system

Abstract

A system is provided.SOLUTION: A system, comprising: means for crawling a plurality of web pages on the Internet and analyzing contents of the pages; means for identifying web pages containing sexual advertisements based on the analyzed contents and storing URLs of these pages in a database; means for receiving a URL of a web page to be accessed by a user and determining whether the URL is stored in the database; and means for summarizing the contents of the web pages containing sexual advertisements.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] The problem on the Internet is that inappropriate advertisements, including sexual content, are frequently displayed even on web pages intended for all ages. This poses a risk to users, especially children, who may be exposed to offensive advertisements. Existing countermeasures are insufficient and difficult to completely eliminate such advertisements. In this situation, there is a need for a system that allows users to access the information they desire without being exposed to offensive advertisements. Furthermore, effective means are needed to promote the integrity of Internet advertising and weed out inappropriate advertisements. [Means for solving the problem]

[0005] This invention provides a means for crawling multiple web pages on the Internet and analyzing the content of those pages. It also provides a means for identifying web pages containing sexually explicit advertisements based on the analyzed content and storing the URLs of those pages in a database. It also uses a means for receiving the URL of a web page a user attempts to access and determining whether the URL is stored in the database. It also implements a means for summarizing the content of web pages containing sexually explicit advertisements, sending the summarized content to the user's device and displaying it as a pop-up window, allowing the user to obtain the desired information without being exposed to offensive advertisements. This configuration is expected to promote the soundness of Internet advertising and eliminate inappropriate advertisements.

[0006] "Crawling" is the process of automatically collecting multiple web pages on the web and analyzing their content.

[0007] "Analysis" is the process of evaluating collected web page source code, metadata, and advertising banner content to identify certain characteristics.

[0008] "Sexual advertising" refers to advertising banners or text ads that contain sexual language or content.

[0009] "Identification" is the act of identifying or classifying objects based on specific patterns or characteristics.

[0010] A "database" is a collection of digital data that is used to systematically organize and store information.

[0011] A "URL" is a unique address that identifies and locates a web page or resource.

[0012] "Judgment" is the act of determining whether an object belongs to a particular category based on specific conditions or criteria.

[0013] "Summarizing" refers to condensing the content of a longer text into a short, to-the-point form.

[0014] A "natural language processing model" is an algorithm or method for understanding and generating human language using machine learning.

[0015] A "pop-up window" is a small window that appears on the screen in response to a user operation or certain conditions.

[0016] A "browser extension" is a software component that adds or extends the functionality of a web browser. [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] Overall program overview

[0039] This invention relates to a system for securely acquiring the contents of web pages on the Internet. This system involves the cooperation of a server, a terminal, and a user, and aims to restrict access to web pages that contain sexually explicit advertisements.

[0040] Server-side processing

[0041] The server first crawls multiple web pages on the Internet. It analyzes the content of the crawled web pages to identify whether they contain sexually explicit advertisements. The URLs of the identified web pages are stored in a database. When analyzing this content, the server uses natural language processing and image recognition technologies to evaluate the content of the advertisements.

[0042] The server then uses the massive dataset to train a natural language processing model. The trained AI model is then used to summarize the content of the web page. This summarization model is designed to quickly generate a summary of the web page in question. The summarized information is also stored in a database, ready to be served immediately if needed.

[0043] Terminal side processing

[0044] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. The extension sends the URL to a server, which receives the URL and checks it against a list of problem pages stored in a database. If a problem is detected, a summary is sent from the server to the device.

[0045] The device displays a pop-up window to provide summary information to the user. The pop-up window displays, "This page contains inappropriate advertisements. Summary of page contents: (Summary text)." Based on this summary information, the user can obtain the necessary information without being exposed to unpleasant advertisements.

[0046] User processing

[0047] Users can view web page summary information in their browsers and use this information to decide whether to visit a particular web page. This allows users to easily access the information they want while avoiding pages that contain sexually explicit advertisements.

[0048] Specific examples

[0049] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." A browser extension sends this URL to a server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model summarizes the content of this page and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..."

[0050] This summary information is sent to the device and displayed as a pop-up window by the browser extension. The user reads the summary and feels that they have obtained the desired information without being exposed to annoying advertisements. In this way, the system provides users with access to the information they need while preventing them from being exposed to unwanted advertisements.

[0051] This system configuration promotes the soundness of Internet advertising and provides users with a safe and comfortable web browsing environment.

[0052] The processing flow will be explained below.

[0053] Step 1:

[0054] Server: Crawl web pages

[0055] The server automatically collects pages on the Internet, crawls new pages based on a periodically updated URL list, obtains the HTML source of the crawled pages, and stores it for content analysis.

[0056] Step 2:

[0057] Server: Content analysis and identification

[0058] The server analyzes the content of the crawled pages, using natural language processing and image recognition technology to examine the ad banners and metadata on the pages, determining whether they contain sexually explicit content and storing the URLs of problematic pages in a database.

[0059] Step 3:

[0060] Server: Training the natural language processing model

[0061] The server trains a natural language processing model (such as BERT or GPT-3) using a large amount of text data. The training data includes high-quality summaries and feedback. This model is then used to summarize the content of web pages.

[0062] Step 4:

[0063] Terminal: Monitoring web page requests

[0064] A browser extension installed on the user's device monitors the URLs the user attempts to access and sends this URL information to a server.

[0065] Step 5:

[0066] Server: URL rating

[0067] The server checks the received URL against a database to see if the URL exists and whether it is a page containing sexual advertisements.

[0068] Step 6:

[0069] Server: Generate summary information

[0070] If the page is determined to be relevant, the server uses a natural language processing model to summarize the content of the page, generates this summary information, and sends it to the terminal.

[0071] Step 7:

[0072] Terminal: Display summary information

[0073] The browser extension displays the received summary information in a popup window, displaying the message "This page contains inappropriate advertising. Summary of page contents: (summary text)" to the user.

[0074] Step 8:

[0075] User: Review summary information and make a decision

[0076] The user reviews the summary information provided in the pop-up window, allowing them to decide whether to obtain the information they need without being exposed to the intrusive advertisement.

[0077] Step 9:

[0078] User: Safe Web Browsing

[0079] Users can safely obtain the information they are looking for while avoiding inappropriate ads, improving the user experience and promoting the integrity of internet advertising.

[0080] Example 1

[0081] 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."

[0082] On the Internet, there are web pages containing inappropriate content, and accessing them can lead to unpleasant experiences or expose users to harmful information. Web pages containing sexually explicit advertisements are particularly undesirable for users. To address this, it is necessary to provide an environment in which users can browse the web with peace of mind. However, existing filtering systems have the problem of being unable to completely filter out harmful content.

[0083] 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.

[0084] In this invention, the server includes means for crawling multiple web pages on the Internet and analyzing the contents of those pages, means for identifying sexual advertisements based on the analyzed contents and saving the addresses of those pages in data storage means, means for receiving the addresses of web pages that a user wishes to access and determining whether the addresses are saved in the data storage means, means for summarizing the contents of web pages that include sexual advertisements using natural language processing, and means for transmitting the summarized contents to the user's information processing device and displaying them as a pop-up window. This allows users to efficiently access necessary information while reducing the risk of accessing web pages that include sexual advertisements.

[0085] "Crawling" is the process of automatically visiting many web pages on the Internet to collect information.

[0086] "Content analysis" refers to the act of extracting text and image information from collected data and evaluating and classifying it for a specific purpose.

[0087] "Sexual advertising" refers to advertising displays that contain sexual content and are considered offensive or harmful to users.

[0088] An "address" is a Uniform Resource Locator (URL) that identifies a specific web page on the Internet.

[0089] "Data storage means" refers to a database or storage device that stores analyzed data and allows the information to be quickly referenced later.

[0090] "Receiving" is the act of receiving data or information sent from an external source.

[0091] "Natural language processing" refers to the technology and methods that allow computers to understand and generate human language.

[0092] "Summarizing" is the act of concisely summarizing long text or complex information and extracting only the important points.

[0093] An "information processing device" is a device such as a computer or smartphone used by a user.

[0094] A "pop-up window" is a small window that appears within a user interface and is used to display information such as notifications or alerts.

[0095] An "extension" is a software module that adds functionality to a browser or other application.

[0096] MODE FOR CARRYING OUT THE INVENTION

[0097] This invention relates to a system for securely retrieving the contents of web pages on the Internet. In particular, it aims to restrict access to web pages containing sexually explicit advertisements, thereby preventing users from viewing inappropriate content. This system operates in cooperation with a server, a terminal, and a user.

[0098] Server-side processing

[0099] The server uses existing web crawling tools, such as Python's BeautifulSoup and Scrapy, to crawl the web. It collects the HTML content of each crawled web page and stores it in temporary storage. The server then applies natural language processing (NLP) techniques using TensorFlow and PyTorch to analyze the text data within the web page. At the same time, it also uses OpenCV and TensorFlow for image recognition to determine whether the web page contains sexually explicit advertisements.

[0100] The server trains a natural language processing model using a huge dataset (e.g., Wikipedia, Common Crawl). The training process is streamlined by using a high-performance GPU (e.g., Nvidia Tesla). This trained AI model is then used to summarize the content of the web page in question. The analysis results, summary information, and URLs are stored in a database. Databases such as MySQL and PostgreSQL are used.

[0101] Terminal side processing

[0102] When a user attempts to access a specific web page, the device utilizes a browser extension (e.g., Google Chrome Extension). This extension is implemented in JavaScript and captures the user's navigation events to obtain the URL. This URL is then sent to the server as an HTTPS request. The server checks the received URL against a list of problem pages in a database, and if a problem is detected, it sends a summary of the URL to the device.

[0103] The device displays a pop-up window based on the summary information received from the server. The pop-up window displays the message "This page contains inappropriate advertisements. Summary of page content: (Summary text)." This allows users to avoid inappropriate content.

[0104] User processing

[0105] Users can decide whether to access a particular web page based on the summary information displayed in their browser. This information allows them to safely and efficiently obtain the information they need. This also allows users to use the Internet safely, avoiding anxiety and unpleasant experiences.

[0106] Specific examples

[0107] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." The browser extension sends this URL to the server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model then summarizes the page's contents and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..." This summary is then sent to the device and displayed by the browser extension in a pop-up window. The user can read the summary and obtain the desired information without being exposed to any offensive advertisements.

[0108] Prompt Sentence Examples

[0109] "Summarize the content of a given web page URL, determine whether it contains sexual advertisements, and notify the user if necessary."

[0110] This system will promote the integrity of Internet advertising and provide users with a safe and comfortable web browsing environment.

[0111] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0112] Processing step flow

[0113] Step 1: Crawl the web page (server)

[0114] The server crawls multiple web pages on the Internet using a web crawler tool such as Python's BeautifulSoup or Scrapy. The input is the URL of the web page to be crawled, and the output is data including the HTML content. This data is stored in temporary storage.

[0115] Specifically, the server runs a Python script, and the web crawler tool crawls through the specified URL list to collect HTML data, which is then stored in local storage.

[0116] Step 2: Analyzing the web page content (server)

[0117] The server analyzes the collected HTML data and extracts text and images. The analysis uses frameworks such as TensorFlow and PyTorch. The HTML content is input, and the extracted text data and image URLs are output.

[0118] Specifically, the server parses the HTML using BeautifulSoup to extract all text content, and also extracts and lists URLs for image tags. This extracted data is used in the next analysis step.

[0119] Step 3: Identifying Sexual Ads (Server)

[0120] The server analyzes the extracted text data and image URLs to identify sexually explicit advertisements. This is done using natural language processing and image recognition technology. The extracted text data and image URLs are used as input, and the server outputs the identification of web pages containing sexually explicit advertisements and their URLs.

[0121] Specifically, the server uses an NLP model to evaluate text data and detect sexually explicit terms and expressions. At the same time, it uses an image recognition model to evaluate images retrieved from extracted image URLs. Based on the identification results, the webpage URLs are added to a problem page list.

[0122] Step 4: Training the summary model (server)

[0123] The server trains a natural language processing model using a large text dataset, taking the text dataset as input and the trained summarization model as output, which is used to quickly and effectively summarize web page content.

[0124] Specifically, the server loads a text dataset and trains a model using a natural language processing framework (e.g., TensorFlow), with the training process being streamlined using high-performance GPU hardware.

[0125] Step 5: Monitor the URL (Device)

[0126] A browser extension installed on a user's device monitors the URLs of web pages accessed by the user. The monitored URLs are used as input, and a request to send the URL to a server is used as output.

[0127] Specifically, the browser extension uses JavaScript scripts to capture user navigation events and retrieve the URL to be visited, which is then sent to the server as an HTTPS request.

[0128] Step 6: Identify the URL and send the summary information (server)

[0129] The server compares the received URL with a list of problem pages in the database, and if it determines that there is a problem, it generates and sends summary information. The received URL is the input, and the summary information is the output.

[0130] Specifically, the server performs a database query to check whether the URL is included in the problem page list, and if so, summarizes the page content using a trained summarization model and sends the summary information in JSON format to the device.

[0131] Step 7: View summary information (terminal)

[0132] The terminal displays a pop-up window based on the summary information received from the server. The received summary information is used as input, and the display in the pop-up window is used as output.

[0133] Specifically, the browser extension uses HTML and CSS to generate a popup window and display summary information. JavaScript DOM manipulation causes the popup to appear on the user's screen. The popup displays the following message: "This page contains inappropriate advertisements. Summary of page contents: (Summary text)."

