable browsing experience.

[0299] Example 2

[0300] 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."

[0301] Current ad delivery systems struggle to provide effective ads that take into account the user's emotional state. Furthermore, the inconsistency between the content of a web page and the ads often leads to a poor user experience. Therefore, a system is needed that can generate and display ads in real time that are optimized for the user's emotions and the content of a web page.

[0302] 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.

[0303] In this invention, the server includes means for acquiring web page content through user web access, means for analyzing the web page content in real time and extracting key topics and keywords, means for collecting data such as the user's facial expressions, voice tone, and text input and recognizing emotions, means for generating a native advertisement using a generation AI based on the extracted topics and keywords and the recognized emotion data, means for adjusting the layout and style of the generated native advertisement to match the design of the web page, means for inserting the generated native advertisement at an appropriate position on the web page, and means for transmitting the updated web page to a terminal and redisplaying it. This makes it possible to generate and display an advertisement optimized for the user's emotional state in real time.

[0304] "Web access" refers to a user accessing a specific web page via the Internet.

[0305] "Web page content" refers to all data elements contained within a Web page, including text, images, style sheets, scripts, etc.

[0306] "Real-time" refers to data processing and analysis occurring almost immediately.

[0307] "Topic" refers to the main theme that describes the content or subject matter of a Web page.

[0308] "Keywords" refer to important words or phrases that represent the content of a web page.

[0309] "Facial expression" refers to facial changes and characteristics that represent the user's emotional state.

[0310] "Voice tone" refers to the emotional inflections and nuances contained in a user's voice.

[0311] "Text input" refers to text data that a user inputs on a web page.

[0312] "Emotion recognition" refers to analyzing and identifying a user's emotional state from their facial expressions, voice, and text.

[0313] "Generative AI" refers to algorithms or models that use artificial intelligence techniques to generate new content (e.g., advertisements) based on specific input data.

[0314] "Native advertising" refers to advertising that blends naturally into the content of a web page.

[0315] "Layout" refers to the presentation and structure of a web page or advertisement.

[0316] "Style" refers to the design elements of a web page or advertisement (e.g., fonts, colors, spacing).

[0317] A "server" refers to a computer system that provides information and services over a network.

[0318] "Terminal" refers to a device, such as a computer or smartphone, that a user uses to access the Internet.

[0319] "Redisplay" refers to displaying the updated web page on the screen again.

[0320] This invention relates to a system that recognizes user emotions and automatically inserts native ads into web pages based on those emotions. The system mainly consists of four main components: a user device, a server, a generative AI model, and an emotion engine.

[0321] System Configuration and Hardware / Software Usage

[0322] 1. User's device

[0323] A device that allows users to access the web. The terminal has the function of sending requests to a web server and displaying the retrieved content. It also has the function of collecting data such as the user's facial expressions, voice tone, and text input and sending it to the server. This device is equipped with input devices such as a camera and microphone.

[0324] 2. Server

[0325] The server plays a central role in retrieving and analyzing the HTML content of web pages. The server retrieves HTML data using web server software such as Apache (registered trademark) or Nginx, and analyzes the content using an HTML parser library such as BeautifulSoup. It also has the function of analyzing user emotion data and generating advertisements using generative AI based on the emotion. It also has the function of generating HTTP responses to send updated web pages to the user's device.

[0326] 3. Generative AI Models

[0327] The generative AI model runs on the server and generates text ads based on the analyzed topics and keywords. This model, such as OpenAI's GPT-3, generates appropriate ads by inputting a prompt. An example of a prompt is "Please create an ad for an autumn foliage tour with a happy tone."

[0328] 4. Emotion Engine

[0329] The emotion engine analyzes the user's facial expressions, voice tone, and text input to recognize the user's emotional state in real time. The collected emotion data is sent to a server and input into a generative AI model. This emotion engine may utilize, for example, Microsoft's Azure Emotion API or Google's Cloud Vision API.

[0330] Specific examples

[0331] As a concrete example, consider the case where a user accesses a blog page with the theme "Autumn Kyoto Sightseeing." The device collects the user's facial expressions and voice tone and sends them to the emotion engine. As a result, if the user is smiling, it is determined to be "happy." This emotional state, along with the topic and keywords of the blog page (e.g., "autumn leaves," "famous places," and "sightseeing"), are input into a generative AI model to generate advertising text with a positive tone, such as "Special feature on Kyoto autumn leaves tours."

[0332] An example of a prompt for the generative AI model is "Please create an advertisement for an autumn foliage tour in a happy tone." The server adjusts the layout and style of the generated advertisement using a CSS stylesheet to blend in with the existing design of the webpage, and inserts it in the appropriate position after the third paragraph.

[0333] The server then sends the updated HTML content to the user's device, which quickly redraws it, allowing the user to smoothly view the updated web page.

[0334] This invention makes it possible to generate and display advertisements optimized for the user's emotional state in real time, thereby maximizing advertising effectiveness and providing a superior user experience.

[0335] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0336] Step 1:

[0337] A user visits a web page. The user sends an HTTP request for the URL of a particular web page (e.g., "Autumn Kyoto Sightseeing"). The input is the URL of the web page, and the output is that the request is sent to the server.

[0338] Step 2:

[0339] The server retrieves the HTML data. The server analyzes the received HTTP request and retrieves the HTML data for the corresponding web page from a database or file system. The input is the URL of the web page, and the specific operation is for web server software such as Apache or Nginx to load the HTML file. The output is the HTML data for the web page.

[0340] Step 3:

[0341] The server parses the HTML data. It uses an HTML parser library such as BeautifulSoup to analyze the structure of the web page and extract the text content. The input is the HTML data of the web page, and the specific operation is to parse the HTML structure. The output is the parsed text content.

[0342] Step 4:

[0343] The server extracts key topics and keywords. The server uses natural language processing (NLP) techniques, specifically SpaCy or NLTK, to extract keywords from the text. The input is the parsed text content, and the specific operation is NLP analysis. The output is key topics and keywords (e.g., "autumn leaves," "famous places," "sightseeing").

[0344] Step 5:

[0345] Collecting user emotional data. The device uses the built-in camera and microphone to collect data such as the user's facial expressions, voice tone, and text input. The input is the user's facial expressions and voice, and the output is the collected emotional data.

[0346] Step 6:

[0347] The emotion engine analyzes the emotional data. The emotion engine analyzes the collected data in real time and identifies the user's emotional state. Specifically, it uses Microsoft's Azure Emotion API or Google's Cloud Vision API. The input is emotional data, and the specific action is analysis via the API. The output is the user's emotional state (e.g., "happy," "unhappy").

[0348] Step 7:

[0349] The server generates advertisements using a generative AI model. The server inputs the extracted keywords and recognized emotion data into the generative AI model (e.g., OpenAI's GPT-3). The advertisement is generated using a prompt (e.g., "Please create an advertisement for an autumn foliage tour in a happy tone"). The input is keywords and emotion data, and the prompt is input into the generative AI model as a specific action. The output is the generated advertisement text.

[0350] Step 8:

[0351] The server adjusts the layout and style of the ad. It adjusts the layout and style using a CSS stylesheet so that the generated ad matches the design of the webpage. The input is the generated ad text, and the specific action is to apply the CSS stylesheet. The output is the adjusted ad layout.

[0352] Step 9:

[0353] The server inserts the advertisement into the web page. The generated advertisement is inserted into the updated HTML content at the appropriate position. For example, an advertisement is added after the third paragraph of a blog post. The input is the adjusted advertisement layout and HTML content, and the specific operation is to reconstruct the HTML document. The output is the updated HTML content.

[0354] Step 10:

[0355] The server sends the updated web page to the device. The device receives the updated content and redraws the web page. The input is the updated HTML content, which is sent to the device as an HTTP response as a specific action. The output is the updated web page displayed on the device.

[0356] This allows the user to view advertisements that match their emotions seamlessly along with the displayed web page.

[0357] (Application example 2)

[0358] 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."

[0359] Current virtual stores display uniform advertisements without considering the user's emotional state, resulting in a poor user experience and limited advertising effectiveness. Furthermore, it is difficult to accurately suggest information and products that users desire in real time, and there is a lack of technology to seamlessly provide advertisements tailored to individual users. This results in lower ad click rates and lower purchasing intent, preventing advertisers from implementing effective marketing strategies.

[0360] 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.

[0361] In this invention, the server includes means for acquiring virtual space content through user access, means for analyzing the virtual space content in real time and extracting main themes and related words, means for recognizing emotions from the user's facial expressions and voice, means for generating native advertisements using a generation AI based on the extracted themes and related words and the recognized emotions, means for inserting the generated native advertisements into appropriate positions in the virtual space, and means for transmitting the updated virtual space to the terminal and redisplaying it, thereby making it possible to provide advertisements optimized for the user's emotional state in real time.

[0362] "User" refers to a person who uses a virtual space or system.

[0363] "Access" refers to the action of a user connecting to a specific virtual space or web page.

[0364] "Virtual space" refers to a space that is generated using digital technology and that can be visually experienced but is different from a real location.

[0365] "Content" refers to information such as text, images, video, and audio displayed in virtual spaces and on web pages.

[0366] "Real-time analysis" refers to processing data immediately in response to user operations or access.

[0367] "Major subject" refers to a topic that is considered particularly important within a document or piece of content.

[0368] "Related words" refer to words that are related to the main topic and provide additional information.

[0369] "Facial expression" refers to an element used to infer emotions by analyzing the user's facial movements and features.

[0370] "Voice" refers to the element used to analyze the user's voice to read their emotions and intentions.

[0371] "Emotion recognition" refers to the process of analyzing and identifying a person's emotional state from their facial expressions and voice.

[0372] "Generative AI" refers to artificial intelligence that uses machine learning models to automatically generate content and advertisements.

[0373] "Native advertising" refers to advertising that blends naturally into the format and design of the medium in which it is displayed.

[0374] "Appropriate location" refers to a location where the advertisement is displayed effectively and without interfering with the user's experience.

[0375] "Terminal" refers to a hardware device that allows a user to access a system or virtual space.

[0376] This invention relates to a system that recognizes user emotions in real time and displays appropriate native advertisements in a virtual store based on those emotions. This system mainly consists of four main components: a user device, a server, a generative AI model, and an emotion engine.

[0377] System Configuration

[0378] User's device

[0379] Users access the virtual store using devices such as smartphones and head-mounted displays (HMDs). These devices have the ability to collect the user's facial expressions and voice data and send it to a server.

[0380] server

[0381] The server retrieves the virtual store's content and performs real-time analysis. It also extracts key themes and related terms and analyzes user sentiment data. It then uses a generative AI model to generate native ads based on the extracted data and inserts them into the virtual store at appropriate locations. Finally, it sends the updated virtual store data to the user's device.

[0382] Generative AI Models

[0383] The generative AI model generates text ads based on the extracted subject and related words, as well as user sentiment data. This model uses advanced language generation techniques such as GPT-4. An example prompt is as follows:

[0384] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[0385] Emotion Engine

[0386] The emotion engine analyzes the user's facial expressions and voice to recognize emotions in real time. It uses emotion analysis technologies such as the Emotion SDK to identify the user's emotional state.

[0387] Processing flow

[0388] When a user accesses a virtual store using a smartphone or HMD, the device captures facial expression and voice data and sends them to the server, which then retrieves the content of the virtual store and analyzes it using a natural language processing library (e.g., SpaCy or NLTK) to extract key themes and related terms.

[0389] In parallel, the emotion engine recognizes the user's emotions and sends the data to the server, which then inputs the extracted subject and related words, as well as the emotion data, into a generative AI model to generate native ads. The generated ads are then adjusted to fit the design of the virtual store and inserted at the appropriate locations.

[0390] The updated virtual store data is sent from the server to the user's device and instantly redisplayed, allowing the user to seamlessly browse the virtual store and experience emotionally optimized advertising.

[0391] As a specific example, if a user accesses a virtual store on their smartphone and smiles while browsing the "latest gadgets" section, the system will recognize the user's emotion as "happiness" and send the following prompt to the generative AI model:

[0392] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[0393] The generated ad text is optimized to match the user's emotional state, such as "We're having a special sale on the latest gadgets right now! Get the latest technology that will make you smile at a great price!" This improves the user experience and maximizes the effectiveness of the ad.

[0394] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0395] Step 1:

[0396] A user accesses a virtual store using a device (smartphone or HMD). The device sends an access request to the server, requesting content from the virtual store. The input is the user's access request, and the output is a request to the server.

[0397] Step 2:

[0398] The server receives the access request and retrieves the content of the virtual store from a database or file system. This content can include text, images, videos, etc. The input is the request, and the output is the content of the virtual store.

[0399] Step 3:

[0400] The device captures the user's facial expressions and voice data and transmits this data to the server in real time. The input is the user's facial expressions and voice data, and the output is data transmission to the server. Specifically, data is collected from the smartphone camera and microphone and transmitted in real time.

[0401] Step 4:

[0402] The server analyzes the received content in real time using natural language processing libraries (e.g., SpaCy or NLTK) to extract key themes and related terms from the text data. The input is the content of the virtual store, and the output is the extracted themes and related terms.

[0403] Step 5:

[0404] The server uses an emotion engine to analyze the transmitted facial expressions and voice data and recognize the user's emotions. It uses emotion analysis technologies such as Emotion SDK. The input is the user's facial expressions and voice data, and the output is the recognized emotional state.

[0405] Step 6:

[0406] The server inputs the extracted topic and related words, as well as the recognized emotion data, into a generative AI model. Specifically, the generative AI model (such as GPT-4) generates a text ad based on the prompt sentence. The input is the topic, related words, and emotion data, and the output is the generated ad text. An example of a prompt sentence is as follows:

[0407] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[0408] Step 7:

[0409] The server adjusts the layout and style of the generated advertisement to match the design of the virtual store, including font size, color, margins, etc. using a CSS stylesheet. The input is the generated advertisement text, and the output is the adjusted advertisement.

[0410] Step 8:

[0411] The server inserts the adjusted advertisement into the appropriate location in the virtual store, for example, at the edge of the display where the user is looking or next to a product shelf. The input is the adjusted advertisement, and the output is the updated virtual store content.

[0412] Step 9:

[0413] The server sends the updated virtual store data to the user's terminal. The terminal receives this data and immediately redraws it. The input is the updated virtual store content, and the output is the virtual store redisplayed on the terminal.

[0414] This series of processes allows users to have a seamless shopping experience in a virtual store and view advertisements that are optimal for their emotions.

[0415] 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.

[0416] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (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.

[0417] 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.

[0418] [Second embodiment]

[0419] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0420] 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.

[0421] 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).

[0422] 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.

[0423] 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.

[0424] 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).

[0425] 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.

[0426] 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.

[0427] 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.

[0428] 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.

[0429] 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.

[0430] 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."

[0431] The present invention relates to a real-time native advertising generation system using generation AI. Specifically, the system analyzes the content of web pages accessed by users, dynamically generates advertisements that match that content, and inserts them into the web pages, thereby enhancing advertising effectiveness without causing discomfort to users.

[0432] Specific Embodiments of the System

[0433] System Configuration

[0434] This system mainly consists of three main components: the user's device, the server, and the generative AI model.

[0435] 1. User's device

[0436] A device that allows users to access the web. The device has the ability to send requests to web servers and display the retrieved content. Devices include PCs, smartphones, tablets, etc.

[0437] 2. Server

[0438] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also generates advertisements using AI and processes them to insert them into web pages. It also has the function of sending updated web pages to users' devices.

[0439] 3. Generative AI Models

[0440] The generative AI model runs on the server and generates text ads based on the analyzed topics and keywords. The generative AI model uses natural language processing (NLP) techniques.

[0441] Program processing

[0442] When a user accesses a web page, the following sequence of events occurs:

[0443] 1. User Web Access

[0444] A user accesses a specific web page. For example, the user clicks on a blog page with the theme "Autumn Kyoto Sightseeing."

[0445] The terminal sends a request to the server requesting the HTML content of the web page.

[0446] 2. Getting HTML data

[0447] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[0448] 3. Content Analysis

[0449] The server analyzes the HTML content and extracts key topics and keywords. For example, the analysis extracts keywords such as "autumn leaves," "famous places," and "guided tours."

[0450] 4. Ad generation

[0451] The server inputs the extracted keywords into a generative AI model to generate ad text related to the content, such as an ad for a "Kyoto Autumn Foliage Tour Special."

[0452] 5. Ad Styling

[0453] The server adjusts the style and layout of the generated advertising text to match the design of the web page.

[0454] 6. Ad Insertion

[0455] The server then regenerates the HTML for the web page, inserting the ad in the appropriate place, for example after the second paragraph.