[0134] Step 8: User decision (User)

[0135] The user checks the summary information displayed in the pop-up window and decides whether to access the page. The summary information is the input, and the action that determines whether to access the page is the output.

[0136] Specifically, users view the pop-up, examine its contents, and, if necessary, close the browser tab or visit another secure site.

[0137] In this way, a system is configured in which the server, terminal, and user work together in sequence to safely obtain the contents of web pages on the Internet.

[0138] (Application example 1)

[0139] 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."

[0140] The goal is to prevent users from being exposed to unintended sexual advertisements when browsing web pages on the Internet, while at the same time providing a means for them to quickly and safely access the information they need. Furthermore, by using dedicated applications, it is necessary to realize a comfortable browsing experience on a variety of devices, including smartphones and head-mounted displays.

[0141] 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.

[0142] In this invention, the server includes a means for crawling multiple web pages on the Internet and analyzing the content of those pages, a means for identifying web pages containing sexual advertisements based on the analyzed content and storing the URLs of those pages in a database, and a means for receiving the URL of a web page a user attempts to access and determining whether the URL is stored in the database. This allows users to enjoy safe and comfortable web browsing without being exposed to inappropriate advertisements. Furthermore, a dedicated application that can be installed on a smartphone or head-mounted display can monitor accessed URLs and provide summaries of the web page content using a generative AI model.

[0143] The "Internet" is a network that interconnects computers and networks around the world, enabling the exchange of information.

[0144] "Crawling" is a technique for automatically crawling through web pages on the Internet and collecting information.

[0145] "Means for analyzing the content of web pages" refers to technology that extracts data such as text and images from web pages and uses them to understand and evaluate their content.

[0146] "Sexual advertising" refers to advertisements that appear on web pages and contain sexual content.

[0147] A "database" is a system for systematically storing and managing data.

[0148] "Means for receiving the URL of the web page that the user wishes to access" refers to a technique for obtaining the URL of a specific web page when the user accesses that web page.

[0149] A "pop-up window" is a small window that temporarily appears on a user's device to provide important information or warnings.

[0150] A "browser extension" is a software component that adds or extends the functionality of a web browser.

[0151] A "generative AI model" is an artificial intelligence model that has been pre-trained on large datasets to perform tasks such as text generation and summarization.

[0152] A "smartphone" is a mobile phone equipped with internet connectivity and many applications.

[0153] A "head-mounted display" is a device worn on the head that displays visual information.

[0154] To implement this invention, the server, terminal, and user must work together. The details of how each part works are described below.

[0155] Server-side processing

[0156] The server first crawls multiple web pages on the Internet and analyzes their content. Specifically, it retrieves the content of the web pages using the requests library and analyzes the HTML content using BeautifulSoup. Based on the analyzed page content, it determines whether or not the pages contain sexual advertisements. The URLs of pages that contain sexual advertisements are stored in a database.

[0157] The server then trains a summarization model using natural language processing techniques, using the pipeline function in the transformers library to train a generative AI model on a large dataset, which is then used to summarize the content of web pages.

[0158] Terminal side processing

[0159] When a user attempts to access a specific web page in a browser, a browser extension on the device monitors the URL. This monitoring is performed using JavaScript and related browser extension APIs. The URL of the web page the user is attempting to access is sent to a server, which determines whether the URL is stored in its database. If the determination is that the page is problematic, a summary is sent from the server to the device, which then displays the summary to the user in a pop-up window.

[0160] User processing

[0161] Users can check the summary information displayed on their browser and decide whether to access the website based on the content. This allows users to quickly and safely access the information they need without being exposed to intrusive advertisements.

[0162] Specific examples

[0163] As a specific use case, consider the case where a user attempts to access the URL "http: / / example.com / sample-page." At this time, the browser extension sends this URL to the server. The server references the database and identifies this URL as a page containing sexually explicit advertisements. The generative AI model summarizes the page content and generates summary information such as, "This website provides reviews of image editing software. The key points of the reviews are that the user interface is easy to use and that a wide range of editing tools is available." This summary information is sent to the device and displayed as a pop-up window by the browser extension.

[0164] Prompt Sentence Examples

[0165] The following is an example of a prompt that should be given to the generative AI model regarding the URL the user attempted to access and a summary of its content:

[0166] URL: http: / / example.com / sample-page

[0167] Description: This website provides reviews of photo editing software. The main points of the reviews are that the user interface is easy to use and that there are a wide range of editing tools available.

[0168] In this way, the present invention provides users with access to necessary information while preventing them from being exposed to unintended sexual advertisements when browsing web pages on the Internet.By using devices such as smartphones and head-mounted displays, a comfortable browsing experience can be achieved on a variety of devices.

[0169] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0170] Step 1:

[0171] The server crawls multiple web pages on the Internet and retrieves their contents. Specifically, it retrieves the HTML content of each web page using the requests library and parses it using BeautifulSoup. The input is the URL of the web page, and the output is the content of the page, such as text and images.

[0172] Step 2:

[0173] The server determines whether the web page contains sexual advertisements based on the analyzed content. It uses specific keywords and image recognition algorithms to detect sexual content. The input is the web page content extracted in step 1, and the output is the result of determining whether the web page contains sexual advertisements.

[0174] Step 3:

[0175] The server stores the URLs of pages that are determined to contain sexually explicit ads in a database. The database records the URLs of problematic web pages and their verdicts. The input is the verdict and URL obtained in step 2, and the output is the records stored in the database.

[0176] Step 4:

[0177] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. This browser extension captures the URL entered by the user in real time and sends it to a server. The input is the URL the user is trying to access, and the output is the data sent to the server.

[0178] Step 5:

[0179] The server checks the received URL against a list of problem pages stored in a database. The input is the URL sent from the device, and the output is a decision as to whether the URL is a problematic page.

[0180] Step 6:

[0181] The server uses a generative AI model to create summaries for URLs that are determined to be problematic. Specifically, it summarizes the page content using the summarization function of the transformers library. The input is the determined web page content, and the output is a summary.

[0182] Step 7:

[0183] The server sends the summarized content to the terminal. The input is the generated summary sentence, and the output is the data sent to the terminal.

[0184] Step 8:

[0185] The terminal displays the received summary information to the user as a pop-up window. Specifically, it displays a pop-up on the browser using JavaScript. The input is the summary data received from the server, and the output is the pop-up window displayed to the user.

[0186] Step 9:

[0187] The user checks the displayed summary information and makes the final decision on whether to access it. The input is the summary information displayed in the popup window, and the output is the user's decision on whether to allow or deny access.

[0188] 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.

[0189] Overall program overview

[0190] This invention relates to a system that safely retrieves the contents of web pages on the Internet and presents appropriate information according to the user's emotional state. This system operates in cooperation with a server, a terminal, and a user, and provides information that takes into consideration the user's emotions, while restricting access to web pages that contain sexually explicit advertisements.

[0191] Server-side processing

[0192] The server first crawls multiple web pages on the Internet and analyzes their content. It uses natural language processing and image recognition technologies to identify whether they contain sexually explicit advertisements. It stores the URLs of web pages determined to contain sexually explicit advertisements in a database. The server then uses a large amount of data to train a natural language processing model. This model is then used to summarize the content of the web pages.

[0193] Furthermore, the server is equipped with an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from information such as the user's facial expressions, voice, and input text. This emotion data is saved for each user and accumulates over time.

[0194] Terminal side processing

[0195] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. The extension sends the URL to a server. The server receives the URL and checks it against a list of problem pages stored in a database. If a problem is detected, a summary is sent from the server to the device.

[0196] The device displays this summary information in a pop-up window. Furthermore, the emotion engine analyzes the user's real-time emotions and adjusts the way information is presented according to the user's emotional state. For example, if the user is in a negative emotional state, an appropriate warning message can be displayed.

[0197] User processing

[0198] Users can view summary information about web pages in their browsers. Based on this information, they can decide whether to access a particular web page. Furthermore, information provided by the emotion engine allows them to learn the best way to respond to their own emotional state. User emotion data is stored over the long term, and an algorithm is run to provide the best display method for each individual user.

[0199] Specific examples

[0200] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." A browser extension sends this URL to a server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model summarizes the content of this page and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..."

[0201] Furthermore, the emotion engine analyzes the user's emotions, and if the user is expressing negative emotions such as stress or anger, a warning message is generated stating, "This page currently contains inappropriate advertisements and is not recommended for viewing." This summary information and warning message are sent to the device and displayed by the browser extension as a pop-up window. Users feel that they have read this summary and obtained the desired information without being exposed to unpleasant advertisements.

[0202] This system configuration can provide a better user experience by preventing users from being exposed to unnecessary advertisements and by taking into consideration their emotional state. It is also expected to contribute to the soundness of internet advertising.

[0203] The processing flow will be explained below.

[0204] Step 1:

[0205] Server: Crawl web pages

[0206] The server periodically crawls pages on the Internet and retrieves the HTML source of new pages based on the specified URL list. This HTML source is saved for content analysis.

[0207] Step 2:

[0208] Server: Analyze content and identify problem pages

[0209] The server analyzes the content of the crawled pages. It uses natural language processing and image recognition technology to examine the ad banners and metadata on the pages. If any sexually explicit ads are found, the URL of the page is saved in a database as a problem page.

[0210] Step 3:

[0211] Server: Training the natural language processing model

[0212] The server trains a natural language processing model (such as BERT or GPT-3) using large amounts of text data, which is then used to summarize the content of web pages.

[0213] Step 4:

[0214] Server: Preparing the emotion engine

[0215] The server prepares the emotion engine and trains a model to analyze the user's facial expression data, voice data, input text, etc. This uses machine learning techniques to enable high-accuracy recognition of the user's emotional state.

[0216] Step 5:

[0217] Terminal: Monitoring web page requests

[0218] A browser extension installed on the user's device monitors the URLs the user attempts to access and sends this URL information to a server.

[0219] Step 6:

[0220] Server: URL rating

[0221] The server checks the received URL against its database, determines whether the URL is included in the list of problem pages, and returns the result.

[0222] Step 7:

[0223] Server: Generate summary information

[0224] If a page is determined to be problematic, the server uses a natural language processing model to summarize the content of the page, and then generates this summary information and sends it to the terminal.

[0225] Step 8:

[0226] Terminal: Acquisition and analysis of emotion data

[0227] The device's browser extension collects the user's facial expression data, voice data, input text, etc. in real time and sends it to the emotion engine, which analyzes this data and determines the user's emotional state.

[0228] Step 9:

[0229] Server: Sends summary information and generates warning messages

[0230] The server generates an appropriate warning message along with summary information based on the user's emotional state determined by the emotion engine, and transmits this information to the terminal.

[0231] Step 10:

[0232] Terminal: Display summary information and warning messages

[0233] The browser extension displays the received summary information and a warning message in a pop-up window, such as "This page contains inappropriate advertising. Page content summary: (summary text) Viewing is not recommended based on your current emotional state."

[0234] Step 11:

[0235] User: Review summary information and emotional feedback

[0236] Users can review the summary information and warning messages provided in the pop-up window, and with the necessary information, they can safely browse the web without being exposed to annoying ads.

[0237] Step 12:

[0238] User: Safe Web Browsing

[0239] Users can enjoy a better user experience by being protected from unnecessary advertisements and receiving information tailored to their emotional state.

[0240] Example 2

[0241] 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."

[0242] In today's internet usage environment, users often encounter inappropriate sexual advertisements when browsing web pages, significantly degrading the user experience. Furthermore, the lack of appropriate information presented in response to the user's emotional state can potentially increase the user's psychological burden. Conventional technologies have not provided a means to effectively solve these problems simultaneously, making it difficult to provide a comfortable browsing environment for users.

[0243] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for crawling multiple web pages on the Internet and analyzing the contents of those pages, a means for identifying web pages containing sexual advertisements based on the analyzed contents and saving the URLs of those pages in a recording device, and a means for receiving the URL of a web page that a user attempts to access and determining whether the URL is saved in the recording device. This makes it possible to determine in advance whether a web page that a user attempts to view contains inappropriate advertisements and provide the user with appropriate information.

[0244] The server further includes means for summarizing the content of web pages containing sexual advertisements, means for transmitting the summarized content to the user's information processing terminal and displaying it as a pop-up window, means for recognizing the user's emotions and storing the emotion data in a recording device, and means for adjusting the information presentation method according to the user's emotional state, thereby realizing information provision that takes into consideration the user's emotional state and improving the user experience.

[0245] Furthermore, by including a means for training a natural language processing model for summarization and an extension function for the information processing terminal that monitors the URLs of web pages that users are attempting to access, it becomes possible to provide more accurate summary information and monitor in real time, thereby realizing a safe and comfortable internet experience for users without being exposed to inappropriate advertisements.

[0246] A "web page" is a document that displays information that is publicly available on the Internet and is written in a language such as HTML.

[0247] "Crawling" is the process of automatically visiting specific web pages and retrieving their content.

[0248] "Analysis" refers to the processing of acquired data using computer algorithms to extract or evaluate necessary information.

[0249] "Sexual advertising" means commercial information containing sexual content that may have an inappropriate effect on users.

[0250] A "storage device" is a physical or virtual device for storing data, such as a database.

[0251] An "information processing terminal" is an electronic device that can process information, such as a computer, smartphone, or tablet.

[0252] A "pop-up window" is a small window that suddenly appears on a user's screen and is used to present specific information.

[0253] "Emotion" refers to a psychological state that is expressed through a person's facial expression, voice, input text, etc.

[0254] "Emotional data" refers to information about the user's psychological state, including facial expressions and voice analysis results.

[0255] A "natural language processing model" is a machine learning model for understanding and processing human language, and is used for tasks such as summarizing and translating text.

[0256] An "extension" is an additional program that extends the functionality of existing software and is used to enhance the functionality of a browser.

[0257] This invention is a system in which a server, terminals, and users work together to provide a safe and comfortable Internet browsing environment. Specifically, the server crawls multiple web pages on the Internet and analyzes the content of those pages. It also monitors the URLs of web pages that users attempt to access, summarizes the content of problematic pages, and displays a warning to the user in real time.

[0258] Server-side processing

[0259] The server first uses a web crawler (e.g., BeautifulSoup, Scrapy) to crawl multiple web pages on the Internet. It then obtains the HTML content of each page and analyzes it using natural language processing technology (e.g., spaCy) or image recognition technology (e.g., OpenCV). This analysis allows it to extract text and images from the web pages.

[0260] The extracted content is then evaluated using an AI model (e.g., a model using TensorFlow) to determine whether it contains sexually explicit content. Based on the results of the evaluation, the URLs of web pages that are found to contain sexually explicit content are stored in a storage device (e.g., a MySQL database).

[0261] Additionally, the server summarizes the content of the web page using a natural language processing model (e.g., BERT), which is pre-trained using a large amount of text data.

[0262] In addition, the server uses an emotion engine (e.g., Microsoft Azure Cognitive Services) to recognize the user's emotions. Emotion data is stored in a recording device for each user.

[0263] Terminal side processing

[0264] A browser extension (e.g., Chrome Extension) is installed on the device, and when a user attempts to access a specific web page, it sends the URL to the server. The server receives the URL and compares it with a list of problem pages stored in a recording device. For URLs determined to be problematic, the server generates summary information and a warning message and sends them to the device.

[0265] The terminal displays the received summary information and warning message in a pop-up window, allowing the user to obtain information without being exposed to inappropriate content.

[0266] User processing

[0267] The user checks the displayed summary information and warning message and decides whether to access the web page. The emotion engine also analyzes the user's real-time emotions and sends the results to the server, which then further optimizes the way information is presented in the future.

[0268] Specific examples

[0269] For example, if a user attempts to access the URL "example-adultcontent.com," the device's browser extension sends the URL to the server. The server compares it with information in the database and determines that the URL contains inappropriate advertising. It generates a summary such as "This website provides reviews of image editing software. The gist of the review is..." Furthermore, the emotion engine analyzes the user's emotions, and if the user is feeling stressed or anxious, it generates a warning message saying, "This page contains inappropriate advertising. Viewing it is not recommended."

[0270] An example of the prompt that the user sees:

[0271] "Users submit the URL of the website they are trying to access. Additionally, they enter their current emotional state. For example, the URL 'example-adultcontent.com' and the emotion 'I'm stressed.'"

[0272] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0273] Step 1:

[0274] The server uses a web crawler to crawl multiple web pages on the Internet. The web crawler (e.g., BeautifulSoup, Scrapy) is configured to retrieve the HTML content of all pages in a specified domain. The input to this process is a list of URLs, and the output is the HTML code for each web page.

[0275] Step 2:

[0276] The server analyzes the acquired HTML code and extracts text and images. This analysis uses natural language processing technology (e.g., spaCy) and image recognition technology (e.g., OpenCV). The input is the HTML code, and the output is text data and image data. The extracted text and images are evaluated in the next step.

[0277] Step 3:

[0278] The server inputs the extracted text data and image data into an AI model (e.g., a model built using TensorFlow) to determine whether or not the data contains sexual advertisements. The input is text data and image data, and the output is a determination result of "contains / does not contain sexual advertisements." Based on this result, the determined URLs are stored in a recording device (e.g., a MySQL database) as those containing sexual advertisements.

[0279] Step 4:

[0280] The server uses a natural language processing model (e.g., BERT) to summarize the content of a web page. This model has been trained on a large amount of text data in advance. The input is the text data of the web page, and the output is the summary text. This summary text is saved for presentation to the user.

[0281] Step 5:

[0282] The server uses an emotion engine (e.g., Microsoft Azure Cognitive Services) to analyze the user's emotions. Inputs include the user's facial expressions, voice, and input text, and the output is the user's emotional data (e.g., "feeling stressed," "relaxed," etc.). This emotional data is stored in a recording device for each user.

[0283] Step 6:

[0284] A browser extension (e.g., Chrome Extension) is installed on the device, and when a user tries to access a specific web page, it monitors the URL and sends it to a server. The input is the URL that the user is trying to access, and the output is that the URL is sent to the server.

[0285] Step 7:

[0286] The server checks the received URL against the problem page list stored in the recording device. For URLs that are determined to be problematic, the server generates a warning message based on the previously generated summary text and the user's emotional data. The input is the URL to be accessed and the user's emotional data, and the output is the summary text and the warning message.

[0287] Step 8:

[0288] The terminal displays the summary information and warning message received from the server as a pop-up window. If a user attempts to access "example-adultcontent.com", it displays the message "This site contains inappropriate advertisements. Viewing is not recommended." The input is the summary text and warning message sent from the server, and the output is the display in the pop-up window.

[0289] Step 9:

[0290] The user checks the displayed summary information and warning message and decides whether to access the web page. At the same time, the emotion engine analyzes the user's real-time emotions and sends the results to the server. The input is the user's decision and real-time emotion data, and the output is whether to access the page and updated emotion data.

[0291] (Application example 2)

[0292] 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."

[0293] Currently, when browsing the Internet, there is a risk of accessing inappropriate advertisements or content. Furthermore, information is provided in a uniform manner without considering the user's emotional state, resulting in a suboptimal user experience. Furthermore, users may experience stress or anxiety when exposed to inappropriate content. There is a need to resolve these issues and provide a safer, more user-friendly Internet browsing environment.

[0294] 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.

[0295] In this invention, the server includes means for crawling multiple web pages on the Internet and analyzing the contents of those pages, means for identifying web pages containing inappropriate advertisements based on the analyzed contents and storing the URLs of those pages in a database, means for receiving the URL of a web page that a user attempts to access and determining whether the URL is stored in the database, means for summarizing the contents of the web page containing the inappropriate advertisement, means for analyzing the user's emotional state and generating a warning message or alternative information according to the emotional state, and means for transmitting the summarized contents and information according to the user's emotions to the user's terminal and displaying them as a pop-up window. This makes it possible to provide optimal information according to the user's emotional state while preventing the user from being exposed to inappropriate advertisements or content.

[0296] "Web pages on the Internet" refers to individual pages of publicly accessible websites connected to the Internet.

[0297] "Crawl" refers to the technical process of automatically visiting designated web pages and collecting their content.

[0298] "Content analysis" refers to the process of understanding collected information, such as text and images, from web pages and extracting specific patterns and features.

[0299] "Inappropriate Advertisements" refers to advertisements that may be offensive to users, especially those that contain sexual or violent content.

[0300] "Identifying web pages" refers to the process of finding and identifying web pages that meet certain criteria based on their analyzed content.

[0301] "Means for storing in a database" refers to the process of storing the URLs of the identified web pages and other related information in a dedicated database.

[0302] "Means for receiving the URL of the web page that the user wishes to access" refers to the process of obtaining the address information of the web page that the user has entered or selected.

[0303] "Means for determining" refers to the process of checking a received URL against a list of inappropriate web pages in a database to determine whether it is an inappropriate page.

[0304] "Means for summarizing content" refers to the process of providing a brief summary of the content of a particular Web page.

[0305] "Means for analyzing emotional state" refers to the technical process of recognizing and analyzing the user's current emotions from data such as facial expressions, voice, and input text.

[0306] "Warning Message" refers to a notification that alerts a user to the risk of a particular action or situation.

[0307] "Alternate Information" refers to safe and appropriate information that is provided in place of the page a user is attempting to access.

[0308] "Pop-up window" refers to a small window that appears floating on a user's screen and displays information or a message.

[0309] "User's Device" refers to the electronic device used by the User, such as a smartphone, tablet, or PC.

[0310] "System" refers to a collection of devices and programs that combine all of the above means and have the overall function.

[0311] This invention is a system for safely removing inappropriate advertisements on web pages and providing users with information optimized for them. This system functions in cooperation with a server, a terminal, and a user.

[0312] Server-side processing

[0313] The server uses a crawling engine and a natural language processing engine to crawl multiple web pages on the Internet and analyze their content. It uses web scraping tools such as Python's requests library and BeautifulSoup library for the analysis. It also uses machine learning models to identify web pages that contain inappropriate ads based on the analyzed content. This information is stored in a database for later use.

[0314] The server then uses generative AI models such as Google's Text-To-Text Transfer Transformer (T5) and Bidirectional Encoder Representations from Transformers (BERT) to summarize the content of the webpage, and then uses Microsoft's Azure Emotion API and DeepFace library to apply facial and speech recognition technologies to analyze the user's emotional state.

[0315] Terminal side processing

[0316] When a user attempts to access a web page, the browser extension on the device monitors the URL and sends it to a server. The server checks the URL against a database to determine if it contains inappropriate content. If it does, a server-generated summary is sent to the device and displayed in a pop-up window.

[0317] The device also uses a camera and microphone to analyze the user's emotions in real time, and generates warning messages and alternative information based on the results of this emotion analysis.

[0318] User processing

[0319] Users use their browsers to view web page summaries and warning messages, and use this information to decide whether to access a particular web page. User emotion data is stored over the long term, and this data can be used to provide information optimized for each individual user.

[0320] Specific examples

[0321] For example, consider a situation where a user attempts to access the URL "https: / / example-adultcontent.com." The browser extension sends this URL to a server, which then consults a database. If the URL is determined to be a page containing inappropriate advertising, the AI ​​summarization model summarizes the page's content and generates a summary that reads, "This website contains inappropriate advertising, but its content is a review of image editing software."

[0322] In addition, the emotion engine analyzes the user's facial expressions, and if a negative emotion is detected, a warning message is generated stating, "This page currently contains inappropriate advertisements and is not recommended for viewing." This information and warning message are sent to the device and displayed as a pop-up window, allowing the user to obtain the desired information without being exposed to unpleasant advertisements.

[0323] Use the following as an example prompt:

[0324] test_url = "https: / / example-adultcontent.com"

[0325] test_image_path = "path / to / user_image.jpg"

[0326] main(test_url, test_image_path)

[0327] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0328] Step 1:

[0329] The server crawls multiple web pages on the Internet. During this process, it uses a specific algorithm to generate a list of web page URLs, and then uses Python's requests library and BeautifulSoup library to retrieve the content of each web page. The HTML data of the retrieved web pages is used as input, and the analysis results are output.

[0330] Step 2:

[0331] The server analyzes the content of the retrieved web pages. This analysis uses a natural language processing engine and an image recognition engine to process the text and image data of each web page. A machine learning model is used to determine whether the web page contains inappropriate advertising. The input data is the text and images of the web page, and the output is a determination of whether the web page contains inappropriate advertising.

[0332] Step 3:

[0333] The server identifies web pages containing inappropriate ads based on the analysis results and stores the URLs of these pages in a dedicated database. The stored information includes the URL of the web page and the reason why it was determined to be inappropriate. The output is to store a list of identified URLs in the database.

[0334] Step 4:

[0335] When a user accesses a web page using a browser on their device, a browser extension on the device monitors the URL. It captures the URL the user types in the address bar and sends it to a server. This URL is the input data, and the sent URL is the output.

[0336] Step 5:

[0337] The server checks the received URL against a database to determine whether it contains inappropriate content. Database search technology is used to check the existence of the URL. The input is the URL sent by the user, and the output is the result of whether the URL is inappropriate.

[0338] Step 6:

[0339] If the server determines that the webpage contains inappropriate ads, it uses an AI generative model such as Google's T5 or BERT to summarize the content of the webpage. The generative AI model takes the text data of the webpage as input and outputs a concise summary text.

[0340] Step 7:

[0341] The server receives real-time captured images of the user's facial expressions or voice data to analyze the user's emotional state. Using this as input data, it determines the user's emotional state using an emotion analysis engine (DeepFace or Microsoft Azure Emotion API). The analysis result of the emotional state is obtained as output.

[0342] Step 8:

[0343] The server generates a warning message or alternative information according to the user's emotional state based on the emotion analysis results. The input data is the emotion analysis results, and the output is an appropriate warning message or alternative information.

[0344] Step 9:

[0345] The server sends the summarized content and emotion-based information to the user's terminal and displays them as a pop-up window on the terminal. The input data are the summary information and the emotion-based message, and the output is transmission to the user's terminal and display of the pop-up window.

[0346] 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.

[0347] 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.

[0348] 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.

[0349] [Second embodiment]

[0350] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0351] 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.

[0352] 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).

[0353] 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.

[0354] 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.

[0355] 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).

[0356] 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. 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.

[0357] 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.

[0358] 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.

[0359] 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.

[0360] 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.

[0361] 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."

[0362] Overall program overview

[0363] This invention relates to a system for securely acquiring the contents of web pages on the Internet. This system involves the cooperation of a server, a terminal, and a user, and aims to restrict access to web pages that contain sexually explicit advertisements.