[0456] 7. Submitting Updated Content

[0457] The server sends the updated HTML content to the user's device, which redisplays it.

[0458] Specific examples

[0459] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following happens:

[0460] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[0461] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[0462] The server inputs these keywords into the generation AI, which generates advertising text such as "Special feature on Kyoto autumn foliage tours."

[0463] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[0464] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[0465] In this way, the system realizes native advertising that does not cause discomfort to users, maximizes advertising effectiveness, and improves the user experience.

[0466] The processing flow will be explained below.

[0467] Step 1:

[0468] A user accesses a web page of interest. For example, a user accesses a blog post about "Autumn Kyoto Sightseeing." The device sends a request to that URL, requesting the HTML content of the web page from the server.

[0469] Step 2:

[0470] The server receives the user's request and retrieves the appropriate HTML content from a database or file system. The web server then prepares a response containing all the data for the web page (text, images, style sheets, etc.).

[0471] Step 3:

[0472] The server analyzes the HTML retrieved. Specifically, it uses an HTML parser to analyze the structure of the web page and extract the text content. For example, it extracts the entire body of an article titled "Sightseeing in Kyoto in Autumn."

[0473] Step 4:

[0474] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the extracted text. For example, keywords such as "autumn leaves," "famous places," and "guided tours" might be found. This process may involve the use of NLP libraries or APIs (e.g., SpaCy or NLTK).

[0475] Step 5:

[0476] The server initializes the generative AI model and provides the extracted keywords as input data. The generative AI model (e.g., GPT) is used to generate text for native ads based on the keywords. For example, an ad copy such as "Special feature on Kyoto autumn foliage tours" is generated. Here, a pre-trained model is used for the generative AI.

[0477] Step 6:

[0478] The server adjusts the layout and style of the generated advertisement to match the design of the web page, specifically by applying a CSS stylesheet so that the advertisement blends into the existing design of the web page. This adjustment may include font size, color, margins, etc.

[0479] Step 7:

[0480] The server updates the HTML content, inserting the generated ad in the appropriate place (e.g., adding an ad after the third paragraph of a blog post), and restructures the entire HTML document to ensure consistency.

[0481] Step 8:

[0482] The server sends the updated HTML content to the device, and the device reads the new HTML data received from the server and redraws the web page.

[0483] Step 9:

[0484] Users can browse the updated web page and continue reading the original content without noticing the generated native ad, making the user experience natural and seamless.

[0485] Example 1

[0486] 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."

[0487] Conventional ad delivery methods often lack a proper blend of web page content and ads, creating a sense of discomfort for users. Furthermore, to maximize advertising effectiveness, it is necessary to generate and display ads in real time that are closely related to the content being viewed by users. However, the technology to accomplish this efficiently has not yet been fully developed.

[0488] 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.

[0489] In this invention, the server includes means for acquiring web page information from user web access, means for analyzing the web page information in real time and extracting main topics and keywords, means for generating advertisements using a generative AI model based on the extracted topics and keywords, means for adjusting the layout and style of the generated advertisement to match the design of the web page, means for inserting the generated advertisement into an appropriate position on the web page, and means for transmitting the updated web page to the terminal and redisplaying it, thereby enabling advertisements to be displayed in a way that does not cause discomfort to the user and maximizing advertising effectiveness.

[0490] "Web access" refers to the act of a user connecting to a web page via the Internet and obtaining information.

[0491] "Web page information" is a collection of data such as text, images, videos, and links displayed on a website.

[0492] "Real-time" refers to processing occurring immediately without delay.

[0493] "Analysis" refers to examining the content of data in detail, especially using natural language processing to extract specific information.

[0494] A "topic" refers to the main topic or subject matter in text data.

[0495] "Keywords" are important words or phrases in a document that help identify the topic.

[0496] A "generative AI model" refers to an algorithm or system that uses artificial intelligence to generate new content or data.

[0497] "Advertising" is information used to promote products and services to consumers.

[0498] "Layout" refers to the position and structure of each element arranged within a web page.

[0499] "Style" refers to the appearance and design of text and images on a web page, including color, font, size, etc.

[0500] "Insertion" means adding new information or elements into an existing Web page.

[0501] A "terminal" is a device that a user uses to access the Internet. Examples include personal computers, smartphones, and tablets.

[0502] "Redisplay" means displaying updated content or data again on the user's terminal.

[0503] This invention relates to a real-time native advertising generation system using generative AI. Specifically, it provides a system that analyzes the content of web pages accessed by users, dynamically generates advertisements that fit the content, and inserts them into the web pages, thereby enhancing advertising effectiveness without causing discomfort to users.

[0504] System Configuration

[0505] The system mainly consists of three main components: the user's device, the server, and the generative AI model.

[0506] 1. User's device

[0507] A device that allows users to access the web. The device has the ability to send requests to web servers and display the retrieved content. Devices include PCs, smartphones, tablets, etc.

[0508] 2. Server

[0509] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also generates advertisements using AI and processes them to insert them into web pages. It also has the function of sending updated web pages to users' devices.

[0510] 3. Generative AI Models

[0511] The generative AI model runs on a server and generates text ads based on the analyzed topics and keywords. Natural language processing (NLP) techniques are used for generative AI models. Specific techniques include generative models such as GPT-3.

[0512] Server Processing

[0513] When a user accesses a specific web page, the device requests the HTML content of that web page from the server. The server receives the request and retrieves the relevant HTML data. The server then analyzes the retrieved HTML content to extract key topics and keywords. For example, NLTK and spaCy are used to extract keywords such as "autumn leaves," "famous places," and "guided tours."

[0514] Next, the server inputs the extracted keywords into a generative AI model to generate advertising text related to the content. For example, an advertisement such as "Special feature on Kyoto autumn foliage tours" may be generated. An example of a prompt used in this case is "Please generate advertising text on the themes of 'Kyoto autumn foliage,' 'famous places,' and 'sightseeing.'"

[0515] The server then adjusts the generated ad text to match the style and layout of the webpage, using CSS and JavaScript to make the ad look consistent with the webpage. The server then inserts the ad in the appropriate position and regenerates the webpage's HTML, for example, inserting the ad after the second paragraph.

[0516] Finally, the server sends the updated HTML content to the user's device, which redisplays it, allowing the user to seamlessly browse the page.

[0517] Specific examples

[0518] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following will happen:

[0519] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[0520] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[0521] The server inputs these keywords into the generation AI, which generates advertising text such as "Special feature on Kyoto autumn foliage tours."

[0522] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[0523] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[0524] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0525] Step 1:

[0526] A user accesses a specific web page. Specifically, the user enters a URL in the browser's address bar or clicks a link. The URL is entered as input, and the device sends an HTTP request to the server requesting the HTML content corresponding to that URL. The server receives this request as output.

[0527] Step 2:

[0528] The server receives a user request and retrieves the relevant HTML content. Specifically, the server loads the HTML file corresponding to the requested URL from a database or file system. The input is the requested URL, and the output is the corresponding HTML content.

[0529] Step 3:

[0530] The server analyzes the retrieved HTML content and extracts key topics and keywords. Specifically, it uses a natural language processing (NLP) library to perform text analysis. For example, it uses "NLTK" or "spaCy." The input is the entire HTML content, and the output is extracted keywords such as "autumn leaves," "famous places," and "guided tours."

[0531] Step 4:

[0532] The server inputs the extracted keywords into the generative AI model and generates advertising text related to the content. Specifically, the server sends a prompt to the generative AI model. An example of this prompt is "Please generate advertising text on the themes of 'Kyoto autumn leaves,' 'famous places,' and 'sightseeing.'" The input is the extracted keywords and the prompt, and the output is advertising text such as "Special feature on Kyoto autumn leaves tours."

[0533] Step 5:

[0534] The server adjusts the style and layout of the generated ad text to match the web page design. Specifically, it uses CSS and JavaScript to adjust the font, color, size, etc. of the ad text to match the web page design. The input is the generated ad text, and the output is the adjusted ad text.

[0535] Step 6:

[0536] The server inserts the ad in the appropriate position and regenerates the HTML for the webpage, using JavaScript DOM manipulation and a server-side template engine. For example, inserting the ad after the second paragraph. The input is the adjusted ad text and the original HTML content, and the output is the updated HTML content with the ad inserted.

[0537] Step 7:

[0538] The server sends the updated HTML content to the user's device, which then redisplays it. Specifically, the newly generated HTML is returned to the device as an HTTP response. The input is the updated HTML content with the ad inserted, and the output is the updated web page displayed on the device. This allows the user to view the page seamlessly.

[0539] (Application example 1)

[0540] 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."

[0541] Modern internet advertising tends to be intrusive to users, often detracting from the user experience. Furthermore, static advertisements tend to be less relevant to the page content, reducing advertising effectiveness. Conventional advertising systems are unable to dynamically generate advertisements that match the content of web pages, which can be unsettling for users. The present invention aims to solve these problems.

[0542] 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.

[0543] In this invention, the server includes means for acquiring web page content through user web access, means for analyzing the web page content in real time and extracting key topics and keywords, and means for generating native advertisements using a generation AI based on the extracted topics and keywords. This makes it possible to improve the user experience while increasing advertising effectiveness by inserting dynamically generated advertisements in appropriate positions based on the content viewed by the user.

[0544] "User web access" refers to the act of a user accessing a web page via the Internet and obtaining or viewing information.

[0545] "Web page content" refers collectively to text, images, video, and other media or information displayed on a web page.

[0546] "Real-time analysis" is the process of analyzing the content of a web page immediately after a user accesses it.

[0547] "Major topics and keywords" refers to the important subjects and keywords in the content of a web page, which serve as the basis for generating advertisements.

[0548] "Generative AI" is a system that uses artificial intelligence techniques to generate new text based on specific prompts.

[0549] "Native advertising" is advertising that is displayed in a way that is integrated with the content of a web page and is designed to not cause any discomfort to the user.

[0550] A "dynamically generated advertisement based on content" is an advertisement that is automatically generated in real time according to the content of the web page that the user is viewing.

[0551] "Inserting in an appropriate position on a web page" refers to placing advertisements in appropriate locations on a web page so as not to impair the user experience.

[0552] "User experience" refers to the overall experience and satisfaction a user has when using a product or service.

[0553] An "application" is a software program that runs on a device and provides a particular function or service.

[0554] This invention is a system that uses generative AI to generate native ads in real time and insert them into the content of web pages accessed by users. This article mainly explains an example of applying this system to a news application for smartphones.

[0555] System Configuration

[0556] The system consists of three main components: the user's device, a server, and a generative AI model.

[0557] 1. User's device

[0558] This includes devices such as smartphones and tablets.

[0559] It provides an interface for users to access articles through a news app.

[0560] 2. Server

[0561] It plays a central role in retrieving and parsing the HTML content of web pages.

[0562] Generative AI is used to generate advertisements, and processing is performed to properly insert the advertisements into web pages.

[0563] 3. Generative AI Models

[0564] Natural language processing (NLP) techniques are used to generate text ads based on analyzed topics and keywords.

[0565] Generative AI models used include, for example, OpenAI's GPT-4.

[0566] Program processing overview

[0567] When a user accesses a web page, the server performs the following process.

[0568] 1. User Web Access

[0569] A user clicks on a particular article in a news app, and the device requests the HTML content of that news article from the server.

[0570] 2. Getting HTML data

[0571] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[0572] 3. Content Analysis

[0573] The server analyzes the HTML content and extracts key topics and keywords, such as "climate change," "renewable energy," and "eco-activities."

[0574] 4. Ad generation

[0575] The server inputs the extracted keywords into a generative AI model to generate advertising text related to the content. An example of a prompt sentence to be used is as follows:

[0576] Generate ad text based on the following keywords: climate change, renewable energy, eco-activism

[0577] 5. Ad Styling

[0578] The server adjusts the style and layout of the generated advertising text to match the design of the news article.

[0579] 6. Ad Insertion

[0580] The server inserts the advertisement at the appropriate position in the news article, for example after the second paragraph of the article.

[0581] 7. Submitting Updated Content

[0582] The server sends the updated HTML content to the user's device, which redisplays it.

[0583] Specific examples

[0584] For example, if a user accesses an article on the topic of "Thinking about Environmental Issues" in a news app, the following steps are taken:

[0585] The user clicks on the URL of a news article, and the device requests the HTML content from the server.

[0586] The server retrieves the HTML content, analyzes it, and extracts the keywords "climate change," "renewable energy," and "eco-activities."

[0587] The server inputs these keywords into a generation AI, which generates ad text such as "Introducing a new electric vehicle."

[0588] The server adapts the generated advertisements to the style of the news article and inserts them at the appropriate locations.

[0589] The updated page will be redisplayed on the user's device, allowing the user to continue reading the article without any discomfort.

[0590] This invention makes it possible to maximize advertising effectiveness by dynamically generating and inserting advertisements that match the content without impairing the user experience.

[0591] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0592] Step 1:

[0593] A user clicks on a specific article in a news app.

[0594] Input: The URL of the news article selected by the user.

[0595] Specific operation: The device sends a request to the server for the HTML content of the news article.

[0596] Output: The server receives the request.

[0597] Step 2:

[0598] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[0599] Input: A news article URL request.

[0600] Specific operation: The server searches for and retrieves the corresponding HTML content from the database.

[0601] Output: The retrieved HTML content.

[0602] Step 3:

[0603] The server parses the retrieved HTML content and extracts key topics and keywords.

[0604] Input: The retrieved HTML content.

[0605] What it does: The server uses natural language processing (NLP) techniques to parse the HTML content and extract key topics and keywords (e.g., "climate change," "renewable energy," "green activities").

[0606] Output: Extracted main topics and keywords.

[0607] Step 4:

[0608] The server inputs the extracted keywords into a generative AI model to generate advertising text relevant to the content.

[0609] Input: Extracted main topics and keywords.

[0610] Specific operation: The server sends the following prompt to the generated AI model:

[0611] Generate ad text based on the following keywords: climate change, renewable energy, eco-activism

[0612] A generative AI model generates ad text.

[0613] Output: The generated ad text.

[0614] Step 5:

[0615] The server adjusts the style and layout of the generated advertising text to match the design of the news article.

[0616] Input: The generated ad text.

[0617] What it does: The server uses CSS and HTML templates to adjust the ad design and integrate it into the style of the news article.

[0618] Output: The adjusted ad text.

[0619] Step 6:

[0620] The server inserts the advertisements into the news articles at appropriate locations.

[0621] Input: Your adjusted ad text.

[0622] Specific behavior: The server regenerates the HTML content of the news article, for example inserting an advertisement after the second paragraph of the article.

[0623] Output: The updated HTML content.

[0624] Step 7:

[0625] The server sends the updated HTML content to the user's device, and the user redisplays it through the news app.

[0626] Input: The updated HTML content.

[0627] What happens: The server sends HTML content to the device, and the news app receives and displays it.

[0628] Output: The user can view the updated news article.

[0629] These steps ensure that users are presented with ads that are tailored to their content and that fit their content seamlessly, improving the user experience.

[0630] 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.

[0631] This invention relates to a system that recognizes user emotions and automatically inserts native ads generated based on those emotions into web pages. Specifically, the system analyzes the content of web pages retrieved by users' web access in real time and extracts key topics and keywords. It also uses an emotion engine to recognize user emotions and generates native ads based on this emotion using a generation AI. The generated native ads are inserted in appropriate positions on the web page, and the updated web page is sent to the terminal. This system provides ads with optimal timing and content to users, achieving higher advertising effectiveness.

[0632] Specific Embodiments of the System

[0633] System Configuration

[0634] This system mainly consists of four main components: the user's device, the server, the generative AI model, and the emotion engine.

[0635] 1. User's device

[0636] A device that allows users to access the web. The terminal has the function of sending requests to a web server and displaying the retrieved content. It also has the function of collecting data such as the user's facial expressions, voice tone, and text input, and sending it to the server.

[0637] 2. Server

[0638] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also analyzes user emotion data, generates advertisements based on the emotions using generative AI, and processes the advertisements to be inserted into web pages. It also has the function of sending updated web pages to users' devices.

[0639] 3. Generative AI Models

[0640] The generative AI model runs on a server and generates text ads based on the analyzed topics and keywords, and also takes into account user sentiment data to generate ads with the appropriate tone and content.

[0641] 4. Emotion Engine

[0642] The emotion engine analyzes users' facial expressions, voice tone, and text input to recognize their emotional state in real time. This emotional data is fed into a generative AI model and used to generate ads.