[0364] Server-side processing

[0365] The server first crawls multiple web pages on the Internet. It analyzes the content of the crawled web pages to identify whether they contain sexually explicit advertisements. The URLs of the identified web pages are stored in a database. When analyzing this content, the server uses natural language processing and image recognition technologies to evaluate the content of the advertisements.

[0366] The server then uses the massive dataset to train a natural language processing model. The trained AI model is then used to summarize the content of the web page. This summarization model is designed to quickly generate a summary of the web page in question. The summarized information is also stored in a database, ready to be served immediately if needed.

[0367] Terminal side processing

[0368] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. The extension sends the URL to a server, which receives the URL and checks it against a list of problem pages stored in a database. If a problem is detected, a summary is sent from the server to the device.

[0369] The device displays a pop-up window to provide summary information to the user. The pop-up window displays, "This page contains inappropriate advertisements. Summary of page contents: (Summary text)." Based on this summary information, the user can obtain the necessary information without being exposed to unpleasant advertisements.

[0370] User processing

[0371] Users can view web page summary information in their browsers and use this information to decide whether to visit a particular web page. This allows users to easily access the information they want while avoiding pages that contain sexually explicit advertisements.

[0372] Specific examples

[0373] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." A browser extension sends this URL to a server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model summarizes the content of this page and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..."

[0374] This summary information is sent to the device and displayed as a pop-up window by the browser extension. The user reads the summary and feels that they have obtained the desired information without being exposed to annoying advertisements. In this way, the system provides users with access to the information they need while preventing them from being exposed to unwanted advertisements.

[0375] This system configuration promotes the soundness of Internet advertising and provides users with a safe and comfortable web browsing environment.

[0376] The processing flow will be explained below.

[0377] Step 1:

[0378] Server: Crawl web pages

[0379] The server automatically collects pages on the Internet, crawls new pages based on a periodically updated URL list, obtains the HTML source of the crawled pages, and stores it for content analysis.

[0380] Step 2:

[0381] Server: Content analysis and identification

[0382] The server analyzes the content of the crawled pages, using natural language processing and image recognition technology to examine the ad banners and metadata on the pages, determining whether they contain sexually explicit content and storing the URLs of problematic pages in a database.

[0383] Step 3:

[0384] Server: Training the natural language processing model

[0385] The server trains a natural language processing model (such as BERT or GPT-3) using a large amount of text data. The training data includes high-quality summaries and feedback. This model is then used to summarize the content of web pages.

[0386] Step 4:

[0387] Terminal: Monitoring web page requests

[0388] A browser extension installed on the user's device monitors the URLs the user attempts to access and sends this URL information to a server.

[0389] Step 5:

[0390] Server: URL rating

[0391] The server checks the received URL against a database to see if the URL exists and whether it is a page containing sexual advertisements.

[0392] Step 6:

[0393] Server: Generate summary information

[0394] If the page is determined to be relevant, the server uses a natural language processing model to summarize the content of the page, generates this summary information, and sends it to the terminal.

[0395] Step 7:

[0396] Terminal: Display summary information

[0397] The browser extension displays the received summary information in a popup window, displaying the message "This page contains inappropriate advertising. Summary of page contents: (summary text)" to the user.

[0398] Step 8:

[0399] User: Review summary information and make a decision

[0400] The user reviews the summary information provided in the pop-up window, allowing them to decide whether to obtain the information they need without being exposed to the intrusive advertisement.

[0401] Step 9:

[0402] User: Safe Web Browsing

[0403] Users can safely obtain the information they are looking for while avoiding inappropriate ads, improving the user experience and promoting the integrity of internet advertising.

[0404] Example 1

[0405] 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."

[0406] On the Internet, there are web pages containing inappropriate content, and accessing them can lead to unpleasant experiences or expose users to harmful information. Web pages containing sexually explicit advertisements are particularly undesirable for users. To address this, it is necessary to provide an environment in which users can browse the web with peace of mind. However, existing filtering systems have the problem of being unable to completely filter out harmful content.

[0407] 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.

[0408] In this invention, the server includes means for crawling multiple web pages on the Internet and analyzing the contents of those pages, means for identifying sexual advertisements based on the analyzed contents and saving the addresses of those pages in data storage means, means for receiving the addresses of web pages that a user wishes to access and determining whether the addresses are saved in the data storage means, means for summarizing the contents of web pages that include sexual advertisements using natural language processing, and means for transmitting the summarized contents to the user's information processing device and displaying them as a pop-up window. This allows users to efficiently access necessary information while reducing the risk of accessing web pages that include sexual advertisements.

[0409] "Crawling" is the process of automatically visiting many web pages on the Internet to collect information.

[0410] "Content analysis" refers to the act of extracting text and image information from collected data and evaluating and classifying it for a specific purpose.

[0411] "Sexual advertising" refers to advertising displays that contain sexual content and are considered offensive or harmful to users.

[0412] An "address" is a Uniform Resource Locator (URL) that identifies a specific web page on the Internet.

[0413] "Data storage means" refers to a database or storage device that stores analyzed data and allows the information to be quickly referenced later.

[0414] "Receiving" is the act of receiving data or information sent from an external source.

[0415] "Natural language processing" refers to the technology and methods that allow computers to understand and generate human language.

[0416] "Summarizing" is the act of concisely summarizing long text or complex information and extracting only the important points.

[0417] An "information processing device" is a device such as a computer or smartphone used by a user.

[0418] A "pop-up window" is a small window that appears within a user interface and is used to display information such as notifications or alerts.

[0419] An "extension" is a software module that adds functionality to a browser or other application.

[0420] MODE FOR CARRYING OUT THE INVENTION

[0421] This invention relates to a system for securely retrieving the contents of web pages on the Internet. In particular, it aims to restrict access to web pages containing sexually explicit advertisements, thereby preventing users from viewing inappropriate content. This system operates in cooperation with a server, a terminal, and a user.

[0422] Server-side processing

[0423] The server uses existing web crawling tools, such as Python's BeautifulSoup and Scrapy, to crawl the web. It collects the HTML content of each crawled web page and stores it in temporary storage. The server then applies natural language processing (NLP) techniques using TensorFlow and PyTorch to analyze the text data within the web page. At the same time, it also uses OpenCV and TensorFlow for image recognition to determine whether the web page contains sexually explicit advertisements.

[0424] The server trains a natural language processing model using a huge dataset (e.g., Wikipedia, Common Crawl). The training process is streamlined by using a high-performance GPU (e.g., Nvidia Tesla). This trained AI model is then used to summarize the content of the web page in question. The analysis results, summary information, and URLs are stored in a database. Databases such as MySQL and PostgreSQL are used.

[0425] Terminal side processing

[0426] When a user attempts to access a specific web page, the device utilizes a browser extension (e.g., Google Chrome Extension). This extension is implemented in JavaScript and captures the user's navigation events to obtain the URL. This URL is then sent to the server as an HTTPS request. The server checks the received URL against a list of problem pages in a database, and if a problem is detected, it sends a summary of the URL to the device.

[0427] The device displays a pop-up window based on the summary information received from the server. The pop-up window displays the message "This page contains inappropriate advertisements. Summary of page content: (Summary text)." This allows users to avoid inappropriate content.

[0428] User processing

[0429] Users can decide whether to access a particular web page based on the summary information displayed in their browser. This information allows them to safely and efficiently obtain the information they need. This also allows users to use the Internet safely, avoiding anxiety and unpleasant experiences.

[0430] Specific examples

[0431] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." The browser extension sends this URL to the server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model then summarizes the page's contents and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..." This summary is then sent to the device and displayed by the browser extension in a pop-up window. The user can read the summary and obtain the desired information without being exposed to any offensive advertisements.

[0432] Prompt Sentence Examples

[0433] "Summarize the content of a given web page URL, determine whether it contains sexual advertisements, and notify the user if necessary."

[0434] This system will promote the integrity of Internet advertising and provide users with a safe and comfortable web browsing environment.

[0435] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0436] Processing step flow

[0437] Step 1: Crawl the web page (server)

[0438] The server crawls multiple web pages on the Internet using a web crawler tool such as Python's BeautifulSoup or Scrapy. The input is the URL of the web page to be crawled, and the output is data including the HTML content. This data is stored in temporary storage.

[0439] Specifically, the server runs a Python script, and the web crawler tool crawls through the specified URL list to collect HTML data, which is then stored in local storage.

[0440] Step 2: Analyzing the web page content (server)

[0441] The server analyzes the collected HTML data and extracts text and images. The analysis uses frameworks such as TensorFlow and PyTorch. The HTML content is input, and the extracted text data and image URLs are output.

[0442] Specifically, the server parses the HTML using BeautifulSoup to extract all text content, and also extracts and lists URLs for image tags. This extracted data is used in the next analysis step.

[0443] Step 3: Identifying Sexual Ads (Server)

[0444] The server analyzes the extracted text data and image URLs to identify sexually explicit advertisements. This is done using natural language processing and image recognition technology. The extracted text data and image URLs are used as input, and the server outputs the identification of web pages containing sexually explicit advertisements and their URLs.

[0445] Specifically, the server uses an NLP model to evaluate text data and detect sexually explicit terms and expressions. At the same time, it uses an image recognition model to evaluate images retrieved from extracted image URLs. Based on the identification results, the webpage URLs are added to a problem page list.

[0446] Step 4: Training the summary model (server)

[0447] The server trains a natural language processing model using a large text dataset, taking the text dataset as input and the trained summarization model as output, which is used to quickly and effectively summarize web page content.

[0448] Specifically, the server loads a text dataset and trains a model using a natural language processing framework (e.g., TensorFlow), with the training process being streamlined using high-performance GPU hardware.

[0449] Step 5: Monitor the URL (Device)

[0450] A browser extension installed on a user's device monitors the URLs of web pages accessed by the user. The monitored URLs are used as input, and a request to send the URL to a server is used as output.

[0451] Specifically, the browser extension uses JavaScript scripts to capture user navigation events and retrieve the URL to be visited, which is then sent to the server as an HTTPS request.

[0452] Step 6: Identify the URL and send the summary information (server)

[0453] The server compares the received URL with a list of problem pages in the database, and if it determines that there is a problem, it generates and sends summary information. The received URL is the input, and the summary information is the output.

[0454] Specifically, the server performs a database query to check whether the URL is included in the problem page list, and if so, summarizes the page content using a trained summarization model and sends the summary information in JSON format to the device.

[0455] Step 7: View summary information (terminal)

[0456] The terminal displays a pop-up window based on the summary information received from the server. The received summary information is used as input, and the display in the pop-up window is used as output.

[0457] Specifically, the browser extension uses HTML and CSS to generate a popup window and display summary information. JavaScript DOM manipulation causes the popup to appear on the user's screen. The popup displays the following message: "This page contains inappropriate advertisements. Summary of page contents: (Summary text)."

[0458] Step 8: User decision (User)

[0459] The user checks the summary information displayed in the pop-up window and decides whether to access the page. The summary information is the input, and the action that determines whether to access the page is the output.

[0460] Specifically, users view the pop-up, examine its contents, and, if necessary, close the browser tab or visit another secure site.

[0461] In this way, a system is configured in which the server, terminal, and user work together in sequence to safely obtain the contents of web pages on the Internet.

[0462] (Application example 1)

[0463] 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."

[0464] The goal is to prevent users from being exposed to unintended sexual advertisements when browsing web pages on the Internet, while at the same time providing a means for them to quickly and safely access the information they need. Furthermore, by using dedicated applications, it is necessary to realize a comfortable browsing experience on a variety of devices, including smartphones and head-mounted displays.

[0465] 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.

[0466] In this invention, the server includes a means for crawling multiple web pages on the Internet and analyzing the content of those pages, a means for identifying web pages containing sexual advertisements based on the analyzed content and storing the URLs of those pages in a database, and a means for receiving the URL of a web page a user attempts to access and determining whether the URL is stored in the database. This allows users to enjoy safe and comfortable web browsing without being exposed to inappropriate advertisements. Furthermore, a dedicated application that can be installed on a smartphone or head-mounted display can monitor accessed URLs and provide summaries of the web page content using a generative AI model.

[0467] The "Internet" is a network that interconnects computers and networks around the world, enabling the exchange of information.

[0468] "Crawling" is a technique for automatically crawling through web pages on the Internet and collecting information.

[0469] "Means for analyzing the content of web pages" refers to technology that extracts data such as text and images from web pages and uses them to understand and evaluate their content.

[0470] "Sexual advertising" refers to advertisements that appear on web pages and contain sexual content.

[0471] A "database" is a system for systematically storing and managing data.

[0472] "Means for receiving the URL of the web page that the user wishes to access" refers to a technique for obtaining the URL of a specific web page when the user accesses that web page.

[0473] A "pop-up window" is a small window that temporarily appears on a user's device to provide important information or warnings.

[0474] A "browser extension" is a software component that adds or extends the functionality of a web browser.

[0475] A "generative AI model" is an artificial intelligence model that has been pre-trained on large datasets to perform tasks such as text generation and summarization.

[0476] A "smartphone" is a mobile phone equipped with internet connectivity and many applications.

[0477] A "head-mounted display" is a device worn on the head that displays visual information.

[0478] To implement this invention, the server, terminal, and user must work together. The details of how each part works are described below.

[0479] Server-side processing

[0480] The server first crawls multiple web pages on the Internet and analyzes their content. Specifically, it retrieves the content of the web pages using the requests library and analyzes the HTML content using BeautifulSoup. Based on the analyzed page content, it determines whether or not the pages contain sexual advertisements. The URLs of pages that contain sexual advertisements are stored in a database.

[0481] The server then trains a summarization model using natural language processing techniques, using the pipeline function in the transformers library to train a generative AI model on a large dataset, which is then used to summarize the content of web pages.

[0482] Terminal side processing

[0483] When a user attempts to access a specific web page in a browser, a browser extension on the device monitors the URL. This monitoring is performed using JavaScript and related browser extension APIs. The URL of the web page the user is attempting to access is sent to a server, which determines whether the URL is stored in its database. If the determination is that the page is problematic, a summary is sent from the server to the device, which then displays the summary to the user in a pop-up window.

[0484] User processing

[0485] Users can check the summary information displayed on their browser and decide whether to access the website based on the content. This allows users to quickly and safely access the information they need without being exposed to intrusive advertisements.

[0486] Specific examples

[0487] As a specific use case, consider the case where a user attempts to access the URL "http: / / example.com / sample-page." At this time, the browser extension sends this URL to the server. The server references the database and identifies this URL as a page containing sexually explicit advertisements. The generative AI model summarizes the page content and generates summary information such as, "This website provides reviews of image editing software. The key points of the reviews are that the user interface is easy to use and that a wide range of editing tools is available." This summary information is sent to the device and displayed as a pop-up window by the browser extension.

[0488] Prompt Sentence Examples

[0489] The following is an example of a prompt that should be given to the generative AI model regarding the URL the user attempted to access and a summary of its content:

[0490] URL: http: / / example.com / sample-page

[0491] Description: This website provides reviews of photo editing software. The main points of the reviews are that the user interface is easy to use and that there are a wide range of editing tools available.

[0492] In this way, the present invention provides users with access to necessary information while preventing them from being exposed to unintended sexual advertisements when browsing web pages on the Internet.By using devices such as smartphones and head-mounted displays, a comfortable browsing experience can be achieved on a variety of devices.

[0493] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0494] Step 1:

[0495] The server crawls multiple web pages on the Internet and retrieves their contents. Specifically, it retrieves the HTML content of each web page using the requests library and parses it using BeautifulSoup. The input is the URL of the web page, and the output is the content of the page, such as text and images.

[0496] Step 2:

[0497] The server determines whether the web page contains sexual advertisements based on the analyzed content. It uses specific keywords and image recognition algorithms to detect sexual content. The input is the web page content extracted in step 1, and the output is the result of determining whether the web page contains sexual advertisements.

[0498] Step 3:

[0499] The server stores the URLs of pages that are determined to contain sexually explicit ads in a database. The database records the URLs of problematic web pages and their verdicts. The input is the verdict and URL obtained in step 2, and the output is the records stored in the database.

[0500] Step 4:

[0501] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. This browser extension captures the URL entered by the user in real time and sends it to a server. The input is the URL the user is trying to access, and the output is the data sent to the server.

[0502] Step 5:

[0503] The server checks the received URL against a list of problem pages stored in a database. The input is the URL sent from the device, and the output is a decision as to whether the URL is a problematic page.

[0504] Step 6:

[0505] The server uses a generative AI model to create summaries for URLs that are determined to be problematic. Specifically, it summarizes the page content using the summarization function of the transformers library. The input is the determined web page content, and the output is a summary.

[0506] Step 7:

[0507] The server sends the summarized content to the terminal. The input is the generated summary sentence, and the output is the data sent to the terminal.

[0508] Step 8:

[0509] The terminal displays the received summary information to the user as a pop-up window. Specifically, it displays a pop-up on the browser using JavaScript. The input is the summary data received from the server, and the output is the pop-up window displayed to the user.

[0510] Step 9:

[0511] The user checks the displayed summary information and makes the final decision on whether to access it. The input is the summary information displayed in the popup window, and the output is the user's decision on whether to allow or deny access.

[0512] 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.

[0513] Overall program overview

[0514] This invention relates to a system that safely retrieves the contents of web pages on the Internet and presents appropriate information according to the user's emotional state. This system operates in cooperation with a server, a terminal, and a user, and provides information that takes into consideration the user's emotions, while restricting access to web pages that contain sexually explicit advertisements.

[0515] Server-side processing

[0516] The server first crawls multiple web pages on the Internet and analyzes their content. It uses natural language processing and image recognition technologies to identify whether they contain sexually explicit advertisements. It stores the URLs of web pages determined to contain sexually explicit advertisements in a database. The server then uses a large amount of data to train a natural language processing model. This model is then used to summarize the content of the web pages.

[0517] Furthermore, the server is equipped with an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from information such as the user's facial expressions, voice, and input text. This emotion data is saved for each user and accumulates over time.

[0518] Terminal side processing

[0519] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. The extension sends the URL to a server. The server receives the URL and checks it against a list of problem pages stored in a database. If a problem is detected, a summary is sent from the server to the device.

[0520] The device displays this summary information in a pop-up window. Furthermore, the emotion engine analyzes the user's real-time emotions and adjusts the way information is presented according to the user's emotional state. For example, if the user is in a negative emotional state, an appropriate warning message can be displayed.

[0521] User processing

[0522] Users can view summary information about web pages in their browsers. Based on this information, they can decide whether to access a particular web page. Furthermore, information provided by the emotion engine allows them to learn the best way to respond to their own emotional state. User emotion data is stored over the long term, and an algorithm is run to provide the best display method for each individual user.

[0523] Specific examples

[0524] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." A browser extension sends this URL to a server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model summarizes the content of this page and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..."

[0525] Furthermore, the emotion engine analyzes the user's emotions, and if the user is expressing negative emotions such as stress or anger, a warning message is generated stating, "This page currently contains inappropriate advertisements and is not recommended for viewing." This summary information and warning message are sent to the device and displayed by the browser extension as a pop-up window. Users feel that they have read this summary and obtained the desired information without being exposed to unpleasant advertisements.

[0526] This system configuration can provide a better user experience by preventing users from being exposed to unnecessary advertisements and by taking into consideration their emotional state. It is also expected to contribute to the soundness of internet advertising.

[0527] The processing flow will be explained below.

[0528] Step 1:

[0529] Server: Crawl web pages

[0530] The server periodically crawls pages on the Internet and retrieves the HTML source of new pages based on the specified URL list. This HTML source is saved for content analysis.

[0531] Step 2:

[0532] Server: Analyze content and identify problem pages

[0533] The server analyzes the content of the crawled pages. It uses natural language processing and image recognition technology to examine the ad banners and metadata on the pages. If any sexually explicit ads are found, the URL of the page is saved in a database as a problem page.

[0534] Step 3:

[0535] Server: Training the natural language processing model

[0536] The server trains a natural language processing model (such as BERT or GPT-3) using large amounts of text data, which is then used to summarize the content of web pages.

[0537] Step 4:

[0538] Server: Preparing the emotion engine

[0539] The server prepares the emotion engine and trains a model to analyze the user's facial expression data, voice data, input text, etc. This uses machine learning techniques to enable high-accuracy recognition of the user's emotional state.

[0540] Step 5:

[0541] Terminal: Monitoring web page requests

[0542] A browser extension installed on the user's device monitors the URLs the user attempts to access and sends this URL information to a server.

[0543] Step 6:

[0544] Server: URL rating

[0545] The server checks the received URL against its database, determines whether the URL is included in the list of problem pages, and returns the result.

[0546] Step 7:

[0547] Server: Generate summary information

[0548] If a page is determined to be problematic, the server uses a natural language processing model to summarize the content of the page, and then generates this summary information and sends it to the terminal.

[0549] Step 8:

[0550] Terminal: Acquisition and analysis of emotion data

[0551] The device's browser extension collects the user's facial expression data, voice data, input text, etc. in real time and sends it to the emotion engine, which analyzes this data and determines the user's emotional state.

[0552] Step 9:

[0553] Server: Sends summary information and generates warning messages

[0554] The server generates an appropriate warning message along with summary information based on the user's emotional state determined by the emotion engine, and transmits this information to the terminal.

[0555] Step 10:

[0556] Terminal: Display summary information and warning messages

[0557] The browser extension displays the received summary information and a warning message in a pop-up window, such as "This page contains inappropriate advertising. Page content summary: (summary text) Viewing is not recommended based on your current emotional state."

[0558] Step 11:

[0559] User: Review summary information and emotional feedback

[0560] Users can review the summary information and warning messages provided in the pop-up window, and with the necessary information, they can safely browse the web without being exposed to annoying ads.

[0561] Step 12:

[0562] User: Safe Web Browsing

[0563] Users can enjoy a better user experience by being protected from unnecessary advertisements and receiving information tailored to their emotional state.

[0564] Example 2

[0565] 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."

[0566] In today's internet usage environment, users often encounter inappropriate sexual advertisements when browsing web pages, significantly degrading the user experience. Furthermore, the lack of appropriate information presented in response to the user's emotional state can potentially increase the user's psychological burden. Conventional technologies have not provided a means to effectively solve these problems simultaneously, making it difficult to provide a comfortable browsing environment for users.

[0567] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for crawling multiple web pages on the Internet and analyzing the contents of those pages, a means for identifying web pages containing sexual advertisements based on the analyzed contents and saving the URLs of those pages in a recording device, and a means for receiving the URL of a web page that a user attempts to access and determining whether the URL is saved in the recording device. This makes it possible to determine in advance whether a web page that a user attempts to view contains inappropriate advertisements and provide the user with appropriate information.

[0568] The server further includes means for summarizing the content of web pages containing sexual advertisements, means for transmitting the summarized content to the user's information processing terminal and displaying it as a pop-up window, means for recognizing the user's emotions and storing the emotion data in a recording device, and means for adjusting the information presentation method according to the user's emotional state, thereby realizing information provision that takes into consideration the user's emotional state and improving the user experience.

[0569] Furthermore, by including a means for training a natural language processing model for summarization and an extension function for the information processing terminal that monitors the URLs of web pages that users are attempting to access, it becomes possible to provide more accurate summary information and monitor in real time, thereby realizing a safe and comfortable internet experience for users without being exposed to inappropriate advertisements.

[0570] A "web page" is a document that displays information that is publicly available on the Internet and is written in a language such as HTML.

[0571] "Crawling" is the process of automatically visiting specific web pages and retrieving their content.

[0572] "Analysis" refers to the processing of acquired data using computer algorithms to extract or evaluate necessary information.

[0573] "Sexual advertising" means commercial information containing sexual content that may have an inappropriate effect on users.

[0574] A "storage device" is a physical or virtual device for storing data, such as a database.

[0575] An "information processing terminal" is an electronic device that can process information, such as a computer, smartphone, or tablet.

[0576] A "pop-up window" is a small window that suddenly appears on a user's screen and is used to present specific information.

[0577] "Emotion" refers to a psychological state that is expressed through a person's facial expression, voice, input text, etc.

[0578] "Emotional data" refers to information about the user's psychological state, including facial expressions and voice analysis results.

[0579] A "natural language processing model" is a machine learning model for understanding and processing human language, and is used for tasks such as summarizing and translating text.

[0580] An "extension" is an additional program that extends the functionality of existing software and is used to enhance the functionality of a browser.

[0581] This invention is a system in which a server, terminals, and users work together to provide a safe and comfortable Internet browsing environment. Specifically, the server crawls multiple web pages on the Internet and analyzes the content of those pages. It also monitors the URLs of web pages that users attempt to access, summarizes the content of problematic pages, and displays a warning to the user in real time.

[0582] Server-side processing

[0583] The server first uses a web crawler (e.g., BeautifulSoup, Scrapy) to crawl multiple web pages on the Internet. It then obtains the HTML content of each page and analyzes it using natural language processing technology (e.g., spaCy) or image recognition technology (e.g., OpenCV). This analysis allows it to extract text and images from the web pages.

[0584] The extracted content is then evaluated using an AI model (e.g., a model using TensorFlow) to determine whether it contains sexually explicit content. Based on the results of the evaluation, the URLs of web pages that are found to contain sexually explicit content are stored in a storage device (e.g., a MySQL database).

[0585] Additionally, the server summarizes the content of the web page using a natural language processing model (e.g., BERT), which is pre-trained using a large amount of text data.

[0586] In addition, the server uses an emotion engine (e.g., Microsoft Azure Cognitive Services) to recognize the user's emotions. Emotion data is stored in a recording device for each user.

[0587] Terminal side processing

[0588] A browser extension (e.g., Chrome Extension) is installed on the device, and when a user attempts to access a specific web page, it sends the URL to the server. The server receives the URL and compares it with a list of problem pages stored in a recording device. For URLs determined to be problematic, the server generates summary information and a warning message and sends them to the device.

[0589] The terminal displays the received summary information and warning message in a pop-up window, allowing the user to obtain information without being exposed to inappropriate content.

[0590] User processing

[0591] The user checks the displayed summary information and warning message and decides whether to access the web page. The emotion engine also analyzes the user's real-time emotions and sends the results to the server, which then further optimizes the way information is presented in the future.

[0592] Specific examples

[0593] For example, if a user attempts to access the URL "example-adultcontent.com," the device's browser extension sends the URL to the server. The server compares it with information in the database and determines that the URL contains inappropriate advertising. It generates a summary such as "This website provides reviews of image editing software. The gist of the review is..." Furthermore, the emotion engine analyzes the user's emotions, and if the user is feeling stressed or anxious, it generates a warning message saying, "This page contains inappropriate advertising. Viewing it is not recommended."