[0643] Program processing

[0644] When a user accesses a web page, the following sequence of events occurs:

[0645] 1. User Web Access

[0646] When a user accesses a specific web page, for example, a blog post about "Autumn Kyoto Sightseeing," the device sends a request to the URL, requesting the HTML content of the web page from the server.

[0647] 2. Getting HTML data

[0648] The server receives the user's request and retrieves the relevant HTML data from a database or file system. The web server then prepares a response containing all the data for the web page (text, images, style sheets, etc.).

[0649] 3. Content Analysis

[0650] The server analyzes the HTML retrieved. Specifically, it uses an HTML parser to analyze the structure of the web page and extract the text content. For example, it extracts the entire body of an article titled "Sightseeing in Kyoto in Autumn."

[0651] 4. Topic and Keyword Extraction

[0652] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the extracted text. For example, keywords such as "autumn leaves," "famous places," and "guided tours" might be found. This process may involve the use of NLP libraries or APIs (e.g., SpaCy or NLTK).

[0653] 5. Emotional Recognition

[0654] While a user is browsing a web page, the device collects data such as facial expressions and voice and sends it to the emotion engine. The emotion engine analyzes this data and identifies the user's emotional state. For example, if the user is smiling, it is judged to be "happy," and if they are frowning, it is judged to be "unhappy."

[0655] 6. Ad Generation

[0656] The server inputs the extracted keywords and recognized emotion data into a generative AI model to generate native ads that match the emotion. For example, if the user is in a "happy" state, it generates an ad with a positive tone, and if the user is in a "dissatisfied" state, it generates a problem-solving ad.

[0657] 7. Ad Styling

[0658] The server adjusts the layout and style of the generated advertisement to match the design of the web page, specifically by applying a CSS stylesheet so that the advertisement blends into the existing design of the web page. This adjustment may include font size, color, margins, etc.

[0659] 8. Ad Insertion

[0660] The server updates the HTML content, inserting the generated ad in the appropriate place (e.g., adding an ad after the third paragraph of a blog post), and restructures the entire HTML document to ensure consistency.

[0661] 9. Submitting Updated Content

[0662] The server sends the updated HTML content to the device, and the device reads the new HTML data received from the server and redraws the web page.

[0663] Specific examples

[0664] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following happens:

[0665] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[0666] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[0667] The server inputs these keywords into the AI ​​generator, which generates ad text such as "Kyoto Autumn Foliage Tour Special." It also generates ad text with a positive tone based on the user's emotional data (e.g., "happiness").

[0668] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[0669] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[0670] In this way, the system provides advertisements that best suit the user's emotional state, maximizing advertising effectiveness and improving the user experience.

[0671] The processing flow will be explained below.

[0672] Step 1:

[0673] A user clicks on a URL in their web browser to access a blog post about "Autumn Kyoto Sightseeing." The device sends a request to the server for that URL, requesting the HTML content of the web page.

[0674] Step 2:

[0675] The server receives the user's request and retrieves the appropriate HTML content from a database or file system.

[0676] Step 3:

[0677] The HTML content retrieved by the server is analyzed using an HTML parser, extracting each element of the web page, such as text content, images, and style sheets.

[0678] Step 4:

[0679] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the text content, such as "autumn leaves," "famous places," and "guided tours."

[0680] Step 5:

[0681] While a user is browsing a web page, the device uses a camera and microphone to transmit the user's facial expressions and voice tone to the emotion engine, which analyzes this data and recognizes the user's emotional state. For example, a smile on the user's face indicates "happiness," while a frown indicates "dissatisfaction."

[0682] Step 6:

[0683] The server inputs the extracted keywords and recognized emotion data into a generative AI model. The generative AI generates native ads based on the keywords while adjusting the tone and content of the ads based on the emotion data. For example, if the user is in a "happy" state, it generates a positive-toned ad for a "Kyoto Autumn Foliage Tour Special Campaign."

[0684] Step 7:

[0685] The server adjusts the layout and style of the generated ad to match the design of the web page, for example by using CSS to adjust the font size, color, and margins of the ad so that it blends seamlessly into the web page.

[0686] Step 8:

[0687] The server generates the updated HTML content and inserts the ad in the appropriate position (e.g. after the third paragraph of the article).

[0688] Step 9:

[0689] The server sends the updated web page to the device, which reads the new HTML data and redraws the web page.

[0690] Step 10:

[0691] When users browse the updated web page, they will notice the generated native ads and can continue reading the page content naturally. Users will receive ads that match their emotions, resulting in a more comfortable browsing experience.

[0692] Example 2

[0693] 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."

[0694] Current ad delivery systems struggle to provide effective ads that take into account the user's emotional state. Furthermore, the inconsistency between the content of a web page and the ads often leads to a poor user experience. Therefore, a system is needed that can generate and display ads in real time that are optimized for the user's emotions and the content of a web page.

[0695] 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.

[0696] In this invention, the server includes means for acquiring web page content through user web access, means for analyzing the web page content in real time and extracting key topics and keywords, means for collecting data such as the user's facial expressions, voice tone, and text input and recognizing emotions, means for generating a native advertisement using a generation AI based on the extracted topics and keywords and the recognized emotion data, means for adjusting the layout and style of the generated native advertisement to match the design of the web page, means for inserting the generated native advertisement at an appropriate position on the web page, and means for transmitting the updated web page to a terminal and redisplaying it. This makes it possible to generate and display an advertisement optimized for the user's emotional state in real time.

[0697] "Web access" refers to a user accessing a specific web page via the Internet.

[0698] "Web page content" refers to all data elements contained within a Web page, including text, images, style sheets, scripts, etc.

[0699] "Real-time" refers to data processing and analysis occurring almost immediately.

[0700] "Topic" refers to the main theme that describes the content or subject matter of a Web page.

[0701] "Keywords" refer to important words or phrases that represent the content of a web page.

[0702] "Facial expression" refers to facial changes and characteristics that represent the user's emotional state.

[0703] "Voice tone" refers to the emotional inflections and nuances contained in a user's voice.

[0704] "Text input" refers to text data that a user inputs on a web page.

[0705] "Emotion recognition" refers to analyzing and identifying a user's emotional state from their facial expressions, voice, and text.

[0706] "Generative AI" refers to algorithms or models that use artificial intelligence techniques to generate new content (e.g., advertisements) based on specific input data.

[0707] "Native advertising" refers to advertising that blends naturally into the content of a web page.

[0708] "Layout" refers to the presentation and structure of a web page or advertisement.

[0709] "Style" refers to the design elements of a web page or advertisement (e.g., fonts, colors, spacing).

[0710] A "server" refers to a computer system that provides information and services over a network.

[0711] "Terminal" refers to a device, such as a computer or smartphone, that a user uses to access the Internet.

[0712] "Redisplay" refers to displaying the updated web page on the screen again.

[0713] This invention relates to a system that recognizes user emotions and automatically inserts native ads into web pages based on those emotions. The system mainly consists of four main components: a user device, a server, a generative AI model, and an emotion engine.

[0714] System Configuration and Hardware / Software Usage

[0715] 1. User's device

[0716] A device that allows users to access the web. The terminal has the function of sending requests to a web server and displaying the retrieved content. It also has the function of collecting data such as the user's facial expressions, voice tone, and text input and sending it to the server. This device is equipped with input devices such as a camera and microphone.

[0717] 2. Server

[0718] The server plays a central role in retrieving and analyzing the HTML content of web pages. The server retrieves HTML data using web server software such as Apache or Nginx, and parses the content using an HTML parser library such as BeautifulSoup. It also has the ability to analyze user emotion data and generate advertisements using generative AI based on the emotion. It also has the ability to generate HTTP responses to send updated web pages to users' devices.

[0719] 3. Generative AI Models

[0720] The generative AI model runs on the server and generates text ads based on the analyzed topics and keywords. This model, such as OpenAI's GPT-3, generates appropriate ads by inputting a prompt. An example of a prompt is "Please create an ad for an autumn foliage tour with a happy tone."

[0721] 4. Emotion Engine

[0722] The emotion engine analyzes the user's facial expressions, voice tone, and text input to recognize the user's emotional state in real time. The collected emotion data is sent to a server and input into a generative AI model. This emotion engine may utilize, for example, Microsoft's Azure Emotion API or Google's Cloud Vision API.

[0723] Specific examples

[0724] As a concrete example, consider the case where a user accesses a blog page with the theme "Autumn Kyoto Sightseeing." The device collects the user's facial expressions and voice tone and sends them to the emotion engine. As a result, if the user is smiling, it is determined to be "happy." This emotional state, along with the topic and keywords of the blog page (e.g., "autumn leaves," "famous places," and "sightseeing"), are input into a generative AI model to generate advertising text with a positive tone, such as "Special feature on Kyoto autumn leaves tours."

[0725] An example of a prompt for the generative AI model is "Please create an advertisement for an autumn foliage tour in a happy tone." The server adjusts the layout and style of the generated advertisement using a CSS stylesheet to blend in with the existing design of the webpage, and inserts it in the appropriate position after the third paragraph.

[0726] The server then sends the updated HTML content to the user's device, which quickly redraws it, allowing the user to smoothly view the updated web page.

[0727] This invention makes it possible to generate and display advertisements optimized for the user's emotional state in real time, thereby maximizing advertising effectiveness and providing a superior user experience.

[0728] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0729] Step 1:

[0730] A user visits a web page. The user sends an HTTP request for the URL of a particular web page (e.g., "Autumn Kyoto Sightseeing"). The input is the URL of the web page, and the output is that the request is sent to the server.

[0731] Step 2:

[0732] The server retrieves the HTML data. The server analyzes the received HTTP request and retrieves the HTML data for the corresponding web page from a database or file system. The input is the URL of the web page, and the specific operation is for web server software such as Apache or Nginx to load the HTML file. The output is the HTML data for the web page.

[0733] Step 3:

[0734] The server parses the HTML data. It uses an HTML parser library such as BeautifulSoup to analyze the structure of the web page and extract the text content. The input is the HTML data of the web page, and the specific operation is to parse the HTML structure. The output is the parsed text content.

[0735] Step 4:

[0736] The server extracts key topics and keywords. The server uses natural language processing (NLP) techniques, specifically SpaCy or NLTK, to extract keywords from the text. The input is the parsed text content, and the specific operation is NLP analysis. The output is key topics and keywords (e.g., "autumn leaves," "famous places," "sightseeing").

[0737] Step 5:

[0738] Collecting user emotional data. The device uses the built-in camera and microphone to collect data such as the user's facial expressions, voice tone, and text input. The input is the user's facial expressions and voice, and the output is the collected emotional data.

[0739] Step 6:

[0740] The emotion engine analyzes the emotional data. The emotion engine analyzes the collected data in real time and identifies the user's emotional state. Specifically, it uses Microsoft's Azure Emotion API or Google's Cloud Vision API. The input is emotional data, and the specific action is analysis via the API. The output is the user's emotional state (e.g., "happy," "unhappy").

[0741] Step 7:

[0742] The server generates advertisements using a generative AI model. The server inputs the extracted keywords and recognized emotion data into the generative AI model (e.g., OpenAI's GPT-3). The advertisement is generated using a prompt (e.g., "Please create an advertisement for an autumn foliage tour in a happy tone"). The input is keywords and emotion data, and the prompt is input into the generative AI model as a specific action. The output is the generated advertisement text.

[0743] Step 8:

[0744] The server adjusts the layout and style of the ad. It adjusts the layout and style using a CSS stylesheet so that the generated ad matches the design of the webpage. The input is the generated ad text, and the specific action is to apply the CSS stylesheet. The output is the adjusted ad layout.

[0745] Step 9:

[0746] The server inserts the advertisement into the web page. The generated advertisement is inserted into the updated HTML content at the appropriate position. For example, an advertisement is added after the third paragraph of a blog post. The input is the adjusted advertisement layout and HTML content, and the specific operation is to reconstruct the HTML document. The output is the updated HTML content.

[0747] Step 10:

[0748] The server sends the updated web page to the device. The device receives the updated content and redraws the web page. The input is the updated HTML content, which is sent to the device as an HTTP response as a specific action. The output is the updated web page displayed on the device.

[0749] This allows the user to view advertisements that match their emotions seamlessly along with the displayed web page.

[0750] (Application example 2)

[0751] 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."

[0752] Current virtual stores display uniform advertisements without considering the user's emotional state, resulting in a poor user experience and limited advertising effectiveness. Furthermore, it is difficult to accurately suggest information and products that users desire in real time, and there is a lack of technology to seamlessly provide advertisements tailored to individual users. This results in lower ad click rates and lower purchasing intent, preventing advertisers from implementing effective marketing strategies.

[0753] 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.

[0754] In this invention, the server includes means for acquiring virtual space content through user access, means for analyzing the virtual space content in real time and extracting main themes and related words, means for recognizing emotions from the user's facial expressions and voice, means for generating native advertisements using a generation AI based on the extracted themes and related words and the recognized emotions, means for inserting the generated native advertisements into appropriate positions in the virtual space, and means for transmitting the updated virtual space to the terminal and redisplaying it, thereby making it possible to provide advertisements optimized for the user's emotional state in real time.

[0755] "User" refers to a person who uses a virtual space or system.

[0756] "Access" refers to the action of a user connecting to a specific virtual space or web page.

[0757] "Virtual space" refers to a space that is generated using digital technology and that can be visually experienced but is different from a real location.

[0758] "Content" refers to information such as text, images, video, and audio displayed in virtual spaces and on web pages.

[0759] "Real-time analysis" refers to processing data immediately in response to user operations or access.

[0760] "Major subject" refers to a topic that is considered particularly important within a document or piece of content.

[0761] "Related words" refer to words that are related to the main topic and provide additional information.

[0762] "Facial expression" refers to an element used to infer emotions by analyzing the user's facial movements and features.

[0763] "Voice" refers to the element used to analyze the user's voice to read their emotions and intentions.

[0764] "Emotion recognition" refers to the process of analyzing and identifying a person's emotional state from their facial expressions and voice.

[0765] "Generative AI" refers to artificial intelligence that uses machine learning models to automatically generate content and advertisements.

[0766] "Native advertising" refers to advertising that blends naturally into the format and design of the medium in which it is displayed.

[0767] "Appropriate location" refers to a location where the advertisement is displayed effectively and without interfering with the user's experience.

[0768] "Terminal" refers to a hardware device that allows a user to access a system or virtual space.

[0769] This invention relates to a system that recognizes user emotions in real time and displays appropriate native advertisements in a virtual store based on those emotions. This system mainly consists of four main components: a user device, a server, a generative AI model, and an emotion engine.

[0770] System Configuration

[0771] User's device

[0772] Users access the virtual store using devices such as smartphones and head-mounted displays (HMDs). These devices have the ability to collect the user's facial expressions and voice data and send it to a server.

[0773] server

[0774] The server retrieves the virtual store's content and performs real-time analysis. It also extracts key themes and related terms and analyzes user sentiment data. It then uses a generative AI model to generate native ads based on the extracted data and inserts them into the virtual store at appropriate locations. Finally, it sends the updated virtual store data to the user's device.

[0775] Generative AI Models

[0776] The generative AI model generates text ads based on the extracted subject and related words, as well as user sentiment data. This model uses advanced language generation techniques such as GPT-4. An example prompt is as follows:

[0777] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[0778] Emotion Engine

[0779] The emotion engine analyzes the user's facial expressions and voice to recognize emotions in real time. It uses emotion analysis technologies such as the Emotion SDK to identify the user's emotional state.

[0780] Processing flow

[0781] When a user accesses a virtual store using a smartphone or HMD, the device captures facial expression and voice data and sends them to the server, which then retrieves the content of the virtual store and analyzes it using a natural language processing library (e.g., SpaCy or NLTK) to extract key themes and related terms.

[0782] In parallel, the emotion engine recognizes the user's emotions and sends the data to the server, which then inputs the extracted subject and related words, as well as the emotion data, into a generative AI model to generate native ads. The generated ads are then adjusted to fit the design of the virtual store and inserted at the appropriate locations.

[0783] The updated virtual store data is sent from the server to the user's device and instantly redisplayed, allowing the user to seamlessly browse the virtual store and experience emotionally optimized advertising.

[0784] As a specific example, if a user accesses a virtual store on their smartphone and smiles while browsing the "latest gadgets" section, the system will recognize the user's emotion as "happiness" and send the following prompt to the generative AI model:

[0785] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[0786] The generated ad text is optimized to match the user's emotional state, such as "We're having a special sale on the latest gadgets right now! Get the latest technology that will make you smile at a great price!" This improves the user experience and maximizes the effectiveness of the ad.