[0594] An example of the prompt that the user sees:

[0595] "Users submit the URL of the website they are trying to access. Additionally, they enter their current emotional state. For example, the URL 'example-adultcontent.com' and the emotion 'I'm stressed.'"

[0596] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0597] Step 1:

[0598] The server uses a web crawler to crawl multiple web pages on the Internet. The web crawler (e.g., BeautifulSoup, Scrapy) is configured to retrieve the HTML content of all pages in a specified domain. The input to this process is a list of URLs, and the output is the HTML code for each web page.

[0599] Step 2:

[0600] The server analyzes the acquired HTML code and extracts text and images. This analysis uses natural language processing technology (e.g., spaCy) and image recognition technology (e.g., OpenCV). The input is the HTML code, and the output is text data and image data. The extracted text and images are evaluated in the next step.

[0601] Step 3:

[0602] The server inputs the extracted text data and image data into an AI model (e.g., a model built using TensorFlow) to determine whether or not the data contains sexual advertisements. The input is text data and image data, and the output is a determination result of "contains / does not contain sexual advertisements." Based on this result, the determined URLs are stored in a recording device (e.g., a MySQL database) as those containing sexual advertisements.

[0603] Step 4:

[0604] The server uses a natural language processing model (e.g., BERT) to summarize the content of a web page. This model has been trained on a large amount of text data in advance. The input is the text data of the web page, and the output is the summary text. This summary text is saved for presentation to the user.

[0605] Step 5:

[0606] The server uses an emotion engine (e.g., Microsoft Azure Cognitive Services) to analyze the user's emotions. Inputs include the user's facial expressions, voice, and input text, and the output is the user's emotional data (e.g., "feeling stressed," "relaxed," etc.). This emotional data is stored in a recording device for each user.

[0607] Step 6:

[0608] A browser extension (e.g., Chrome Extension) is installed on the device, and when a user tries to access a specific web page, it monitors the URL and sends it to a server. The input is the URL that the user is trying to access, and the output is that the URL is sent to the server.

[0609] Step 7:

[0610] The server checks the received URL against the problem page list stored in the recording device. For URLs that are determined to be problematic, the server generates a warning message based on the previously generated summary text and the user's emotional data. The input is the URL to be accessed and the user's emotional data, and the output is the summary text and the warning message.

[0611] Step 8:

[0612] The terminal displays the summary information and warning message received from the server as a pop-up window. If a user attempts to access "example-adultcontent.com", it displays the message "This site contains inappropriate advertisements. Viewing is not recommended." The input is the summary text and warning message sent from the server, and the output is the display in the pop-up window.

[0613] Step 9:

[0614] The user checks the displayed summary information and warning message and decides whether to access the web page. At the same time, the emotion engine analyzes the user's real-time emotions and sends the results to the server. The input is the user's decision and real-time emotion data, and the output is whether to access the page and updated emotion data.

[0615] (Application example 2)

[0616] 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."

[0617] Currently, when browsing the Internet, there is a risk of accessing inappropriate advertisements or content. Furthermore, information is provided in a uniform manner without considering the user's emotional state, resulting in a suboptimal user experience. Furthermore, users may experience stress or anxiety when exposed to inappropriate content. There is a need to resolve these issues and provide a safer, more user-friendly Internet browsing environment.

[0618] 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.

[0619] In this invention, the server includes means for crawling multiple web pages on the Internet and analyzing the contents of those pages, means for identifying web pages containing inappropriate advertisements based on the analyzed contents and storing the URLs of those pages in a database, means for receiving the URL of a web page that a user attempts to access and determining whether the URL is stored in the database, means for summarizing the contents of the web page containing the inappropriate advertisement, means for analyzing the user's emotional state and generating a warning message or alternative information according to the emotional state, and means for transmitting the summarized contents and information according to the user's emotions to the user's terminal and displaying them as a pop-up window. This makes it possible to provide optimal information according to the user's emotional state while preventing the user from being exposed to inappropriate advertisements or content.

[0620] "Web pages on the Internet" refers to individual pages of publicly accessible websites connected to the Internet.

[0621] "Crawl" refers to the technical process of automatically visiting designated web pages and collecting their content.

[0622] "Content analysis" refers to the process of understanding collected information, such as text and images, from web pages and extracting specific patterns and features.

[0623] "Inappropriate Advertisements" refers to advertisements that may be offensive to users, especially those that contain sexual or violent content.

[0624] "Identifying web pages" refers to the process of finding and identifying web pages that meet certain criteria based on their analyzed content.

[0625] "Means for storing in a database" refers to the process of storing the URLs of the identified web pages and other related information in a dedicated database.

[0626] "Means for receiving the URL of the web page that the user wishes to access" refers to the process of obtaining the address information of the web page that the user has entered or selected.

[0627] "Means for determining" refers to the process of checking a received URL against a list of inappropriate web pages in a database to determine whether it is an inappropriate page.

[0628] "Means for summarizing content" refers to the process of providing a brief summary of the content of a particular Web page.

[0629] "Means for analyzing emotional state" refers to the technical process of recognizing and analyzing the user's current emotions from data such as facial expressions, voice, and input text.

[0630] "Warning Message" refers to a notification that alerts a user to the risk of a particular action or situation.

[0631] "Alternate Information" refers to safe and appropriate information that is provided in place of the page a user is attempting to access.

[0632] "Pop-up window" refers to a small window that appears floating on a user's screen and displays information or a message.

[0633] "User's Device" refers to the electronic device used by the User, such as a smartphone, tablet, or PC.

[0634] "System" refers to a collection of devices and programs that combine all of the above means and have the overall function.

[0635] This invention is a system for safely removing inappropriate advertisements on web pages and providing users with information optimized for them. This system functions in cooperation with a server, a terminal, and a user.

[0636] Server-side processing

[0637] The server uses a crawling engine and a natural language processing engine to crawl multiple web pages on the Internet and analyze their content. It uses web scraping tools such as Python's requests library and BeautifulSoup library for the analysis. It also uses machine learning models to identify web pages that contain inappropriate ads based on the analyzed content. This information is stored in a database for later use.

[0638] The server then uses generative AI models such as Google's Text-To-Text Transfer Transformer (T5) and Bidirectional Encoder Representations from Transformers (BERT) to summarize the content of the webpage, and then uses Microsoft's Azure Emotion API and DeepFace library to apply facial and speech recognition technologies to analyze the user's emotional state.

[0639] Terminal side processing

[0640] When a user attempts to access a web page, the browser extension on the device monitors the URL and sends it to a server. The server checks the URL against a database to determine if it contains inappropriate content. If it does, a server-generated summary is sent to the device and displayed in a pop-up window.

[0641] The device also uses a camera and microphone to analyze the user's emotions in real time, and generates warning messages and alternative information based on the results of this emotion analysis.

[0642] User processing

[0643] Users use their browsers to view web page summaries and warning messages, and use this information to decide whether to access a particular web page. User emotion data is stored over the long term, and this data can be used to provide information optimized for each individual user.

[0644] Specific examples

[0645] For example, consider a situation where a user attempts to access the URL "https: / / example-adultcontent.com." The browser extension sends this URL to a server, which then consults a database. If the URL is determined to be a page containing inappropriate advertising, the AI ​​summarization model summarizes the page's content and generates a summary that reads, "This website contains inappropriate advertising, but its content is a review of image editing software."

[0646] In addition, the emotion engine analyzes the user's facial expressions, and if a negative emotion is detected, a warning message is generated stating, "This page currently contains inappropriate advertisements and is not recommended for viewing." This information and warning message are sent to the device and displayed as a pop-up window, allowing the user to obtain the desired information without being exposed to unpleasant advertisements.

[0647] Use the following as an example prompt:

[0648] test_url = "https: / / example-adultcontent.com"

[0649] test_image_path = "path / to / user_image.jpg"

[0650] main(test_url, test_image_path)

[0651] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0652] Step 1:

[0653] The server crawls multiple web pages on the Internet. During this process, it uses a specific algorithm to generate a list of web page URLs, and then uses Python's requests library and BeautifulSoup library to retrieve the content of each web page. The HTML data of the retrieved web pages is used as input, and the analysis results are output.

[0654] Step 2:

[0655] The server analyzes the content of the retrieved web pages. This analysis uses a natural language processing engine and an image recognition engine to process the text and image data of each web page. A machine learning model is used to determine whether the web page contains inappropriate advertising. The input data is the text and images of the web page, and the output is a determination of whether the web page contains inappropriate advertising.

[0656] Step 3:

[0657] The server identifies web pages containing inappropriate ads based on the analysis results and stores the URLs of these pages in a dedicated database. The stored information includes the URL of the web page and the reason why it was determined to be inappropriate. The output is to store a list of identified URLs in the database.

[0658] Step 4:

[0659] When a user accesses a web page using a browser on their device, a browser extension on the device monitors the URL. It captures the URL the user types in the address bar and sends it to a server. This URL is the input data, and the sent URL is the output.

[0660] Step 5:

[0661] The server checks the received URL against a database to determine whether it contains inappropriate content. Database search technology is used to check the existence of the URL. The input is the URL sent by the user, and the output is the result of whether the URL is inappropriate.

[0662] Step 6:

[0663] If the server determines that the webpage contains inappropriate ads, it uses an AI generative model such as Google's T5 or BERT to summarize the content of the webpage. The generative AI model takes the text data of the webpage as input and outputs a concise summary text.

[0664] Step 7:

[0665] The server receives real-time captured images of the user's facial expressions or voice data to analyze the user's emotional state. Using this as input data, it determines the user's emotional state using an emotion analysis engine (DeepFace or Microsoft Azure Emotion API). The analysis result of the emotional state is obtained as output.

[0666] Step 8:

[0667] The server generates a warning message or alternative information according to the user's emotional state based on the emotion analysis results. The input data is the emotion analysis results, and the output is an appropriate warning message or alternative information.

[0668] Step 9:

[0669] The server sends the summarized content and emotion-based information to the user's terminal and displays them as a pop-up window on the terminal. The input data are the summary information and the emotion-based message, and the output is transmission to the user's terminal and display of the pop-up window.

[0670] 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.

[0671] 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.

[0672] 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.

[0673] [Third embodiment]

[0674] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0675] 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.

[0676] 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).

[0677] 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.

[0678] 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.

[0679] 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).

[0680] 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. 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.

[0681] 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.

[0682] 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.

[0683] 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.

[0684] 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.

[0685] 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."

[0686] Overall program overview

[0687] This invention relates to a system for securely acquiring the contents of web pages on the Internet. This system involves the cooperation of a server, a terminal, and a user, and aims to restrict access to web pages that contain sexually explicit advertisements.

[0688] Server-side processing

[0689] The server first crawls multiple web pages on the Internet. It analyzes the content of the crawled web pages to identify whether they contain sexually explicit advertisements. The URLs of the identified web pages are stored in a database. When analyzing this content, the server uses natural language processing and image recognition technologies to evaluate the content of the advertisements.

[0690] The server then uses the massive dataset to train a natural language processing model. The trained AI model is then used to summarize the content of the web page. This summarization model is designed to quickly generate a summary of the web page in question. The summarized information is also stored in a database, ready to be served immediately if needed.

[0691] Terminal side processing

[0692] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. The extension sends the URL to a server, which receives the URL and checks it against a list of problem pages stored in a database. If a problem is detected, a summary is sent from the server to the device.

[0693] The device displays a pop-up window to provide summary information to the user. The pop-up window displays, "This page contains inappropriate advertisements. Summary of page contents: (Summary text)." Based on this summary information, the user can obtain the necessary information without being exposed to unpleasant advertisements.

[0694] User processing

[0695] Users can view web page summary information in their browsers and use this information to decide whether to visit a particular web page. This allows users to easily access the information they want while avoiding pages that contain sexually explicit advertisements.

[0696] Specific examples

[0697] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." A browser extension sends this URL to a server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model summarizes the content of this page and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..."

[0698] This summary information is sent to the device and displayed as a pop-up window by the browser extension. The user reads the summary and feels that they have obtained the desired information without being exposed to annoying advertisements. In this way, the system provides users with access to the information they need while preventing them from being exposed to unwanted advertisements.

[0699] This system configuration promotes the soundness of Internet advertising and provides users with a safe and comfortable web browsing environment.

[0700] The processing flow will be explained below.

[0701] Step 1:

[0702] Server: Crawl web pages

[0703] The server automatically collects pages on the Internet, crawls new pages based on a periodically updated URL list, obtains the HTML source of the crawled pages, and stores it for content analysis.

[0704] Step 2:

[0705] Server: Content analysis and identification

[0706] The server analyzes the content of the crawled pages, using natural language processing and image recognition technology to examine the ad banners and metadata on the pages, determining whether they contain sexually explicit content and storing the URLs of problematic pages in a database.

[0707] Step 3:

[0708] Server: Training the natural language processing model

[0709] The server trains a natural language processing model (such as BERT or GPT-3) using a large amount of text data. The training data includes high-quality summaries and feedback. This model is then used to summarize the content of web pages.

[0710] Step 4:

[0711] Terminal: Monitoring web page requests

[0712] A browser extension installed on the user's device monitors the URLs the user attempts to access and sends this URL information to a server.

[0713] Step 5:

[0714] Server: URL rating

[0715] The server checks the received URL against a database to see if the URL exists and whether it is a page containing sexual advertisements.

[0716] Step 6:

[0717] Server: Generate summary information

[0718] If the page is determined to be relevant, the server uses a natural language processing model to summarize the content of the page, generates this summary information, and sends it to the terminal.

[0719] Step 7:

[0720] Terminal: Display summary information

[0721] The browser extension displays the received summary information in a popup window, displaying the message "This page contains inappropriate advertising. Summary of page contents: (summary text)" to the user.

[0722] Step 8:

[0723] User: Review summary information and make a decision

[0724] The user reviews the summary information provided in the pop-up window, allowing them to decide whether to obtain the information they need without being exposed to the intrusive advertisement.

[0725] Step 9:

[0726] User: Safe Web Browsing

[0727] Users can safely obtain the information they are looking for while avoiding inappropriate ads, improving the user experience and promoting the integrity of internet advertising.

[0728] Example 1

[0729] 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."

[0730] On the Internet, there are web pages containing inappropriate content, and accessing them can lead to unpleasant experiences or expose users to harmful information. Web pages containing sexually explicit advertisements are particularly undesirable for users. To address this, it is necessary to provide an environment in which users can browse the web with peace of mind. However, existing filtering systems have the problem of being unable to completely filter out harmful content.

[0731] 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.

[0732] In this invention, the server includes means for crawling multiple web pages on the Internet and analyzing the contents of those pages, means for identifying sexual advertisements based on the analyzed contents and saving the addresses of those pages in data storage means, means for receiving the addresses of web pages that a user wishes to access and determining whether the addresses are saved in the data storage means, means for summarizing the contents of web pages that include sexual advertisements using natural language processing, and means for transmitting the summarized contents to the user's information processing device and displaying them as a pop-up window. This allows users to efficiently access necessary information while reducing the risk of accessing web pages that include sexual advertisements.

[0733] "Crawling" is the process of automatically visiting many web pages on the Internet to collect information.

[0734] "Content analysis" refers to the act of extracting text and image information from collected data and evaluating and classifying it for a specific purpose.

[0735] "Sexual advertising" refers to advertising displays that contain sexual content and are considered offensive or harmful to users.

[0736] An "address" is a Uniform Resource Locator (URL) that identifies a specific web page on the Internet.

[0737] "Data storage means" refers to a database or storage device that stores analyzed data and allows the information to be quickly referenced later.

[0738] "Receiving" is the act of receiving data or information sent from an external source.

[0739] "Natural language processing" refers to the technology and methods that allow computers to understand and generate human language.

[0740] "Summarizing" is the act of concisely summarizing long text or complex information and extracting only the important points.

[0741] An "information processing device" is a device such as a computer or smartphone used by a user.

[0742] A "pop-up window" is a small window that appears within a user interface and is used to display information such as notifications or alerts.

[0743] An "extension" is a software module that adds functionality to a browser or other application.

[0744] MODE FOR CARRYING OUT THE INVENTION

[0745] This invention relates to a system for securely retrieving the contents of web pages on the Internet. In particular, it aims to restrict access to web pages containing sexually explicit advertisements, thereby preventing users from viewing inappropriate content. This system operates in cooperation with a server, a terminal, and a user.

[0746] Server-side processing

[0747] The server uses existing web crawling tools, such as Python's BeautifulSoup and Scrapy, to crawl the web. It collects the HTML content of each crawled web page and stores it in temporary storage. The server then applies natural language processing (NLP) techniques using TensorFlow and PyTorch to analyze the text data within the web page. At the same time, it also uses OpenCV and TensorFlow for image recognition to determine whether the web page contains sexually explicit advertisements.

[0748] The server trains a natural language processing model using a huge dataset (e.g., Wikipedia, Common Crawl). The training process is streamlined by using a high-performance GPU (e.g., Nvidia Tesla). This trained AI model is then used to summarize the content of the web page in question. The analysis results, summary information, and URLs are stored in a database. Databases such as MySQL and PostgreSQL are used.

[0749] Terminal side processing

[0750] When a user attempts to access a specific web page, the device utilizes a browser extension (e.g., Google Chrome Extension). This extension is implemented in JavaScript and captures the user's navigation events to obtain the URL. This URL is then sent to the server as an HTTPS request. The server checks the received URL against a list of problem pages in a database, and if a problem is detected, it sends a summary of the URL to the device.

[0751] The device displays a pop-up window based on the summary information received from the server. The pop-up window displays the message "This page contains inappropriate advertisements. Summary of page content: (Summary text)." This allows users to avoid inappropriate content.

[0752] User processing

[0753] Users can decide whether to access a particular web page based on the summary information displayed in their browser. This information allows them to safely and efficiently obtain the information they need. This also allows users to use the Internet safely, avoiding anxiety and unpleasant experiences.

[0754] Specific examples

[0755] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." The browser extension sends this URL to the server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model then summarizes the page's contents and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..." This summary is then sent to the device and displayed by the browser extension in a pop-up window. The user can read the summary and obtain the desired information without being exposed to any offensive advertisements.

[0756] Prompt Sentence Examples

[0757] "Summarize the content of a given web page URL, determine whether it contains sexual advertisements, and notify the user if necessary."

[0758] This system will promote the integrity of Internet advertising and provide users with a safe and comfortable web browsing environment.

[0759] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0760] Processing step flow

[0761] Step 1: Crawl the web page (server)

[0762] The server crawls multiple web pages on the Internet using a web crawler tool such as Python's BeautifulSoup or Scrapy. The input is the URL of the web page to be crawled, and the output is data including the HTML content. This data is stored in temporary storage.

[0763] Specifically, the server runs a Python script, and the web crawler tool crawls through the specified URL list to collect HTML data, which is then stored in local storage.

[0764] Step 2: Analyzing the web page content (server)

[0765] The server analyzes the collected HTML data and extracts text and images. The analysis uses frameworks such as TensorFlow and PyTorch. The HTML content is input, and the extracted text data and image URLs are output.

[0766] Specifically, the server parses the HTML using BeautifulSoup to extract all text content, and also extracts and lists URLs for image tags. This extracted data is used in the next analysis step.

[0767] Step 3: Identifying Sexual Ads (Server)

[0768] The server analyzes the extracted text data and image URLs to identify sexually explicit advertisements. This is done using natural language processing and image recognition technology. The extracted text data and image URLs are used as input, and the server outputs the identification of web pages containing sexually explicit advertisements and their URLs.

[0769] Specifically, the server uses an NLP model to evaluate text data and detect sexually explicit terms and expressions. At the same time, it uses an image recognition model to evaluate images retrieved from extracted image URLs. Based on the identification results, the webpage URLs are added to a problem page list.

[0770] Step 4: Training the summary model (server)

[0771] The server trains a natural language processing model using a large text dataset, taking the text dataset as input and the trained summarization model as output, which is used to quickly and effectively summarize web page content.

[0772] Specifically, the server loads a text dataset and trains a model using a natural language processing framework (e.g., TensorFlow), with the training process being streamlined using high-performance GPU hardware.

[0773] Step 5: Monitor the URL (Device)

[0774] A browser extension installed on a user's device monitors the URLs of web pages accessed by the user. The monitored URLs are used as input, and a request to send the URL to a server is used as output.

[0775] Specifically, the browser extension uses JavaScript scripts to capture user navigation events and retrieve the URL to be visited, which is then sent to the server as an HTTPS request.

[0776] Step 6: Identify the URL and send the summary information (server)

[0777] The server compares the received URL with a list of problem pages in the database, and if it determines that there is a problem, it generates and sends summary information. The received URL is the input, and the summary information is the output.

[0778] Specifically, the server performs a database query to check whether the URL is included in the problem page list, and if so, summarizes the page content using a trained summarization model and sends the summary information in JSON format to the device.

[0779] Step 7: View summary information (terminal)

[0780] The terminal displays a pop-up window based on the summary information received from the server. The received summary information is used as input, and the display in the pop-up window is used as output.

[0781] Specifically, the browser extension uses HTML and CSS to generate a popup window and display summary information. JavaScript DOM manipulation causes the popup to appear on the user's screen. The popup displays the following message: "This page contains inappropriate advertisements. Summary of page contents: (Summary text)."

[0782] Step 8: User decision (User)

[0783] The user checks the summary information displayed in the pop-up window and decides whether to access the page. The summary information is the input, and the action that determines whether to access the page is the output.

[0784] Specifically, users view the pop-up, examine its contents, and, if necessary, close the browser tab or visit another secure site.

[0785] In this way, a system is configured in which the server, terminal, and user work together in sequence to safely obtain the contents of web pages on the Internet.

[0786] (Application example 1)

[0787] 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."

[0788] The goal is to prevent users from being exposed to unintended sexual advertisements when browsing web pages on the Internet, while at the same time providing a means for them to quickly and safely access the information they need. Furthermore, by using dedicated applications, it is necessary to realize a comfortable browsing experience on a variety of devices, including smartphones and head-mounted displays.

[0789] 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.

[0790] In this invention, the server includes a means for crawling multiple web pages on the Internet and analyzing the content of those pages, a means for identifying web pages containing sexual advertisements based on the analyzed content and storing the URLs of those pages in a database, and a means for receiving the URL of a web page a user attempts to access and determining whether the URL is stored in the database. This allows users to enjoy safe and comfortable web browsing without being exposed to inappropriate advertisements. Furthermore, a dedicated application that can be installed on a smartphone or head-mounted display can monitor accessed URLs and provide summaries of the web page content using a generative AI model.

[0791] The "Internet" is a network that interconnects computers and networks around the world, enabling the exchange of information.

[0792] "Crawling" is a technique for automatically crawling through web pages on the Internet and collecting information.

[0793] "Means for analyzing the content of web pages" refers to technology that extracts data such as text and images from web pages and uses them to understand and evaluate their content.

[0794] "Sexual advertising" refers to advertisements that appear on web pages and contain sexual content.

[0795] A "database" is a system for systematically storing and managing data.

[0796] "Means for receiving the URL of the web page that the user wishes to access" refers to a technique for obtaining the URL of a specific web page when the user accesses that web page.

[0797] A "pop-up window" is a small window that temporarily appears on a user's device to provide important information or warnings.

[0798] A "browser extension" is a software component that adds or extends the functionality of a web browser.

[0799] A "generative AI model" is an artificial intelligence model that has been pre-trained on large datasets to perform tasks such as text generation and summarization.

[0800] A "smartphone" is a mobile phone equipped with internet connectivity and many applications.

[0801] A "head-mounted display" is a device worn on the head that displays visual information.

[0802] To implement this invention, the server, terminal, and user must work together. The details of how each part works are described below.

[0803] Server-side processing

[0804] The server first crawls multiple web pages on the Internet and analyzes their content. Specifically, it retrieves the content of the web pages using the requests library and analyzes the HTML content using BeautifulSoup. Based on the analyzed page content, it determines whether or not the pages contain sexual advertisements. The URLs of pages that contain sexual advertisements are stored in a database.

[0805] The server then trains a summarization model using natural language processing techniques, using the pipeline function in the transformers library to train a generative AI model on a large dataset, which is then used to summarize the content of web pages.

[0806] Terminal side processing

[0807] When a user attempts to access a specific web page in a browser, a browser extension on the device monitors the URL. This monitoring is performed using JavaScript and related browser extension APIs. The URL of the web page the user is attempting to access is sent to a server, which determines whether the URL is stored in its database. If the determination is that the page is problematic, a summary is sent from the server to the device, which then displays the summary to the user in a pop-up window.

[0808] User processing

[0809] Users can check the summary information displayed on their browser and decide whether to access the website based on the content. This allows users to quickly and safely access the information they need without being exposed to intrusive advertisements.

[0810] Specific examples

[0811] As a specific use case, consider the case where a user attempts to access the URL "http: / / example.com / sample-page." At this time, the browser extension sends this URL to the server. The server references the database and identifies this URL as a page containing sexually explicit advertisements. The generative AI model summarizes the page content and generates summary information such as, "This website provides reviews of image editing software. The key points of the reviews are that the user interface is easy to use and that a wide range of editing tools is available." This summary information is sent to the device and displayed as a pop-up window by the browser extension.

[0812] Prompt Sentence Examples

[0813] The following is an example of a prompt that should be given to the generative AI model regarding the URL the user attempted to access and a summary of its content:

[0814] URL: http: / / example.com / sample-page

[0815] Description: This website provides reviews of photo editing software. The main points of the reviews are that the user interface is easy to use and that there are a wide range of editing tools available.

[0816] In this way, the present invention provides users with access to necessary information while preventing them from being exposed to unintended sexual advertisements when browsing web pages on the Internet.By using devices such as smartphones and head-mounted displays, a comfortable browsing experience can be achieved on a variety of devices.

[0817] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0818] Step 1:

[0819] The server crawls multiple web pages on the Internet and retrieves their contents. Specifically, it retrieves the HTML content of each web page using the requests library and parses it using BeautifulSoup. The input is the URL of the web page, and the output is the content of the page, such as text and images.

[0820] Step 2:

[0821] The server determines whether the web page contains sexual advertisements based on the analyzed content. It uses specific keywords and image recognition algorithms to detect sexual content. The input is the web page content extracted in step 1, and the output is the result of determining whether the web page contains sexual advertisements.

[0822] Step 3:

[0823] The server stores the URLs of pages that are determined to contain sexually explicit ads in a database. The database records the URLs of problematic web pages and their verdicts. The input is the verdict and URL obtained in step 2, and the output is the records stored in the database.