[0787] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0788] Step 1:

[0789] A user accesses a virtual store using a device (smartphone or HMD). The device sends an access request to the server, requesting content from the virtual store. The input is the user's access request, and the output is a request to the server.

[0790] Step 2:

[0791] The server receives the access request and retrieves the content of the virtual store from a database or file system. This content can include text, images, videos, etc. The input is the request, and the output is the content of the virtual store.

[0792] Step 3:

[0793] The device captures the user's facial expressions and voice data and transmits this data to the server in real time. The input is the user's facial expressions and voice data, and the output is data transmission to the server. Specifically, data is collected from the smartphone camera and microphone and transmitted in real time.

[0794] Step 4:

[0795] The server analyzes the received content in real time using natural language processing libraries (e.g., SpaCy or NLTK) to extract key themes and related terms from the text data. The input is the content of the virtual store, and the output is the extracted themes and related terms.

[0796] Step 5:

[0797] The server uses an emotion engine to analyze the transmitted facial expressions and voice data and recognize the user's emotions. It uses emotion analysis technologies such as Emotion SDK. The input is the user's facial expressions and voice data, and the output is the recognized emotional state.

[0798] Step 6:

[0799] The server inputs the extracted topic and related words, as well as the recognized emotion data, into a generative AI model. Specifically, the generative AI model (such as GPT-4) generates a text ad based on the prompt sentence. The input is the topic, related words, and emotion data, and the output is the generated ad text. An example of a prompt sentence is as follows:

[0800] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[0801] Step 7:

[0802] The server adjusts the layout and style of the generated advertisement to match the design of the virtual store, including font size, color, margins, etc. using a CSS stylesheet. The input is the generated advertisement text, and the output is the adjusted advertisement.

[0803] Step 8:

[0804] The server inserts the adjusted advertisement into the appropriate location in the virtual store, for example, at the edge of the display where the user is looking or next to a product shelf. The input is the adjusted advertisement, and the output is the updated virtual store content.

[0805] Step 9:

[0806] The server sends the updated virtual store data to the user's terminal. The terminal receives this data and immediately redraws it. The input is the updated virtual store content, and the output is the virtual store redisplayed on the terminal.

[0807] This series of processes allows users to have a seamless shopping experience in a virtual store and view advertisements that are optimal for their emotions.

[0808] 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.

[0809] 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.

[0810] 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.

[0811] [Third embodiment]

[0812] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0813] 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.

[0814] 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).

[0815] 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.

[0816] 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.

[0817] 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).

[0818] 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.

[0819] 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.

[0820] 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.

[0821] 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.

[0822] 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.

[0823] 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."

[0824] The present invention relates to a real-time native advertising generation system using generation AI. Specifically, the system analyzes the content of web pages accessed by users, dynamically generates advertisements that match that content, and inserts them into the web pages, thereby enhancing advertising effectiveness without causing discomfort to users.

[0825] Specific Embodiments of the System

[0826] System Configuration

[0827] This system mainly consists of three main components: the user's device, the server, and the generative AI model.

[0828] 1. User's device

[0829] A device that allows users to access the web. The device has the ability to send requests to web servers and display the retrieved content. Devices include PCs, smartphones, tablets, etc.

[0830] 2. Server

[0831] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also generates advertisements using AI and processes them to insert them into web pages. It also has the function of sending updated web pages to users' devices.

[0832] 3. Generative AI Models

[0833] The generative AI model runs on the server and generates text ads based on the analyzed topics and keywords. The generative AI model uses natural language processing (NLP) techniques.

[0834] Program processing

[0835] When a user accesses a web page, the following sequence of events occurs:

[0836] 1. User Web Access

[0837] A user accesses a specific web page. For example, the user clicks on a blog page with the theme "Autumn Kyoto Sightseeing."

[0838] The terminal sends a request to the server requesting the HTML content of the web page.

[0839] 2. Getting HTML data

[0840] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[0841] 3. Content Analysis

[0842] The server analyzes the HTML content and extracts key topics and keywords. For example, the analysis extracts keywords such as "autumn leaves," "famous places," and "guided tours."

[0843] 4. Ad generation

[0844] The server inputs the extracted keywords into a generative AI model to generate ad text related to the content, such as an ad for a "Kyoto Autumn Foliage Tour Special."

[0845] 5. Ad Styling

[0846] The server adjusts the style and layout of the generated advertising text to match the design of the web page.

[0847] 6. Ad Insertion

[0848] The server then regenerates the HTML for the web page, inserting the ad in the appropriate place, for example after the second paragraph.

[0849] 7. Submitting Updated Content

[0850] The server sends the updated HTML content to the user's device, which redisplays it.

[0851] Specific examples

[0852] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following happens:

[0853] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[0854] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[0855] The server inputs these keywords into the generation AI, which generates advertising text such as "Special feature on Kyoto autumn foliage tours."

[0856] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[0857] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[0858] In this way, the system realizes native advertising that does not cause discomfort to users, maximizes advertising effectiveness, and improves the user experience.

[0859] The processing flow will be explained below.

[0860] Step 1:

[0861] A user accesses a web page of interest. For example, a user accesses a blog post about "Autumn Kyoto Sightseeing." The device sends a request to that URL, requesting the HTML content of the web page from the server.

[0862] Step 2:

[0863] The server receives the user's request and retrieves the appropriate HTML content from a database or file system. The web server then prepares a response containing all the data for the web page (text, images, style sheets, etc.).

[0864] Step 3:

[0865] The server analyzes the HTML retrieved. Specifically, it uses an HTML parser to analyze the structure of the web page and extract the text content. For example, it extracts the entire body of an article titled "Sightseeing in Kyoto in Autumn."

[0866] Step 4:

[0867] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the extracted text. For example, keywords such as "autumn leaves," "famous places," and "guided tours" might be found. This process may involve the use of NLP libraries or APIs (e.g., SpaCy or NLTK).

[0868] Step 5:

[0869] The server initializes the generative AI model and provides the extracted keywords as input data. The generative AI model (e.g., GPT) is used to generate text for native ads based on the keywords. For example, an ad copy such as "Special feature on Kyoto autumn foliage tours" is generated. Here, a pre-trained model is used for the generative AI.

[0870] Step 6:

[0871] The server adjusts the layout and style of the generated advertisement to match the design of the web page, specifically by applying a CSS stylesheet so that the advertisement blends into the existing design of the web page. This adjustment may include font size, color, margins, etc.

[0872] Step 7:

[0873] The server updates the HTML content, inserting the generated ad in the appropriate place (e.g., adding an ad after the third paragraph of a blog post), and restructures the entire HTML document to ensure consistency.

[0874] Step 8:

[0875] The server sends the updated HTML content to the device, and the device reads the new HTML data received from the server and redraws the web page.

[0876] Step 9:

[0877] Users can browse the updated web page and continue reading the original content without noticing the generated native ad, making the user experience natural and seamless.

[0878] Example 1

[0879] 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."

[0880] Conventional ad delivery methods often lack a proper blend of web page content and ads, creating a sense of discomfort for users. Furthermore, to maximize advertising effectiveness, it is necessary to generate and display ads in real time that are closely related to the content being viewed by users. However, the technology to accomplish this efficiently has not yet been fully developed.

[0881] 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.

[0882] In this invention, the server includes means for acquiring web page information from user web access, means for analyzing the web page information in real time and extracting main topics and keywords, means for generating advertisements using a generative AI model based on the extracted topics and keywords, means for adjusting the layout and style of the generated advertisement to match the design of the web page, means for inserting the generated advertisement into an appropriate position on the web page, and means for transmitting the updated web page to the terminal and redisplaying it, thereby enabling advertisements to be displayed in a way that does not cause discomfort to the user and maximizing advertising effectiveness.

[0883] "Web access" refers to the act of a user connecting to a web page via the Internet and obtaining information.

[0884] "Web page information" is a collection of data such as text, images, videos, and links displayed on a website.

[0885] "Real-time" refers to processing occurring immediately without delay.

[0886] "Analysis" refers to examining the content of data in detail, especially using natural language processing to extract specific information.

[0887] A "topic" refers to the main topic or subject matter in text data.

[0888] "Keywords" are important words or phrases in a document that help identify the topic.

[0889] A "generative AI model" refers to an algorithm or system that uses artificial intelligence to generate new content or data.

[0890] "Advertising" is information used to promote products and services to consumers.

[0891] "Layout" refers to the position and structure of each element arranged within a web page.

[0892] "Style" refers to the appearance and design of text and images on a web page, including color, font, size, etc.

[0893] "Insertion" means adding new information or elements into an existing Web page.

[0894] A "terminal" is a device that a user uses to access the Internet. Examples include personal computers, smartphones, and tablets.

[0895] "Redisplay" means displaying updated content or data again on the user's terminal.

[0896] This invention relates to a real-time native advertising generation system using generative AI. Specifically, it provides a system that analyzes the content of web pages accessed by users, dynamically generates advertisements that fit the content, and inserts them into the web pages, thereby enhancing advertising effectiveness without causing discomfort to users.

[0897] System Configuration

[0898] The system mainly consists of three main components: the user's device, the server, and the generative AI model.

[0899] 1. User's device

[0900] A device that allows users to access the web. The device has the ability to send requests to web servers and display the retrieved content. Devices include PCs, smartphones, tablets, etc.

[0901] 2. Server

[0902] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also generates advertisements using AI and processes them to insert them into web pages. It also has the function of sending updated web pages to users' devices.

[0903] 3. Generative AI Models

[0904] The generative AI model runs on a server and generates text ads based on the analyzed topics and keywords. Natural language processing (NLP) techniques are used for generative AI models. Specific techniques include generative models such as GPT-3.

[0905] Server Processing

[0906] When a user accesses a specific web page, the device requests the HTML content of that web page from the server. The server receives the request and retrieves the relevant HTML data. The server then analyzes the retrieved HTML content to extract key topics and keywords. For example, NLTK and spaCy are used to extract keywords such as "autumn leaves," "famous places," and "guided tours."

[0907] Next, the server inputs the extracted keywords into a generative AI model to generate advertising text related to the content. For example, an advertisement such as "Special feature on Kyoto autumn foliage tours" may be generated. An example of a prompt used in this case is "Please generate advertising text on the themes of 'Kyoto autumn foliage,' 'famous places,' and 'sightseeing.'"

[0908] The server then adjusts the generated ad text to match the style and layout of the webpage, using CSS and JavaScript to make the ad look consistent with the webpage. The server then inserts the ad in the appropriate position and regenerates the webpage's HTML, for example, inserting the ad after the second paragraph.

[0909] Finally, the server sends the updated HTML content to the user's device, which redisplays it, allowing the user to seamlessly browse the page.

[0910] Specific examples

[0911] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following will happen:

[0912] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[0913] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[0914] The server inputs these keywords into the generation AI, which generates advertising text such as "Special feature on Kyoto autumn foliage tours."

[0915] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[0916] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[0917] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0918] Step 1:

[0919] A user accesses a specific web page. Specifically, the user enters a URL in the browser's address bar or clicks a link. The URL is entered as input, and the device sends an HTTP request to the server requesting the HTML content corresponding to that URL. The server receives this request as output.

[0920] Step 2:

[0921] The server receives a user request and retrieves the relevant HTML content. Specifically, the server loads the HTML file corresponding to the requested URL from a database or file system. The input is the requested URL, and the output is the corresponding HTML content.

[0922] Step 3:

[0923] The server analyzes the retrieved HTML content and extracts key topics and keywords. Specifically, it uses a natural language processing (NLP) library to perform text analysis. For example, it uses "NLTK" or "spaCy." The input is the entire HTML content, and the output is extracted keywords such as "autumn leaves," "famous places," and "guided tours."

[0924] Step 4:

[0925] The server inputs the extracted keywords into the generative AI model and generates advertising text related to the content. Specifically, the server sends a prompt to the generative AI model. An example of this prompt is "Please generate advertising text on the themes of 'Kyoto autumn leaves,' 'famous places,' and 'sightseeing.'" The input is the extracted keywords and the prompt, and the output is advertising text such as "Special feature on Kyoto autumn leaves tours."

[0926] Step 5:

[0927] The server adjusts the style and layout of the generated ad text to match the web page design. Specifically, it uses CSS and JavaScript to adjust the font, color, size, etc. of the ad text to match the web page design. The input is the generated ad text, and the output is the adjusted ad text.

[0928] Step 6:

[0929] The server inserts the ad in the appropriate position and regenerates the HTML for the webpage, using JavaScript DOM manipulation and a server-side template engine. For example, inserting the ad after the second paragraph. The input is the adjusted ad text and the original HTML content, and the output is the updated HTML content with the ad inserted.

[0930] Step 7:

[0931] The server sends the updated HTML content to the user's device, which then redisplays it. Specifically, the newly generated HTML is returned to the device as an HTTP response. The input is the updated HTML content with the ad inserted, and the output is the updated web page displayed on the device. This allows the user to view the page seamlessly.

[0932] (Application example 1)

[0933] 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."

[0934] Modern internet advertising tends to be intrusive to users, often detracting from the user experience. Furthermore, static advertisements tend to be less relevant to the page content, reducing advertising effectiveness. Conventional advertising systems are unable to dynamically generate advertisements that match the content of web pages, which can be unsettling for users. The present invention aims to solve these problems.

[0935] 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.

[0936] In this invention, the server includes means for acquiring web page content through user web access, means for analyzing the web page content in real time and extracting key topics and keywords, and means for generating native advertisements using a generation AI based on the extracted topics and keywords. This makes it possible to improve the user experience while increasing advertising effectiveness by inserting dynamically generated advertisements in appropriate positions based on the content viewed by the user.

[0937] "User web access" refers to the act of a user accessing a web page via the Internet and obtaining or viewing information.

[0938] "Web page content" refers collectively to text, images, video, and other media or information displayed on a web page.

[0939] "Real-time analysis" is the process of analyzing the content of a web page immediately after a user accesses it.

[0940] "Major topics and keywords" refers to the important subjects and keywords in the content of a web page, which serve as the basis for generating advertisements.

[0941] "Generative AI" is a system that uses artificial intelligence techniques to generate new text based on specific prompts.

[0942] "Native advertising" is advertising that is displayed in a way that is integrated with the content of a web page and is designed to not cause any discomfort to the user.

[0943] A "dynamically generated advertisement based on content" is an advertisement that is automatically generated in real time according to the content of the web page that the user is viewing.

[0944] "Inserting in an appropriate position on a web page" refers to placing advertisements in appropriate locations on a web page so as not to impair the user experience.

[0945] "User experience" refers to the overall experience and satisfaction a user has when using a product or service.

[0946] An "application" is a software program that runs on a device and provides a particular function or service.

[0947] This invention is a system that uses generative AI to generate native ads in real time and insert them into the content of web pages accessed by users. This article mainly explains an example of applying this system to a news application for smartphones.

[0948] System Configuration

[0949] The system consists of three main components: the user's device, a server, and a generative AI model.

[0950] 1. User's device

[0951] This includes devices such as smartphones and tablets.

[0952] It provides an interface for users to access articles through a news app.

[0953] 2. Server

[0954] It plays a central role in retrieving and parsing the HTML content of web pages.

[0955] Generative AI is used to generate advertisements, and processing is performed to properly insert the advertisements into web pages.

[0956] 3. Generative AI Models

[0957] Natural language processing (NLP) techniques are used to generate text ads based on analyzed topics and keywords.

[0958] Generative AI models used include, for example, OpenAI's GPT-4.

[0959] Program processing overview

[0960] When a user accesses a web page, the server performs the following process.

[0961] 1. User Web Access

[0962] A user clicks on a particular article in a news app, and the device requests the HTML content of that news article from the server.

[0963] 2. Getting HTML data

[0964] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[0965] 3. Content Analysis

[0966] The server analyzes the HTML content and extracts key topics and keywords, such as "climate change," "renewable energy," and "eco-activities."

[0967] 4. Ad generation

[0968] The server inputs the extracted keywords into a generative AI model to generate advertising text related to the content. An example of a prompt sentence to be used is as follows:

[0969] Generate ad text based on the following keywords: climate change, renewable energy, eco-activism

[0970] 5. Ad Styling

[0971] The server adjusts the style and layout of the generated advertising text to match the design of the news article.

[0972] 6. Ad Insertion

[0973] The server inserts the advertisement at the appropriate position in the news article, for example after the second paragraph of the article.

[0974] 7. Submitting Updated Content

[0975] The server sends the updated HTML content to the user's device, which redisplays it.

[0976] Specific examples

[0977] For example, if a user accesses an article on the topic of "Thinking about Environmental Issues" in a news app, the following steps are taken:

[0978] The user clicks on the URL of a news article, and the device requests the HTML content from the server.

[0979] The server retrieves the HTML content, analyzes it, and extracts the keywords "climate change," "renewable energy," and "eco-activities."

[0980] The server inputs these keywords into a generation AI, which generates ad text such as "Introducing a new electric vehicle."

[0981] The server adapts the generated advertisements to the style of the news article and inserts them at the appropriate locations.

[0982] The updated page will be redisplayed on the user's device, allowing the user to continue reading the article without any discomfort.

[0983] This invention makes it possible to maximize advertising effectiveness by dynamically generating and inserting advertisements that match the content without impairing the user experience.

[0984] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0985] Step 1:

[0986] A user clicks on a specific article in a news app.

[0987] Input: The URL of the news article selected by the user.

[0988] Specific operation: The device sends a request to the server for the HTML content of the news article.

[0989] Output: The server receives the request.

[0990] Step 2:

[0991] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[0992] Input: A news article URL request.

[0993] Specific operation: The server searches for and retrieves the corresponding HTML content from the database.

[0994] Output: The retrieved HTML content.

[0995] Step 3:

[0996] The server parses the retrieved HTML content and extracts key topics and keywords.

[0997] Input: The retrieved HTML content.

[0998] What it does: The server uses natural language processing (NLP) techniques to parse the HTML content and extract key topics and keywords (e.g., "climate change," "renewable energy," "green activities").

[0999] Output: Extracted main topics and keywords.

[1000] Step 4:

[1001] The server inputs the extracted keywords into a generative AI model to generate advertising text relevant to the content.

[1002] Input: Extracted main topics and keywords.

[1003] Specific operation: The server sends the following prompt to the generated AI model:

[1004] Generate ad text based on the following keywords: climate change, renewable energy, eco-activism

[1005] A generative AI model generates ad text.

[1006] Output: The generated ad text.

[1007] Step 5:

[1008] The server adjusts the style and layout of the generated advertising text to match the design of the news article.

[1009] Input: The generated ad text.

[1010] What it does: The server uses CSS and HTML templates to adjust the ad design and integrate it into the style of the news article.

[1011] Output: The adjusted ad text.

[1012] Step 6:

[1013] The server inserts the advertisements into the news articles at appropriate locations.

[1014] Input: Your adjusted ad text.

[1015] Specific behavior: The server regenerates the HTML content of the news article, for example inserting an advertisement after the second paragraph of the article.

[1016] Output: The updated HTML content.

[1017] Step 7:

[1018] The server sends the updated HTML content to the user's device, and the user redisplays it through the news app.

[1019] Input: The updated HTML content.

[1020] What happens: The server sends HTML content to the device, and the news app receives and displays it.

[1021] Output: The user can view the updated news article.

[1022] These steps ensure that users are presented with ads that are tailored to their content and that fit their content seamlessly, improving the user experience.

[1023] 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.

[1024] This invention relates to a system that recognizes user emotions and automatically inserts native ads generated based on those emotions into web pages. Specifically, the system analyzes the content of web pages retrieved by users' web access in real time and extracts key topics and keywords. It also uses an emotion engine to recognize user emotions and generates native ads based on this emotion using a generation AI. The generated native ads are inserted in appropriate positions on the web page, and the updated web page is sent to the terminal. This system provides ads with optimal timing and content to users, achieving higher advertising effectiveness.

[1025] Specific Embodiments of the System

[1026] System Configuration

[1027] This system mainly consists of four main components: the user's device, the server, the generative AI model, and the emotion engine.

[1028] 1. User's device

[1029] A device that allows users to access the web. The terminal has the function of sending requests to a web server and displaying the retrieved content. It also has the function of collecting data such as the user's facial expressions, voice tone, and text input, and sending it to the server.

[1030] 2. Server

[1031] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also analyzes user emotion data, generates advertisements based on the emotions using generative AI, and processes the advertisements to be inserted into web pages. It also has the function of sending updated web pages to users' devices.

[1032] 3. Generative AI Models

[1033] The generative AI model runs on a server and generates text ads based on the analyzed topics and keywords, and also takes into account user sentiment data to generate ads with the appropriate tone and content.

[1034] 4. Emotion Engine

[1035] The emotion engine analyzes users' facial expressions, voice tone, and text input to recognize their emotional state in real time. This emotional data is fed into a generative AI model and used to generate ads.

[1036] Program processing

[1037] When a user accesses a web page, the following sequence of events occurs:

[1038] 1. User Web Access

[1039] When a user accesses a specific web page, for example, a blog post about "Autumn Kyoto Sightseeing," the device sends a request to the URL, requesting the HTML content of the web page from the server.

[1040] 2. Getting HTML data

[1041] The server receives the user's request and retrieves the relevant HTML data from a database or file system. The web server then prepares a response containing all the data for the web page (text, images, style sheets, etc.).

[1042] 3. Content Analysis

[1043] The server analyzes the HTML retrieved. Specifically, it uses an HTML parser to analyze the structure of the web page and extract the text content. For example, it extracts the entire body of an article titled "Sightseeing in Kyoto in Autumn."

[1044] 4. Topic and Keyword Extraction

[1045] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the extracted text. For example, keywords such as "autumn leaves," "famous places," and "guided tours" might be found. This process may involve the use of NLP libraries or APIs (e.g., SpaCy or NLTK).

[1046] 5. Emotional Recognition

[1047] While a user is browsing a web page, the device collects data such as facial expressions and voice and sends it to the emotion engine. The emotion engine analyzes this data and identifies the user's emotional state. For example, if the user is smiling, it is judged to be "happy," and if they are frowning, it is judged to be "unhappy."

[1048] 6. Ad Generation

[1049] The server inputs the extracted keywords and recognized emotion data into a generative AI model to generate native ads that match the emotion. For example, if the user is in a "happy" state, it generates an ad with a positive tone, and if the user is in a "dissatisfied" state, it generates a problem-solving ad.

[1050] 7. Ad Styling

[1051] The server adjusts the layout and style of the generated advertisement to match the design of the web page, specifically by applying a CSS stylesheet so that the advertisement blends into the existing design of the web page. This adjustment may include font size, color, margins, etc.

[1052] 8. Ad Insertion

[1053] The server updates the HTML content, inserting the generated ad in the appropriate place (e.g., adding an ad after the third paragraph of a blog post), and restructures the entire HTML document to ensure consistency.

[1054] 9. Submitting Updated Content

[1055] The server sends the updated HTML content to the device, and the device reads the new HTML data received from the server and redraws the web page.

[1056] Specific examples

[1057] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following happens:

[1058] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[1059] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[1060] The server inputs these keywords into the AI ​​generator, which generates ad text such as "Kyoto Autumn Foliage Tour Special." It also generates ad text with a positive tone based on the user's emotional data (e.g., "happiness").

[1061] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[1062] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[1063] In this way, the system provides advertisements that best suit the user's emotional state, maximizing advertising effectiveness and improving the user experience.

[1064] The processing flow will be explained below.

[1065] Step 1:

[1066] A user clicks on a URL in their web browser to access a blog post about "Autumn Kyoto Sightseeing." The device sends a request to the server for that URL, requesting the HTML content of the web page.

[1067] Step 2:

[1068] The server receives the user's request and retrieves the appropriate HTML content from a database or file system.

[1069] Step 3:

[1070] The HTML content retrieved by the server is analyzed using an HTML parser, extracting each element of the web page, such as text content, images, and style sheets.

[1071] Step 4:

[1072] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the text content, such as "autumn leaves," "famous places," and "guided tours."

[1073] Step 5:

[1074] While a user is browsing a web page, the device uses a camera and microphone to transmit the user's facial expressions and voice tone to the emotion engine, which analyzes this data and recognizes the user's emotional state. For example, a smile on the user's face indicates "happiness," while a frown indicates "dissatisfaction."

[1075] Step 6:

[1076] The server inputs the extracted keywords and recognized emotion data into a generative AI model. The generative AI generates native ads based on the keywords while adjusting the tone and content of the ads based on the emotion data. For example, if the user is in a "happy" state, it generates a positive-toned ad for a "Kyoto Autumn Foliage Tour Special Campaign."

[1077] Step 7:

[1078] The server adjusts the layout and style of the generated ad to match the design of the web page, for example by using CSS to adjust the font size, color, and margins of the ad so that it blends seamlessly into the web page.

[1079] Step 8:

[1080] The server generates the updated HTML content and inserts the ad in the appropriate position (e.g. after the third paragraph of the article).

[1081] Step 9:

[1082] The server sends the updated web page to the device, which reads the new HTML data and redraws the web page.

[1083] Step 10:

[1084] When users browse the updated web page, they will notice the generated native ads and can continue reading the page content naturally. Users will receive ads that match their emotions, resulting in a more comfortable browsing experience.

[1085] Example 2

[1086] 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."

[1087] Current ad delivery systems struggle to provide effective ads that take into account the user's emotional state. Furthermore, the inconsistency between the content of a web page and the ads often leads to a poor user experience. Therefore, a system is needed that can generate and display ads in real time that are optimized for the user's emotions and the content of a web page.

[1088] 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.

[1089] In this invention, the server includes means for acquiring web page content through user web access, means for analyzing the web page content in real time and extracting key topics and keywords, means for collecting data such as the user's facial expressions, voice tone, and text input and recognizing emotions, means for generating a native advertisement using a generation AI based on the extracted topics and keywords and the recognized emotion data, means for adjusting the layout and style of the generated native advertisement to match the design of the web page, means for inserting the generated native advertisement at an appropriate position on the web page, and means for transmitting the updated web page to a terminal and redisplaying it. This makes it possible to generate and display an advertisement optimized for the user's emotional state in real time.

[1090] "Web access" refers to a user accessing a specific web page via the Internet.

[1091] "Web page content" refers to all data elements contained within a Web page, including text, images, style sheets, scripts, etc.

[1092] "Real-time" refers to data processing and analysis occurring almost immediately.

[1093] "Topic" refers to the main theme that describes the content or subject matter of a Web page.

[1094] "Keywords" refer to important words or phrases that represent the content of a web page.

[1095] "Facial expression" refers to facial changes and characteristics that represent the user's emotional state.

[1096] "Voice tone" refers to the emotional inflections and nuances contained in a user's voice.

[1097] "Text input" refers to text data that a user inputs on a web page.

[1098] "Emotion recognition" refers to analyzing and identifying a user's emotional state from their facial expressions, voice, and text.

[1099] "Generative AI" refers to algorithms or models that use artificial intelligence techniques to generate new content (e.g., advertisements) based on specific input data.

[1100] "Native advertising" refers to advertising that blends naturally into the content of a web page.

[1101] "Layout" refers to the presentation and structure of a web page or advertisement.

[1102] "Style" refers to the design elements of a web page or advertisement (e.g., fonts, colors, spacing).

[1103] A "server" refers to a computer system that provides information and services over a network.

[1104] "Terminal" refers to a device, such as a computer or smartphone, that a user uses to access the Internet.

[1105] "Redisplay" refers to displaying the updated web page on the screen again.

[1106] This invention relates to a system that recognizes user emotions and automatically inserts native ads into web pages based on those emotions. The system mainly consists of four main components: a user device, a server, a generative AI model, and an emotion engine.

[1107] System Configuration and Hardware / Software Usage

[1108] 1. User's device

[1109] A device that allows users to access the web. The terminal has the function of sending requests to a web server and displaying the retrieved content. It also has the function of collecting data such as the user's facial expressions, voice tone, and text input and sending it to the server. This device is equipped with input devices such as a camera and microphone.

[1110] 2. Server

[1111] The server plays a central role in retrieving and analyzing the HTML content of web pages. The server retrieves HTML data using web server software such as Apache or Nginx, and parses the content using an HTML parser library such as BeautifulSoup. It also has the ability to analyze user emotion data and generate advertisements using generative AI based on the emotion. It also has the ability to generate HTTP responses to send updated web pages to users' devices.

[1112] 3. Generative AI Models

[1113] The generative AI model runs on the server and generates text ads based on the analyzed topics and keywords. This model, such as OpenAI's GPT-3, generates appropriate ads by inputting a prompt. An example of a prompt is "Please create an ad for an autumn foliage tour with a happy tone."

[1114] 4. Emotion Engine

[1115] The emotion engine analyzes the user's facial expressions, voice tone, and text input to recognize the user's emotional state in real time. The collected emotion data is sent to a server and input into a generative AI model. This emotion engine may utilize, for example, Microsoft's Azure Emotion API or Google's Cloud Vision API.

[1116] Specific examples

[1117] As a concrete example, consider the case where a user accesses a blog page with the theme "Autumn Kyoto Sightseeing." The device collects the user's facial expressions and voice tone and sends them to the emotion engine. As a result, if the user is smiling, it is determined to be "happy." This emotional state, along with the topic and keywords of the blog page (e.g., "autumn leaves," "famous places," and "sightseeing"), are input into a generative AI model to generate advertising text with a positive tone, such as "Special feature on Kyoto autumn leaves tours."

[1118] An example of a prompt for the generative AI model is "Please create an advertisement for an autumn foliage tour in a happy tone." The server adjusts the layout and style of the generated advertisement using a CSS stylesheet to blend in with the existing design of the webpage, and inserts it in the appropriate position after the third paragraph.

[1119] The server then sends the updated HTML content to the user's device, which quickly redraws it, allowing the user to smoothly view the updated web page.

[1120] This invention makes it possible to generate and display advertisements optimized for the user's emotional state in real time, thereby maximizing advertising effectiveness and providing a superior user experience.

[1121] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1122] Step 1:

[1123] A user visits a web page. The user sends an HTTP request for the URL of a particular web page (e.g., "Autumn Kyoto Sightseeing"). The input is the URL of the web page, and the output is that the request is sent to the server.

[1124] Step 2:

[1125] The server retrieves the HTML data. The server analyzes the received HTTP request and retrieves the HTML data for the corresponding web page from a database or file system. The input is the URL of the web page, and the specific operation is for web server software such as Apache or Nginx to load the HTML file. The output is the HTML data for the web page.

[1126] Step 3:

[1127] The server parses the HTML data. It uses an HTML parser library such as BeautifulSoup to analyze the structure of the web page and extract the text content. The input is the HTML data of the web page, and the specific operation is to parse the HTML structure. The output is the parsed text content.

[1128] Step 4:

[1129] The server extracts key topics and keywords. The server uses natural language processing (NLP) techniques, specifically SpaCy or NLTK, to extract keywords from the text. The input is the parsed text content, and the specific operation is NLP analysis. The output is key topics and keywords (e.g., "autumn leaves," "famous places," "sightseeing").

[1130] Step 5:

[1131] Collecting user emotional data. The device uses the built-in camera and microphone to collect data such as the user's facial expressions, voice tone, and text input. The input is the user's facial expressions and voice, and the output is the collected emotional data.

[1132] Step 6:

[1133] The emotion engine analyzes the emotional data. The emotion engine analyzes the collected data in real time and identifies the user's emotional state. Specifically, it uses Microsoft's Azure Emotion API or Google's Cloud Vision API. The input is emotional data, and the specific action is analysis via the API. The output is the user's emotional state (e.g., "happy," "unhappy").

[1134] Step 7:

[1135] The server generates advertisements using a generative AI model. The server inputs the extracted keywords and recognized emotion data into the generative AI model (e.g., OpenAI's GPT-3). The advertisement is generated using a prompt (e.g., "Please create an advertisement for an autumn foliage tour in a happy tone"). The input is keywords and emotion data, and the prompt is input into the generative AI model as a specific action. The output is the generated advertisement text.

[1136] Step 8:

[1137] The server adjusts the layout and style of the ad. It adjusts the layout and style using a CSS stylesheet so that the generated ad matches the design of the webpage. The input is the generated ad text, and the specific action is to apply the CSS stylesheet. The output is the adjusted ad layout.

[1138] Step 9:

[1139] The server inserts the advertisement into the web page. The generated advertisement is inserted into the updated HTML content at the appropriate position. For example, an advertisement is added after the third paragraph of a blog post. The input is the adjusted advertisement layout and HTML content, and the specific operation is to reconstruct the HTML document. The output is the updated HTML content.

[1140] Step 10:

[1141] The server sends the updated web page to the device. The device receives the updated content and redraws the web page. The input is the updated HTML content, which is sent to the device as an HTTP response as a specific action. The output is the updated web page displayed on the device.

[1142] This allows the user to view advertisements that match their emotions seamlessly along with the displayed web page.

[1143] (Application example 2)

[1144] 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."

[1145] Current virtual stores display uniform advertisements without considering the user's emotional state, resulting in a poor user experience and limited advertising effectiveness. Furthermore, it is difficult to accurately suggest information and products that users desire in real time, and there is a lack of technology to seamlessly provide advertisements tailored to individual users. This results in lower ad click rates and lower purchasing intent, preventing advertisers from implementing effective marketing strategies.

[1146] 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.

[1147] In this invention, the server includes means for acquiring virtual space content through user access, means for analyzing the virtual space content in real time and extracting main themes and related words, means for recognizing emotions from the user's facial expressions and voice, means for generating native advertisements using a generation AI based on the extracted themes and related words and the recognized emotions, means for inserting the generated native advertisements into appropriate positions in the virtual space, and means for transmitting the updated virtual space to the terminal and redisplaying it, thereby making it possible to provide advertisements optimized for the user's emotional state in real time.

[1148] "User" refers to a person who uses a virtual space or system.

[1149] "Access" refers to the action of a user connecting to a specific virtual space or web page.

[1150] "Virtual space" refers to a space that is generated using digital technology and that can be visually experienced but is different from a real location.

[1151] "Content" refers to information such as text, images, video, and audio displayed in virtual spaces and on web pages.

[1152] "Real-time analysis" refers to processing data immediately in response to user operations or access.

[1153] "Major subject" refers to a topic that is considered particularly important within a document or piece of content.

[1154] "Related words" refer to words that are related to the main topic and provide additional information.

[1155] "Facial expression" refers to an element used to infer emotions by analyzing the user's facial movements and features.

[1156] "Voice" refers to the element used to analyze the user's voice to read their emotions and intentions.

[1157] "Emotion recognition" refers to the process of analyzing and identifying a person's emotional state from their facial expressions and voice.

[1158] "Generative AI" refers to artificial intelligence that uses machine learning models to automatically generate content and advertisements.

[1159] "Native advertising" refers to advertising that blends naturally into the format and design of the medium in which it is displayed.

[1160] "Appropriate location" refers to a location where the advertisement is displayed effectively and without interfering with the user's experience.

[1161] "Terminal" refers to a hardware device that allows a user to access a system or virtual space.

[1162] This invention relates to a system that recognizes user emotions in real time and displays appropriate native advertisements in a virtual store based on those emotions. This system mainly consists of four main components: a user device, a server, a generative AI model, and an emotion engine.

[1163] System Configuration

[1164] User's device

[1165] Users access the virtual store using devices such as smartphones and head-mounted displays (HMDs). These devices have the ability to collect the user's facial expressions and voice data and send it to a server.

[1166] server

[1167] The server retrieves the virtual store's content and performs real-time analysis. It also extracts key themes and related terms and analyzes user sentiment data. It then uses a generative AI model to generate native ads based on the extracted data and inserts them into the virtual store at appropriate locations. Finally, it sends the updated virtual store data to the user's device.

[1168] Generative AI Models

[1169] The generative AI model generates text ads based on the extracted subject and related words, as well as user sentiment data. This model uses advanced language generation techniques such as GPT-4. An example prompt is as follows:

[1170] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[1171] Emotion Engine

[1172] The emotion engine analyzes the user's facial expressions and voice to recognize emotions in real time. It uses emotion analysis technologies such as the Emotion SDK to identify the user's emotional state.

[1173] Processing flow

[1174] When a user accesses a virtual store using a smartphone or HMD, the device captures facial expression and voice data and sends them to the server, which then retrieves the content of the virtual store and analyzes it using a natural language processing library (e.g., SpaCy or NLTK) to extract key themes and related terms.

[1175] In parallel, the emotion engine recognizes the user's emotions and sends the data to the server, which then inputs the extracted subject and related words, as well as the emotion data, into a generative AI model to generate native ads. The generated ads are then adjusted to fit the design of the virtual store and inserted at the appropriate locations.

[1176] The updated virtual store data is sent from the server to the user's device and instantly redisplayed, allowing the user to seamlessly browse the virtual store and experience emotionally optimized advertising.

[1177] As a specific example, if a user accesses a virtual store on their smartphone and smiles while browsing the "latest gadgets" section, the system will recognize the user's emotion as "happiness" and send the following prompt to the generative AI model:

[1178] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[1179] The generated ad text is optimized to match the user's emotional state, such as "We're having a special sale on the latest gadgets right now! Get the latest technology that will make you smile at a great price!" This improves the user experience and maximizes the effectiveness of the ad.

[1180] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1181] Step 1:

[1182] A user accesses a virtual store using a device (smartphone or HMD). The device sends an access request to the server, requesting content from the virtual store. The input is the user's access request, and the output is a request to the server.

[1183] Step 2:

[1184] The server receives the access request and retrieves the content of the virtual store from a database or file system. This content can include text, images, videos, etc. The input is the request, and the output is the content of the virtual store.

[1185] Step 3:

[1186] The device captures the user's facial expressions and voice data and transmits this data to the server in real time. The input is the user's facial expressions and voice data, and the output is data transmission to the server. Specifically, data is collected from the smartphone camera and microphone and transmitted in real time.

[1187] Step 4:

[1188] The server analyzes the received content in real time using natural language processing libraries (e.g., SpaCy or NLTK) to extract key themes and related terms from the text data. The input is the content of the virtual store, and the output is the extracted themes and related terms.

[1189] Step 5:

[1190] The server uses an emotion engine to analyze the transmitted facial expressions and voice data and recognize the user's emotions. It uses emotion analysis technologies such as Emotion SDK. The input is the user's facial expressions and voice data, and the output is the recognized emotional state.

[1191] Step 6:

[1192] The server inputs the extracted topic and related words, as well as the recognized emotion data, into a generative AI model. Specifically, the generative AI model (such as GPT-4) generates a text ad based on the prompt sentence. The input is the topic, related words, and emotion data, and the output is the generated ad text. An example of a prompt sentence is as follows:

[1193] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[1194] Step 7:

[1195] The server adjusts the layout and style of the generated advertisement to match the design of the virtual store, including font size, color, margins, etc. using a CSS stylesheet. The input is the generated advertisement text, and the output is the adjusted advertisement.

[1196] Step 8:

[1197] The server inserts the adjusted advertisement into the appropriate location in the virtual store, for example, at the edge of the display where the user is looking or next to a product shelf. The input is the adjusted advertisement, and the output is the updated virtual store content.

[1198] Step 9:

[1199] The server sends the updated virtual store data to the user's terminal. The terminal receives this data and immediately redraws it. The input is the updated virtual store content, and the output is the virtual store redisplayed on the terminal.

[1200] This series of processes allows users to have a seamless shopping experience in a virtual store and view advertisements that are optimal for their emotions.

[1201] 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.

[1202] 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.

[1203] 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.

[1204] [Fourth embodiment]

[1205] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1206] 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.

[1207] 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).

[1208] 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.

[1209] 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.

[1210] 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).

[1211] 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.

[1212] 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.

[1213] 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.

[1214] 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.

[1215] 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.

[1216] 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.

[1217] 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."

[1218] The present invention relates to a real-time native advertising generation system using generation AI. Specifically, the system analyzes the content of web pages accessed by users, dynamically generates advertisements that match that content, and inserts them into the web pages, thereby enhancing advertising effectiveness without causing discomfort to users.

[1219] Specific Embodiments of the System

[1220] System Configuration

[1221] This system mainly consists of three main components: the user's device, the server, and the generative AI model.

[1222] 1. User's device

[1223] A device that allows users to access the web. The device has the ability to send requests to web servers and display the retrieved content. Devices include PCs, smartphones, tablets, etc.

[1224] 2. Server

[1225] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also generates advertisements using AI and processes them to insert them into web pages. It also has the function of sending updated web pages to users' devices.

[1226] 3. Generative AI Models

[1227] The generative AI model runs on the server and generates text ads based on the analyzed topics and keywords. The generative AI model uses natural language processing (NLP) techniques.

[1228] Program processing

[1229] When a user accesses a web page, the following sequence of events occurs:

[1230] 1. User Web Access

[1231] A user accesses a specific web page. For example, the user clicks on a blog page with the theme "Autumn Kyoto Sightseeing."

[1232] The terminal sends a request to the server requesting the HTML content of the web page.

[1233] 2. Getting HTML data

[1234] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[1235] 3. Content Analysis

[1236] The server analyzes the HTML content and extracts key topics and keywords. For example, the analysis extracts keywords such as "autumn leaves," "famous places," and "guided tours."

[1237] 4. Ad generation

[1238] The server inputs the extracted keywords into a generative AI model to generate ad text related to the content, such as an ad for a "Kyoto Autumn Foliage Tour Special."

[1239] 5. Ad Styling

[1240] The server adjusts the style and layout of the generated advertising text to match the design of the web page.

[1241] 6. Ad Insertion

[1242] The server then regenerates the HTML for the web page, inserting the ad in the appropriate place, for example after the second paragraph.

[1243] 7. Submitting Updated Content

[1244] The server sends the updated HTML content to the user's device, which redisplays it.

[1245] Specific examples

[1246] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following happens:

[1247] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[1248] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[1249] The server inputs these keywords into the generation AI, which generates advertising text such as "Special feature on Kyoto autumn foliage tours."

[1250] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[1251] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[1252] In this way, the system realizes native advertising that does not cause discomfort to users, maximizes advertising effectiveness, and improves the user experience.

[1253] The processing flow will be explained below.

[1254] Step 1:

[1255] A user accesses a web page of interest. For example, a user accesses a blog post about "Autumn Kyoto Sightseeing." The device sends a request to that URL, requesting the HTML content of the web page from the server.

[1256] Step 2:

[1257] The server receives the user's request and retrieves the appropriate HTML content from a database or file system. The web server then prepares a response containing all the data for the web page (text, images, style sheets, etc.).

[1258] Step 3:

[1259] The server analyzes the HTML retrieved. Specifically, it uses an HTML parser to analyze the structure of the web page and extract the text content. For example, it extracts the entire body of an article titled "Sightseeing in Kyoto in Autumn."

[1260] Step 4:

[1261] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the extracted text. For example, keywords such as "autumn leaves," "famous places," and "guided tours" might be found. This process may involve the use of NLP libraries or APIs (e.g., SpaCy or NLTK).

[1262] Step 5:

[1263] The server initializes the generative AI model and provides the extracted keywords as input data. The generative AI model (e.g., GPT) is used to generate text for native ads based on the keywords. For example, an ad copy such as "Special feature on Kyoto autumn foliage tours" is generated. Here, a pre-trained model is used for the generative AI.

[1264] Step 6:

[1265] The server adjusts the layout and style of the generated advertisement to match the design of the web page, specifically by applying a CSS stylesheet so that the advertisement blends into the existing design of the web page. This adjustment may include font size, color, margins, etc.

[1266] Step 7:

[1267] The server updates the HTML content, inserting the generated ad in the appropriate place (e.g., adding an ad after the third paragraph of a blog post), and restructures the entire HTML document to ensure consistency.

[1268] Step 8:

[1269] The server sends the updated HTML content to the device, and the device reads the new HTML data received from the server and redraws the web page.

[1270] Step 9:

[1271] Users can browse the updated web page and continue reading the original content without noticing the generated native ad, making the user experience natural and seamless.

[1272] Example 1

[1273] 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."

[1274] Conventional ad delivery methods often lack a proper blend of web page content and ads, creating a sense of discomfort for users. Furthermore, to maximize advertising effectiveness, it is necessary to generate and display ads in real time that are closely related to the content being viewed by users. However, the technology to accomplish this efficiently has not yet been fully developed.

[1275] 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.

[1276] In this invention, the server includes means for acquiring web page information from user web access, means for analyzing the web page information in real time and extracting main topics and keywords, means for generating advertisements using a generative AI model based on the extracted topics and keywords, means for adjusting the layout and style of the generated advertisement to match the design of the web page, means for inserting the generated advertisement into an appropriate position on the web page, and means for transmitting the updated web page to the terminal and redisplaying it, thereby enabling advertisements to be displayed in a way that does not cause discomfort to the user and maximizing advertising effectiveness.

[1277] "Web access" refers to the act of a user connecting to a web page via the Internet and obtaining information.

[1278] "Web page information" is a collection of data such as text, images, videos, and links displayed on a website.

[1279] "Real-time" refers to processing occurring immediately without delay.

[1280] "Analysis" refers to examining the content of data in detail, especially using natural language processing to extract specific information.

[1281] A "topic" refers to the main topic or subject matter in text data.

[1282] "Keywords" are important words or phrases in a document that help identify the topic.

[1283] A "generative AI model" refers to an algorithm or system that uses artificial intelligence to generate new content or data.

[1284] "Advertising" is information used to promote products and services to consumers.

[1285] "Layout" refers to the position and structure of each element arranged within a web page.

[1286] "Style" refers to the appearance and design of text and images on a web page, including color, font, size, etc.

[1287] "Insertion" means adding new information or elements into an existing Web page.

[1288] A "terminal" is a device that a user uses to access the Internet. Examples include personal computers, smartphones, and tablets.

[1289] "Redisplay" means displaying updated content or data again on the user's terminal.

[1290] This invention relates to a real-time native advertising generation system using generative AI. Specifically, it provides a system that analyzes the content of web pages accessed by users, dynamically generates advertisements that fit the content, and inserts them into the web pages, thereby enhancing advertising effectiveness without causing discomfort to users.

[1291] System Configuration

[1292] The system mainly consists of three main components: the user's device, the server, and the generative AI model.

[1293] 1. User's device

[1294] A device that allows users to access the web. The device has the ability to send requests to web servers and display the retrieved content. Devices include PCs, smartphones, tablets, etc.

[1295] 2. Server

[1296] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also generates advertisements using AI and processes them to insert them into web pages. It also has the function of sending updated web pages to users' devices.

[1297] 3. Generative AI Models

[1298] The generative AI model runs on a server and generates text ads based on the analyzed topics and keywords. Natural language processing (NLP) techniques are used for generative AI models. Specific techniques include generative models such as GPT-3.

[1299] Server Processing

[1300] When a user accesses a specific web page, the device requests the HTML content of that web page from the server. The server receives the request and retrieves the relevant HTML data. The server then analyzes the retrieved HTML content to extract key topics and keywords. For example, NLTK and spaCy are used to extract keywords such as "autumn leaves," "famous places," and "guided tours."

[1301] Next, the server inputs the extracted keywords into a generative AI model to generate advertising text related to the content. For example, an advertisement such as "Special feature on Kyoto autumn foliage tours" may be generated. An example of a prompt used in this case is "Please generate advertising text on the themes of 'Kyoto autumn foliage,' 'famous places,' and 'sightseeing.'"

[1302] The server then adjusts the generated ad text to match the style and layout of the webpage, using CSS and JavaScript to make the ad look consistent with the webpage. The server then inserts the ad in the appropriate position and regenerates the webpage's HTML, for example, inserting the ad after the second paragraph.

[1303] Finally, the server sends the updated HTML content to the user's device, which redisplays it, allowing the user to seamlessly browse the page.

[1304] Specific examples

[1305] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following will happen:

[1306] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[1307] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[1308] The server inputs these keywords into the generation AI, which generates advertising text such as "Special feature on Kyoto autumn foliage tours."

[1309] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[1310] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[1311] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1312] Step 1:

[1313] A user accesses a specific web page. Specifically, the user enters a URL in the browser's address bar or clicks a link. The URL is entered as input, and the device sends an HTTP request to the server requesting the HTML content corresponding to that URL. The server receives this request as output.

[1314] Step 2:

[1315] The server receives a user request and retrieves the relevant HTML content. Specifically, the server loads the HTML file corresponding to the requested URL from a database or file system. The input is the requested URL, and the output is the corresponding HTML content.

[1316] Step 3:

[1317] The server analyzes the retrieved HTML content and extracts key topics and keywords. Specifically, it uses a natural language processing (NLP) library to perform text analysis. For example, it uses "NLTK" or "spaCy." The input is the entire HTML content, and the output is extracted keywords such as "autumn leaves," "famous places," and "guided tours."

[1318] Step 4:

[1319] The server inputs the extracted keywords into the generative AI model and generates advertising text related to the content. Specifically, the server sends a prompt to the generative AI model. An example of this prompt is "Please generate advertising text on the themes of 'Kyoto autumn leaves,' 'famous places,' and 'sightseeing.'" The input is the extracted keywords and the prompt, and the output is advertising text such as "Special feature on Kyoto autumn leaves tours."

[1320] Step 5:

[1321] The server adjusts the style and layout of the generated ad text to match the web page design. Specifically, it uses CSS and JavaScript to adjust the font, color, size, etc. of the ad text to match the web page design. The input is the generated ad text, and the output is the adjusted ad text.

[1322] Step 6:

[1323] The server inserts the ad in the appropriate position and regenerates the HTML for the webpage, using JavaScript DOM manipulation and a server-side template engine. For example, inserting the ad after the second paragraph. The input is the adjusted ad text and the original HTML content, and the output is the updated HTML content with the ad inserted.

[1324] Step 7:

[1325] The server sends the updated HTML content to the user's device, which then redisplays it. Specifically, the newly generated HTML is returned to the device as an HTTP response. The input is the updated HTML content with the ad inserted, and the output is the updated web page displayed on the device. This allows the user to view the page seamlessly.

[1326] (Application example 1)

[1327] 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."

[1328] Modern internet advertising tends to be intrusive to users, often detracting from the user experience. Furthermore, static advertisements tend to be less relevant to the page content, reducing advertising effectiveness. Conventional advertising systems are unable to dynamically generate advertisements that match the content of web pages, which can be unsettling for users. The present invention aims to solve these problems.

[1329] 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.

[1330] In this invention, the server includes means for acquiring web page content through user web access, means for analyzing the web page content in real time and extracting key topics and keywords, and means for generating native advertisements using a generation AI based on the extracted topics and keywords. This makes it possible to improve the user experience while increasing advertising effectiveness by inserting dynamically generated advertisements in appropriate positions based on the content viewed by the user.

[1331] "User web access" refers to the act of a user accessing a web page via the Internet and obtaining or viewing information.

[1332] "Web page content" refers collectively to text, images, video, and other media or information displayed on a web page.

[1333] "Real-time analysis" is the process of analyzing the content of a web page immediately after a user accesses it.

[1334] "Major topics and keywords" refers to the important subjects and keywords in the content of a web page, which serve as the basis for generating advertisements.

[1335] "Generative AI" is a system that uses artificial intelligence techniques to generate new text based on specific prompts.

[1336] "Native advertising" is advertising that is displayed in a way that is integrated with the content of a web page and is designed to not cause any discomfort to the user.

[1337] A "dynamically generated advertisement based on content" is an advertisement that is automatically generated in real time according to the content of the web page that the user is viewing.

[1338] "Inserting in an appropriate position on a web page" refers to placing advertisements in appropriate locations on a web page so as not to impair the user experience.

[1339] "User experience" refers to the overall experience and satisfaction a user has when using a product or service.

[1340] An "application" is a software program that runs on a device and provides a particular function or service.

[1341] This invention is a system that uses generative AI to generate native ads in real time and insert them into the content of web pages accessed by users. This article mainly explains an example of applying this system to a news application for smartphones.

[1342] System Configuration

[1343] The system consists of three main components: the user's device, a server, and a generative AI model.

[1344] 1. User's device

[1345] This includes devices such as smartphones and tablets.

[1346] It provides an interface for users to access articles through a news app.

[1347] 2. Server

[1348] It plays a central role in retrieving and parsing the HTML content of web pages.

[1349] Generative AI is used to generate advertisements, and processing is performed to properly insert the advertisements into web pages.

[1350] 3. Generative AI Models

[1351] Natural language processing (NLP) techniques are used to generate text ads based on analyzed topics and keywords.

[1352] Generative AI models used include, for example, OpenAI's GPT-4.

[1353] Program processing overview

[1354] When a user accesses a web page, the server performs the following process.

[1355] 1. User Web Access

[1356] A user clicks on a particular article in a news app, and the device requests the HTML content of that news article from the server.

[1357] 2. Getting HTML data

[1358] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[1359] 3. Content Analysis

[1360] The server analyzes the HTML content and extracts key topics and keywords, such as "climate change," "renewable energy," and "eco-activities."

[1361] 4. Ad generation

[1362] The server inputs the extracted keywords into a generative AI model to generate advertising text related to the content. An example of a prompt sentence to be used is as follows:

[1363] Generate ad text based on the following keywords: climate change, renewable energy, eco-activism

[1364] 5. Ad Styling

[1365] The server adjusts the style and layout of the generated advertising text to match the design of the news article.

[1366] 6. Ad Insertion

[1367] The server inserts the advertisement at the appropriate position in the news article, for example after the second paragraph of the article.

[1368] 7. Submitting Updated Content

[1369] The server sends the updated HTML content to the user's device, which redisplays it.

[1370] Specific examples

[1371] For example, if a user accesses an article on the topic of "Thinking about Environmental Issues" in a news app, the following steps are taken:

[1372] The user clicks on the URL of a news article, and the device requests the HTML content from the server.

[1373] The server retrieves the HTML content, analyzes it, and extracts the keywords "climate change," "renewable energy," and "eco-activities."

[1374] The server inputs these keywords into a generation AI, which generates ad text such as "Introducing a new electric vehicle."

[1375] The server adapts the generated advertisements to the style of the news article and inserts them at the appropriate locations.

[1376] The updated page will be redisplayed on the user's device, allowing the user to continue reading the article without any discomfort.

[1377] This invention makes it possible to maximize advertising effectiveness by dynamically generating and inserting advertisements that match the content without impairing the user experience.

[1378] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1379] Step 1:

[1380] A user clicks on a specific article in a news app.

[1381] Input: The URL of the news article selected by the user.

[1382] Specific operation: The device sends a request to the server for the HTML content of the news article.

[1383] Output: The server receives the request.

[1384] Step 2:

[1385] The server receives the user's request and retrieves the relevant HTML data from a database or file system.

[1386] Input: A news article URL request.

[1387] Specific operation: The server searches for and retrieves the corresponding HTML content from the database.

[1388] Output: The retrieved HTML content.

[1389] Step 3:

[1390] The server parses the retrieved HTML content and extracts key topics and keywords.

[1391] Input: The retrieved HTML content.

[1392] What it does: The server uses natural language processing (NLP) techniques to parse the HTML content and extract key topics and keywords (e.g., "climate change," "renewable energy," "green activities").

[1393] Output: Extracted main topics and keywords.

[1394] Step 4:

[1395] The server inputs the extracted keywords into a generative AI model to generate advertising text relevant to the content.

[1396] Input: Extracted main topics and keywords.

[1397] Specific operation: The server sends the following prompt to the generated AI model:

[1398] Generate ad text based on the following keywords: climate change, renewable energy, eco-activism

[1399] A generative AI model generates ad text.

[1400] Output: The generated ad text.

[1401] Step 5:

[1402] The server adjusts the style and layout of the generated advertising text to match the design of the news article.

[1403] Input: The generated ad text.

[1404] What it does: The server uses CSS and HTML templates to adjust the ad design and integrate it into the style of the news article.

[1405] Output: The adjusted ad text.

[1406] Step 6:

[1407] The server inserts the advertisements into the news articles at appropriate locations.

[1408] Input: Your adjusted ad text.

[1409] Specific behavior: The server regenerates the HTML content of the news article, for example inserting an advertisement after the second paragraph of the article.

[1410] Output: The updated HTML content.

[1411] Step 7:

[1412] The server sends the updated HTML content to the user's device, and the user redisplays it through the news app.

[1413] Input: The updated HTML content.

[1414] What happens: The server sends HTML content to the device, and the news app receives and displays it.

[1415] Output: The user can view the updated news article.

[1416] These steps ensure that users are presented with ads that are tailored to their content and that fit their content seamlessly, improving the user experience.

[1417] 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.

[1418] This invention relates to a system that recognizes user emotions and automatically inserts native ads generated based on those emotions into web pages. Specifically, the system analyzes the content of web pages retrieved by users' web access in real time and extracts key topics and keywords. It also uses an emotion engine to recognize user emotions and generates native ads based on this emotion using a generation AI. The generated native ads are inserted in appropriate positions on the web page, and the updated web page is sent to the terminal. This system provides ads with optimal timing and content to users, achieving higher advertising effectiveness.

[1419] Specific Embodiments of the System

[1420] System Configuration

[1421] This system mainly consists of four main components: the user's device, the server, the generative AI model, and the emotion engine.

[1422] 1. User's device

[1423] A device that allows users to access the web. The terminal has the function of sending requests to a web server and displaying the retrieved content. It also has the function of collecting data such as the user's facial expressions, voice tone, and text input, and sending it to the server.

[1424] 2. Server

[1425] The server plays a central role in acquiring and analyzing the HTML content of web pages. It also analyzes user emotion data, generates advertisements based on the emotions using generative AI, and processes the advertisements to be inserted into web pages. It also has the function of sending updated web pages to users' devices.

[1426] 3. Generative AI Models

[1427] The generative AI model runs on a server and generates text ads based on the analyzed topics and keywords, and also takes into account user sentiment data to generate ads with the appropriate tone and content.

[1428] 4. Emotion Engine

[1429] The emotion engine analyzes users' facial expressions, voice tone, and text input to recognize their emotional state in real time. This emotional data is fed into a generative AI model and used to generate ads.

[1430] Program processing

[1431] When a user accesses a web page, the following sequence of events occurs:

[1432] 1. User Web Access

[1433] When a user accesses a specific web page, for example, a blog post about "Autumn Kyoto Sightseeing," the device sends a request to the URL, requesting the HTML content of the web page from the server.

[1434] 2. Getting HTML data

[1435] The server receives the user's request and retrieves the relevant HTML data from a database or file system. The web server then prepares a response containing all the data for the web page (text, images, style sheets, etc.).

[1436] 3. Content Analysis

[1437] The server analyzes the HTML retrieved. Specifically, it uses an HTML parser to analyze the structure of the web page and extract the text content. For example, it extracts the entire body of an article titled "Sightseeing in Kyoto in Autumn."

[1438] 4. Topic and Keyword Extraction

[1439] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the extracted text. For example, keywords such as "autumn leaves," "famous places," and "guided tours" might be found. This process may involve the use of NLP libraries or APIs (e.g., SpaCy or NLTK).

[1440] 5. Emotional Recognition

[1441] While a user is browsing a web page, the device collects data such as facial expressions and voice and sends it to the emotion engine. The emotion engine analyzes this data and identifies the user's emotional state. For example, if the user is smiling, it is judged to be "happy," and if they are frowning, it is judged to be "unhappy."

[1442] 6. Ad Generation

[1443] The server inputs the extracted keywords and recognized emotion data into a generative AI model to generate native ads that match the emotion. For example, if the user is in a "happy" state, it generates an ad with a positive tone, and if the user is in a "dissatisfied" state, it generates a problem-solving ad.

[1444] 7. Ad Styling

[1445] The server adjusts the layout and style of the generated advertisement to match the design of the web page, specifically by applying a CSS stylesheet so that the advertisement blends into the existing design of the web page. This adjustment may include font size, color, margins, etc.

[1446] 8. Ad Insertion

[1447] The server updates the HTML content, inserting the generated ad in the appropriate place (e.g., adding an ad after the third paragraph of a blog post), and restructures the entire HTML document to ensure consistency.

[1448] 9. Submitting Updated Content

[1449] The server sends the updated HTML content to the device, and the device reads the new HTML data received from the server and redraws the web page.

[1450] Specific examples

[1451] For example, if a user visits a blog page with the theme "Autumn Kyoto Sightseeing," the following happens:

[1452] The user clicks on the URL of the web page, and the terminal requests the HTML content from the server.

[1453] The server retrieves the HTML content, analyzes it, and extracts the keywords "autumn leaves," "famous places," and "sightseeing."

[1454] The server inputs these keywords into the AI ​​generator, which generates ad text such as "Kyoto Autumn Foliage Tour Special." It also generates ad text with a positive tone based on the user's emotional data (e.g., "happiness").

[1455] The server adapts the generated advertisement to the style of the web page and inserts it at the appropriate location.

[1456] The terminal redisplays the updated page, allowing the user to seamlessly view the page.

[1457] In this way, the system provides advertisements that best suit the user's emotional state, maximizing advertising effectiveness and improving the user experience.

[1458] The processing flow will be explained below.

[1459] Step 1:

[1460] A user clicks on a URL in their web browser to access a blog post about "Autumn Kyoto Sightseeing." The device sends a request to the server for that URL, requesting the HTML content of the web page.

[1461] Step 2:

[1462] The server receives the user's request and retrieves the appropriate HTML content from a database or file system.

[1463] Step 3:

[1464] The HTML content retrieved by the server is analyzed using an HTML parser, extracting each element of the web page, such as text content, images, and style sheets.

[1465] Step 4:

[1466] The server uses natural language processing (NLP) techniques to analyze and extract key topics and keywords from the text content, such as "autumn leaves," "famous places," and "guided tours."

[1467] Step 5:

[1468] While a user is browsing a web page, the device uses a camera and microphone to transmit the user's facial expressions and voice tone to the emotion engine, which analyzes this data and recognizes the user's emotional state. For example, a smile on the user's face indicates "happiness," while a frown indicates "dissatisfaction."

[1469] Step 6:

[1470] The server inputs the extracted keywords and recognized emotion data into a generative AI model. The generative AI generates native ads based on the keywords while adjusting the tone and content of the ads based on the emotion data. For example, if the user is in a "happy" state, it generates a positive-toned ad for a "Kyoto Autumn Foliage Tour Special Campaign."

[1471] Step 7:

[1472] The server adjusts the layout and style of the generated ad to match the design of the web page, for example by using CSS to adjust the font size, color, and margins of the ad so that it blends seamlessly into the web page.

[1473] Step 8:

[1474] The server generates the updated HTML content and inserts the ad in the appropriate position (e.g. after the third paragraph of the article).

[1475] Step 9:

[1476] The server sends the updated web page to the device, which reads the new HTML data and redraws the web page.

[1477] Step 10:

[1478] When users browse the updated web page, they will notice the generated native ads and can continue reading the page content naturally. Users will receive ads that match their emotions, resulting in a more comfortable browsing experience.

[1479] Example 2

[1480] 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."

[1481] Current ad delivery systems struggle to provide effective ads that take into account the user's emotional state. Furthermore, the inconsistency between the content of a web page and the ads often leads to a poor user experience. Therefore, a system is needed that can generate and display ads in real time that are optimized for the user's emotions and the content of a web page.

[1482] 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.

[1483] In this invention, the server includes means for acquiring web page content through user web access, means for analyzing the web page content in real time and extracting key topics and keywords, means for collecting data such as the user's facial expressions, voice tone, and text input and recognizing emotions, means for generating a native advertisement using a generation AI based on the extracted topics and keywords and the recognized emotion data, means for adjusting the layout and style of the generated native advertisement to match the design of the web page, means for inserting the generated native advertisement at an appropriate position on the web page, and means for transmitting the updated web page to a terminal and redisplaying it. This makes it possible to generate and display an advertisement optimized for the user's emotional state in real time.

[1484] "Web access" refers to a user accessing a specific web page via the Internet.

[1485] "Web page content" refers to all data elements contained within a Web page, including text, images, style sheets, scripts, etc.

[1486] "Real-time" refers to data processing and analysis occurring almost immediately.

[1487] "Topic" refers to the main theme that describes the content or subject matter of a Web page.

[1488] "Keywords" refer to important words or phrases that represent the content of a web page.

[1489] "Facial expression" refers to facial changes and characteristics that represent the user's emotional state.

[1490] "Voice tone" refers to the emotional inflections and nuances contained in a user's voice.

[1491] "Text input" refers to text data that a user inputs on a web page.

[1492] "Emotion recognition" refers to analyzing and identifying a user's emotional state from their facial expressions, voice, and text.

[1493] "Generative AI" refers to algorithms or models that use artificial intelligence techniques to generate new content (e.g., advertisements) based on specific input data.

[1494] "Native advertising" refers to advertising that blends naturally into the content of a web page.

[1495] "Layout" refers to the presentation and structure of a web page or advertisement.

[1496] "Style" refers to the design elements of a web page or advertisement (e.g., fonts, colors, spacing).

[1497] A "server" refers to a computer system that provides information and services over a network.

[1498] "Terminal" refers to a device, such as a computer or smartphone, that a user uses to access the Internet.

[1499] "Redisplay" refers to displaying the updated web page on the screen again.

[1500] This invention relates to a system that recognizes user emotions and automatically inserts native ads into web pages based on those emotions. The system mainly consists of four main components: a user device, a server, a generative AI model, and an emotion engine.

[1501] System Configuration and Hardware / Software Usage

[1502] 1. User's device

[1503] A device that allows users to access the web. The terminal has the function of sending requests to a web server and displaying the retrieved content. It also has the function of collecting data such as the user's facial expressions, voice tone, and text input and sending it to the server. This device is equipped with input devices such as a camera and microphone.

[1504] 2. Server

[1505] The server plays a central role in retrieving and analyzing the HTML content of web pages. The server retrieves HTML data using web server software such as Apache or Nginx, and parses the content using an HTML parser library such as BeautifulSoup. It also has the ability to analyze user emotion data and generate advertisements using generative AI based on the emotion. It also has the ability to generate HTTP responses to send updated web pages to users' devices.

[1506] 3. Generative AI Models

[1507] The generative AI model runs on the server and generates text ads based on the analyzed topics and keywords. This model, such as OpenAI's GPT-3, generates appropriate ads by inputting a prompt. An example of a prompt is "Please create an ad for an autumn foliage tour with a happy tone."

[1508] 4. Emotion Engine

[1509] The emotion engine analyzes the user's facial expressions, voice tone, and text input to recognize the user's emotional state in real time. The collected emotion data is sent to a server and input into a generative AI model. This emotion engine may utilize, for example, Microsoft's Azure Emotion API or Google's Cloud Vision API.

[1510] Specific examples

[1511] As a concrete example, consider the case where a user accesses a blog page with the theme "Autumn Kyoto Sightseeing." The device collects the user's facial expressions and voice tone and sends them to the emotion engine. As a result, if the user is smiling, it is determined to be "happy." This emotional state, along with the topic and keywords of the blog page (e.g., "autumn leaves," "famous places," and "sightseeing"), are input into a generative AI model to generate advertising text with a positive tone, such as "Special feature on Kyoto autumn leaves tours."

[1512] An example of a prompt for the generative AI model is "Please create an advertisement for an autumn foliage tour in a happy tone." The server adjusts the layout and style of the generated advertisement using a CSS stylesheet to blend in with the existing design of the webpage, and inserts it in the appropriate position after the third paragraph.

[1513] The server then sends the updated HTML content to the user's device, which quickly redraws it, allowing the user to smoothly view the updated web page.

[1514] This invention makes it possible to generate and display advertisements optimized for the user's emotional state in real time, thereby maximizing advertising effectiveness and providing a superior user experience.

[1515] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1516] Step 1:

[1517] A user visits a web page. The user sends an HTTP request for the URL of a particular web page (e.g., "Autumn Kyoto Sightseeing"). The input is the URL of the web page, and the output is that the request is sent to the server.

[1518] Step 2:

[1519] The server retrieves the HTML data. The server analyzes the received HTTP request and retrieves the HTML data for the corresponding web page from a database or file system. The input is the URL of the web page, and the specific operation is for web server software such as Apache or Nginx to load the HTML file. The output is the HTML data for the web page.

[1520] Step 3:

[1521] The server parses the HTML data. It uses an HTML parser library such as BeautifulSoup to analyze the structure of the web page and extract the text content. The input is the HTML data of the web page, and the specific operation is to parse the HTML structure. The output is the parsed text content.

[1522] Step 4:

[1523] The server extracts key topics and keywords. The server uses natural language processing (NLP) techniques, specifically SpaCy or NLTK, to extract keywords from the text. The input is the parsed text content, and the specific operation is NLP analysis. The output is key topics and keywords (e.g., "autumn leaves," "famous places," "sightseeing").

[1524] Step 5:

[1525] Collecting user emotional data. The device uses the built-in camera and microphone to collect data such as the user's facial expressions, voice tone, and text input. The input is the user's facial expressions and voice, and the output is the collected emotional data.

[1526] Step 6:

[1527] The emotion engine analyzes the emotional data. The emotion engine analyzes the collected data in real time and identifies the user's emotional state. Specifically, it uses Microsoft's Azure Emotion API or Google's Cloud Vision API. The input is emotional data, and the specific action is analysis via the API. The output is the user's emotional state (e.g., "happy," "unhappy").

[1528] Step 7:

[1529] The server generates advertisements using a generative AI model. The server inputs the extracted keywords and recognized emotion data into the generative AI model (e.g., OpenAI's GPT-3). The advertisement is generated using a prompt (e.g., "Please create an advertisement for an autumn foliage tour in a happy tone"). The input is keywords and emotion data, and the prompt is input into the generative AI model as a specific action. The output is the generated advertisement text.

[1530] Step 8:

[1531] The server adjusts the layout and style of the ad. It adjusts the layout and style using a CSS stylesheet so that the generated ad matches the design of the webpage. The input is the generated ad text, and the specific action is to apply the CSS stylesheet. The output is the adjusted ad layout.

[1532] Step 9:

[1533] The server inserts the advertisement into the web page. The generated advertisement is inserted into the updated HTML content at the appropriate position. For example, an advertisement is added after the third paragraph of a blog post. The input is the adjusted advertisement layout and HTML content, and the specific operation is to reconstruct the HTML document. The output is the updated HTML content.

[1534] Step 10:

[1535] The server sends the updated web page to the device. The device receives the updated content and redraws the web page. The input is the updated HTML content, which is sent to the device as an HTTP response as a specific action. The output is the updated web page displayed on the device.

[1536] This allows the user to view advertisements that match their emotions seamlessly along with the displayed web page.

[1537] (Application example 2)

[1538] 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."

[1539] Current virtual stores display uniform advertisements without considering the user's emotional state, resulting in a poor user experience and limited advertising effectiveness. Furthermore, it is difficult to accurately suggest information and products that users desire in real time, and there is a lack of technology to seamlessly provide advertisements tailored to individual users. This results in lower ad click rates and lower purchasing intent, preventing advertisers from implementing effective marketing strategies.

[1540] 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.

[1541] In this invention, the server includes means for acquiring virtual space content through user access, means for analyzing the virtual space content in real time and extracting main themes and related words, means for recognizing emotions from the user's facial expressions and voice, means for generating native advertisements using a generation AI based on the extracted themes and related words and the recognized emotions, means for inserting the generated native advertisements into appropriate positions in the virtual space, and means for transmitting the updated virtual space to the terminal and redisplaying it, thereby making it possible to provide advertisements optimized for the user's emotional state in real time.

[1542] "User" refers to a person who uses a virtual space or system.

[1543] "Access" refers to the action of a user connecting to a specific virtual space or web page.

[1544] "Virtual space" refers to a space that is generated using digital technology and that can be visually experienced but is different from a real location.

[1545] "Content" refers to information such as text, images, video, and audio displayed in virtual spaces and on web pages.

[1546] "Real-time analysis" refers to processing data immediately in response to user operations or access.

[1547] "Major subject" refers to a topic that is considered particularly important within a document or piece of content.

[1548] "Related words" refer to words that are related to the main topic and provide additional information.

[1549] "Facial expression" refers to an element used to infer emotions by analyzing the user's facial movements and features.

[1550] "Voice" refers to the element used to analyze the user's voice to read their emotions and intentions.

[1551] "Emotion recognition" refers to the process of analyzing and identifying a person's emotional state from their facial expressions and voice.

[1552] "Generative AI" refers to artificial intelligence that uses machine learning models to automatically generate content and advertisements.

[1553] "Native advertising" refers to advertising that blends naturally into the format and design of the medium in which it is displayed.

[1554] "Appropriate location" refers to a location where the advertisement is displayed effectively and without interfering with the user's experience.

[1555] "Terminal" refers to a hardware device that allows a user to access a system or virtual space.

[1556] This invention relates to a system that recognizes user emotions in real time and displays appropriate native advertisements in a virtual store based on those emotions. This system mainly consists of four main components: a user device, a server, a generative AI model, and an emotion engine.

[1557] System Configuration

[1558] User's device

[1559] Users access the virtual store using devices such as smartphones and head-mounted displays (HMDs). These devices have the ability to collect the user's facial expressions and voice data and send it to a server.

[1560] server

[1561] The server retrieves the virtual store's content and performs real-time analysis. It also extracts key themes and related terms and analyzes user sentiment data. It then uses a generative AI model to generate native ads based on the extracted data and inserts them into the virtual store at appropriate locations. Finally, it sends the updated virtual store data to the user's device.

[1562] Generative AI Models

[1563] The generative AI model generates text ads based on the extracted subject and related words, as well as user sentiment data. This model uses advanced language generation techniques such as GPT-4. An example prompt is as follows:

[1564] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[1565] Emotion Engine

[1566] The emotion engine analyzes the user's facial expressions and voice to recognize emotions in real time. It uses emotion analysis technologies such as the Emotion SDK to identify the user's emotional state.

[1567] Processing flow

[1568] When a user accesses a virtual store using a smartphone or HMD, the device captures facial expression and voice data and sends them to the server, which then retrieves the content of the virtual store and analyzes it using a natural language processing library (e.g., SpaCy or NLTK) to extract key themes and related terms.

[1569] In parallel, the emotion engine recognizes the user's emotions and sends the data to the server, which then inputs the extracted subject and related words, as well as the emotion data, into a generative AI model to generate native ads. The generated ads are then adjusted to fit the design of the virtual store and inserted at the appropriate locations.

[1570] The updated virtual store data is sent from the server to the user's device and instantly redisplayed, allowing the user to seamlessly browse the virtual store and experience emotionally optimized advertising.

[1571] As a specific example, if a user accesses a virtual store on their smartphone and smiles while browsing the "latest gadgets" section, the system will recognize the user's emotion as "happiness" and send the following prompt to the generative AI model:

[1572] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[1573] The generated ad text is optimized to match the user's emotional state, such as "We're having a special sale on the latest gadgets right now! Get the latest technology that will make you smile at a great price!" This improves the user experience and maximizes the effectiveness of the ad.

[1574] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1575] Step 1:

[1576] A user accesses a virtual store using a device (smartphone or HMD). The device sends an access request to the server, requesting content from the virtual store. The input is the user's access request, and the output is a request to the server.

[1577] Step 2:

[1578] The server receives the access request and retrieves the content of the virtual store from a database or file system. This content can include text, images, videos, etc. The input is the request, and the output is the content of the virtual store.

[1579] Step 3:

[1580] The device captures the user's facial expressions and voice data and transmits this data to the server in real time. The input is the user's facial expressions and voice data, and the output is data transmission to the server. Specifically, data is collected from the smartphone camera and microphone and transmitted in real time.

[1581] Step 4:

[1582] The server analyzes the received content in real time using natural language processing libraries (e.g., SpaCy or NLTK) to extract key themes and related terms from the text data. The input is the content of the virtual store, and the output is the extracted themes and related terms.

[1583] Step 5:

[1584] The server uses an emotion engine to analyze the transmitted facial expressions and voice data and recognize the user's emotions. It uses emotion analysis technologies such as Emotion SDK. The input is the user's facial expressions and voice data, and the output is the recognized emotional state.

[1585] Step 6:

[1586] The server inputs the extracted topic and related words, as well as the recognized emotion data, into a generative AI model. Specifically, the generative AI model (such as GPT-4) generates a text ad based on the prompt sentence. The input is the topic, related words, and emotion data, and the output is the generated ad text. An example of a prompt sentence is as follows:

[1587] "Users are happy. They're interested in the latest gadgets, deals, and recommendations. Create ads with a positive tone."

[1588] Step 7:

[1589] The server adjusts the layout and style of the generated advertisement to match the design of the virtual store, including font size, color, margins, etc. using a CSS stylesheet. The input is the generated advertisement text, and the output is the adjusted advertisement.

[1590] Step 8:

[1591] The server inserts the adjusted advertisement into the appropriate location in the virtual store, for example, at the edge of the display where the user is looking or next to a product shelf. The input is the adjusted advertisement, and the output is the updated virtual store content.

[1592] Step 9:

[1593] The server sends the updated virtual store data to the user's terminal. The terminal receives this data and immediately redraws it. The input is the updated virtual store content, and the output is the virtual store redisplayed on the terminal.

[1594] This series of processes allows users to have a seamless shopping experience in a virtual store and view advertisements that are optimal for their emotions.

[1595] 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.

[1596] 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.

[1597] 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.

[1598] 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.

[1599] FIG. 9 illustrates 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 behaviors 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.

[1600] 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.

[1601] 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).

[1602] 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.

[1603] 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."

[1604] 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.

[1605] 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).

[1606] 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.

[1607] 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.

[1608] 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.

[1609] 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.

[1610] The hardware resource for executing a specific process can be any of the following types of processors: A processor, for example, 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 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.

[1611] 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.

[1612] 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.

[1613] 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.

[1614] 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.

[1615] 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.

[1616] The following is further disclosed regarding the above embodiment.

[1617] (Claim 1)

[1618] A means for acquiring web page content through a user's web access;

[1619] means for analyzing web page content in real time and extracting key topics and keywords;

[1620] A means for generating native advertisements using a generative AI based on the extracted topics and keywords;

[1621] means for inserting the generated native advertisement into an appropriate position on a web page;

[1622] A means of sending updated web pages to the device and redisplaying them.

[1623] A system including:

[1624] (Claim 2)

[1625] 10. The system of claim 1, wherein natural language processing is used to extract key topics and keywords.

[1626] (Claim 3)

[1627] 10. The system of claim 1, wherein the layout and style of the generated native advertisement is adjusted to match the design of the web page.

[1628] "Example 1"

[1629] (Claim 1)

[1630] A means for acquiring web page information through a user's web access;

[1631] means for analyzing web page information in real time and extracting key topics and keywords;

[1632] means for generating advertisements using a generative AI model based on the extracted topics and keywords;

[1633] means for inserting the generated advertisements into appropriate locations on the web page;

[1634] A means of sending updated web pages to the device and redisplaying them.

[1635] A system including:

[1636] (Claim 2)

[1637] 10. The system of claim 1, wherein natural language processing is used to extract key topics and keywords.

[1638] (Claim 3)

[1639] 10. The system of claim 1, wherein the layout and style of the generated advertisement is adjusted to match the design of the web page.

[1640] "Application Example 1"

[1641] (Claim 1)

[1642] A means for acquiring web page content through a user's web access;

[1643] means for analyzing web page content in real time and extracting key topics and keywords;

[1644] A means for generating native advertisements using a generative AI based on the extracted topics and keywords;

[1645] means for inserting the generated native advertisement into an appropriate position on a web page;

[1646] means for transmitting the updated web page to the terminal and redisplaying it;

[1647] A means for having an application that enhances user experience by inserting dynamically generated advertisements at appropriate locations based on the content viewed by the user;

[1648] A system including:

[1649] (Claim 2)

[1650] 10. The system of claim 1, wherein natural language processing is used to extract key topics and keywords.

[1651] (Claim 3)

[1652] 10. The system of claim 1, wherein the layout and style of the generated native advertisement is adjusted to match the design of the web page.

[1653] "Example 2: Combining Emotion Engines"

[1654] (Claim 1)

[1655] A means for acquiring web page content through a user's web access;

[1656] means for analyzing web page content in real time and extracting key topics and keywords;

[1657] A means for collecting data such as a user's facial expression, voice tone, and text input to recognize emotions;

[1658] a means for generating native advertisements using generative AI based on the extracted topics and keywords and the recognized sentiment data;

[1659] means for adjusting the layout and style of the generated native advertisement to match the design of the web page;

[1660] means for inserting the generated native advertisement into an appropriate position on a web page;

[1661] A means of sending updated web pages to the device and redisplaying them.

[1662] A system including:

[1663] (Claim 2)

[1664] 10. The system of claim 1, wherein natural language processing is used to extract key topics and keywords.

[1665] (Claim 3)

[1666] 10. The system of claim 1, wherein the system analyzes data such as facial expressions, voice tone, and text input to recognize user emotions in real time.

[1667] "Application example 2 when combining emotion engines"

[1668] (Claim 1)

[1669] A means for acquiring content of the virtual space through user access;

[1670] means for analyzing the content of the virtual space in real time and extracting key themes and related terms;

[1671] means for recognizing emotions from the user's facial expressions and voice;

[1672] a means for generating native advertisements using generative AI based on the extracted themes and related terms and the recognized sentiment;

[1673] A means for inserting the generated native advertisement at an appropriate position in the virtual space;

[1674] A means to send the updated virtual space to the terminal and redisplay it.

[1675] A system including:

[1676] (Claim 2)

[1677] 10. The system of claim 1, wherein natural language processing is used to extract key themes and related terms.

[1678] (Claim 3)

[1679] 10. The system of claim 1, wherein the layout and style of the generated native advertisement is adjusted to match the design of the virtual space. [Explanation of symbols]

[1680] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for acquiring web page content through a user's web access; means for analyzing web page content in real time and extracting key topics and keywords; A means for generating native advertisements using a generative AI based on the extracted topics and keywords; means for inserting the generated native advertisement into an appropriate position on a web page; A means of sending updated web pages to the device and redisplaying them. A system including:

2. 10. The system of claim 1, wherein natural language processing is used to extract key topics and keywords.

3. The system of claim 1 , wherein the layout and style of the generated native advertisement is adjusted to match the design of the web page.

Images