[0824] Step 4:

[0825] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. This browser extension captures the URL entered by the user in real time and sends it to a server. The input is the URL the user is trying to access, and the output is the data sent to the server.

[0826] Step 5:

[0827] The server checks the received URL against a list of problem pages stored in a database. The input is the URL sent from the device, and the output is a decision as to whether the URL is a problematic page.

[0828] Step 6:

[0829] The server uses a generative AI model to create summaries for URLs that are determined to be problematic. Specifically, it summarizes the page content using the summarization function of the transformers library. The input is the determined web page content, and the output is a summary.

[0830] Step 7:

[0831] The server sends the summarized content to the terminal. The input is the generated summary sentence, and the output is the data sent to the terminal.

[0832] Step 8:

[0833] The terminal displays the received summary information to the user as a pop-up window. Specifically, it displays a pop-up on the browser using JavaScript. The input is the summary data received from the server, and the output is the pop-up window displayed to the user.

[0834] Step 9:

[0835] The user checks the displayed summary information and makes the final decision on whether to access it. The input is the summary information displayed in the popup window, and the output is the user's decision on whether to allow or deny access.

[0836] 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.

[0837] Overall program overview

[0838] This invention relates to a system that safely retrieves the contents of web pages on the Internet and presents appropriate information according to the user's emotional state. This system operates in cooperation with a server, a terminal, and a user, and provides information that takes into consideration the user's emotions, while restricting access to web pages that contain sexually explicit advertisements.

[0839] Server-side processing

[0840] The server first crawls multiple web pages on the Internet and analyzes their content. It uses natural language processing and image recognition technologies to identify whether they contain sexually explicit advertisements. It stores the URLs of web pages determined to contain sexually explicit advertisements in a database. The server then uses a large amount of data to train a natural language processing model. This model is then used to summarize the content of the web pages.

[0841] Furthermore, the server is equipped with an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from information such as the user's facial expressions, voice, and input text. This emotion data is saved for each user and accumulates over time.

[0842] Terminal side processing

[0843] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. The extension sends the URL to a server. The server receives the URL and checks it against a list of problem pages stored in a database. If a problem is detected, a summary is sent from the server to the device.

[0844] The device displays this summary information in a pop-up window. Furthermore, the emotion engine analyzes the user's real-time emotions and adjusts the way information is presented according to the user's emotional state. For example, if the user is in a negative emotional state, an appropriate warning message can be displayed.

[0845] User processing

[0846] Users can view summary information about web pages in their browsers. Based on this information, they can decide whether to access a particular web page. Furthermore, information provided by the emotion engine allows them to learn the best way to respond to their own emotional state. User emotion data is stored over the long term, and an algorithm is run to provide the best display method for each individual user.

[0847] Specific examples

[0848] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." A browser extension sends this URL to a server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model summarizes the content of this page and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..."

[0849] Furthermore, the emotion engine analyzes the user's emotions, and if the user is expressing negative emotions such as stress or anger, a warning message is generated stating, "This page currently contains inappropriate advertisements and is not recommended for viewing." This summary information and warning message are sent to the device and displayed by the browser extension as a pop-up window. Users feel that they have read this summary and obtained the desired information without being exposed to unpleasant advertisements.

[0850] This system configuration can provide a better user experience by preventing users from being exposed to unnecessary advertisements and by taking into consideration their emotional state. It is also expected to contribute to the soundness of internet advertising.

[0851] The processing flow will be explained below.

[0852] Step 1:

[0853] Server: Crawl web pages

[0854] The server periodically crawls pages on the Internet and retrieves the HTML source of new pages based on the specified URL list. This HTML source is saved for content analysis.

[0855] Step 2:

[0856] Server: Analyze content and identify problem pages

[0857] The server analyzes the content of the crawled pages. It uses natural language processing and image recognition technology to examine the ad banners and metadata on the pages. If any sexually explicit ads are found, the URL of the page is saved in a database as a problem page.

[0858] Step 3:

[0859] Server: Training the natural language processing model

[0860] The server trains a natural language processing model (such as BERT or GPT-3) using large amounts of text data, which is then used to summarize the content of web pages.

[0861] Step 4:

[0862] Server: Preparing the emotion engine

[0863] The server prepares the emotion engine and trains a model to analyze the user's facial expression data, voice data, input text, etc. This uses machine learning techniques to enable high-accuracy recognition of the user's emotional state.

[0864] Step 5:

[0865] Terminal: Monitoring web page requests

[0866] A browser extension installed on the user's device monitors the URLs the user attempts to access and sends this URL information to a server.

[0867] Step 6:

[0868] Server: URL rating

[0869] The server checks the received URL against its database, determines whether the URL is included in the list of problem pages, and returns the result.

[0870] Step 7:

[0871] Server: Generate summary information

[0872] If a page is determined to be problematic, the server uses a natural language processing model to summarize the content of the page, and then generates this summary information and sends it to the terminal.

[0873] Step 8:

[0874] Terminal: Acquisition and analysis of emotion data

[0875] The device's browser extension collects the user's facial expression data, voice data, input text, etc. in real time and sends it to the emotion engine, which analyzes this data and determines the user's emotional state.

[0876] Step 9:

[0877] Server: Sends summary information and generates warning messages

[0878] The server generates an appropriate warning message along with summary information based on the user's emotional state determined by the emotion engine, and transmits this information to the terminal.

[0879] Step 10:

[0880] Terminal: Display summary information and warning messages

[0881] The browser extension displays the received summary information and a warning message in a pop-up window, such as "This page contains inappropriate advertising. Page content summary: (summary text) Viewing is not recommended based on your current emotional state."

[0882] Step 11:

[0883] User: Review summary information and emotional feedback

[0884] Users can review the summary information and warning messages provided in the pop-up window, and with the necessary information, they can safely browse the web without being exposed to annoying ads.

[0885] Step 12:

[0886] User: Safe Web Browsing

[0887] Users can enjoy a better user experience by being protected from unnecessary advertisements and receiving information tailored to their emotional state.

[0888] Example 2

[0889] 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."

[0890] In today's internet usage environment, users often encounter inappropriate sexual advertisements when browsing web pages, significantly degrading the user experience. Furthermore, the lack of appropriate information presented in response to the user's emotional state can potentially increase the user's psychological burden. Conventional technologies have not provided a means to effectively solve these problems simultaneously, making it difficult to provide a comfortable browsing environment for users.

[0891] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for crawling multiple web pages on the Internet and analyzing the contents of those pages, a means for identifying web pages containing sexual advertisements based on the analyzed contents and saving the URLs of those pages in a recording device, and a means for receiving the URL of a web page that a user attempts to access and determining whether the URL is saved in the recording device. This makes it possible to determine in advance whether a web page that a user attempts to view contains inappropriate advertisements and provide the user with appropriate information.

[0892] The server further includes means for summarizing the content of web pages containing sexual advertisements, means for transmitting the summarized content to the user's information processing terminal and displaying it as a pop-up window, means for recognizing the user's emotions and storing the emotion data in a recording device, and means for adjusting the information presentation method according to the user's emotional state, thereby realizing information provision that takes into consideration the user's emotional state and improving the user experience.

[0893] Furthermore, by including a means for training a natural language processing model for summarization and an extension function for the information processing terminal that monitors the URLs of web pages that users are attempting to access, it becomes possible to provide more accurate summary information and monitor in real time, thereby realizing a safe and comfortable internet experience for users without being exposed to inappropriate advertisements.

[0894] A "web page" is a document that displays information that is publicly available on the Internet and is written in a language such as HTML.

[0895] "Crawling" is the process of automatically visiting specific web pages and retrieving their content.

[0896] "Analysis" refers to the processing of acquired data using computer algorithms to extract or evaluate necessary information.

[0897] "Sexual advertising" means commercial information containing sexual content that may have an inappropriate effect on users.

[0898] A "storage device" is a physical or virtual device for storing data, such as a database.

[0899] An "information processing terminal" is an electronic device that can process information, such as a computer, smartphone, or tablet.

[0900] A "pop-up window" is a small window that suddenly appears on a user's screen and is used to present specific information.

[0901] "Emotion" refers to a psychological state that is expressed through a person's facial expression, voice, input text, etc.

[0902] "Emotional data" refers to information about the user's psychological state, including facial expressions and voice analysis results.

[0903] A "natural language processing model" is a machine learning model for understanding and processing human language, and is used for tasks such as summarizing and translating text.

[0904] An "extension" is an additional program that extends the functionality of existing software and is used to enhance the functionality of a browser.

[0905] This invention is a system in which a server, terminals, and users work together to provide a safe and comfortable Internet browsing environment. Specifically, the server crawls multiple web pages on the Internet and analyzes the content of those pages. It also monitors the URLs of web pages that users attempt to access, summarizes the content of problematic pages, and displays a warning to the user in real time.

[0906] Server-side processing

[0907] The server first uses a web crawler (e.g., BeautifulSoup, Scrapy) to crawl multiple web pages on the Internet. It then obtains the HTML content of each page and analyzes it using natural language processing technology (e.g., spaCy) or image recognition technology (e.g., OpenCV). This analysis allows it to extract text and images from the web pages.

[0908] The extracted content is then evaluated using an AI model (e.g., a model using TensorFlow) to determine whether it contains sexually explicit content. Based on the results of the evaluation, the URLs of web pages that are found to contain sexually explicit content are stored in a storage device (e.g., a MySQL database).

[0909] Additionally, the server summarizes the content of the web page using a natural language processing model (e.g., BERT), which is pre-trained using a large amount of text data.

[0910] In addition, the server uses an emotion engine (e.g., Microsoft Azure Cognitive Services) to recognize the user's emotions. Emotion data is stored in a recording device for each user.

[0911] Terminal side processing

[0912] A browser extension (e.g., Chrome Extension) is installed on the device, and when a user attempts to access a specific web page, it sends the URL to the server. The server receives the URL and compares it with a list of problem pages stored in a recording device. For URLs determined to be problematic, the server generates summary information and a warning message and sends them to the device.

[0913] The terminal displays the received summary information and warning message in a pop-up window, allowing the user to obtain information without being exposed to inappropriate content.

[0914] User processing

[0915] The user checks the displayed summary information and warning message and decides whether to access the web page. The emotion engine also analyzes the user's real-time emotions and sends the results to the server, which then further optimizes the way information is presented in the future.

[0916] Specific examples

[0917] For example, if a user attempts to access the URL "example-adultcontent.com," the device's browser extension sends the URL to the server. The server compares it with information in the database and determines that the URL contains inappropriate advertising. It generates a summary such as "This website provides reviews of image editing software. The gist of the review is..." Furthermore, the emotion engine analyzes the user's emotions, and if the user is feeling stressed or anxious, it generates a warning message saying, "This page contains inappropriate advertising. Viewing it is not recommended."

[0918] An example of the prompt that the user sees:

[0919] "Users submit the URL of the website they are trying to access. Additionally, they enter their current emotional state. For example, the URL 'example-adultcontent.com' and the emotion 'I'm stressed.'"

[0920] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0921] Step 1:

[0922] The server uses a web crawler to crawl multiple web pages on the Internet. The web crawler (e.g., BeautifulSoup, Scrapy) is configured to retrieve the HTML content of all pages in a specified domain. The input to this process is a list of URLs, and the output is the HTML code for each web page.

[0923] Step 2:

[0924] The server analyzes the acquired HTML code and extracts text and images. This analysis uses natural language processing technology (e.g., spaCy) and image recognition technology (e.g., OpenCV). The input is the HTML code, and the output is text data and image data. The extracted text and images are evaluated in the next step.

[0925] Step 3:

[0926] The server inputs the extracted text data and image data into an AI model (e.g., a model built using TensorFlow) to determine whether or not the data contains sexual advertisements. The input is text data and image data, and the output is a determination result of "contains / does not contain sexual advertisements." Based on this result, the determined URLs are stored in a recording device (e.g., a MySQL database) as those containing sexual advertisements.

[0927] Step 4:

[0928] The server uses a natural language processing model (e.g., BERT) to summarize the content of a web page. This model has been trained on a large amount of text data in advance. The input is the text data of the web page, and the output is the summary text. This summary text is saved for presentation to the user.

[0929] Step 5:

[0930] The server uses an emotion engine (e.g., Microsoft Azure Cognitive Services) to analyze the user's emotions. Inputs include the user's facial expressions, voice, and input text, and the output is the user's emotional data (e.g., "feeling stressed," "relaxed," etc.). This emotional data is stored in a recording device for each user.

[0931] Step 6:

[0932] A browser extension (e.g., Chrome Extension) is installed on the device, and when a user tries to access a specific web page, it monitors the URL and sends it to a server. The input is the URL that the user is trying to access, and the output is that the URL is sent to the server.

[0933] Step 7:

[0934] The server checks the received URL against the problem page list stored in the recording device. For URLs that are determined to be problematic, the server generates a warning message based on the previously generated summary text and the user's emotional data. The input is the URL to be accessed and the user's emotional data, and the output is the summary text and the warning message.

[0935] Step 8:

[0936] The terminal displays the summary information and warning message received from the server as a pop-up window. If a user attempts to access "example-adultcontent.com", it displays the message "This site contains inappropriate advertisements. Viewing is not recommended." The input is the summary text and warning message sent from the server, and the output is the display in the pop-up window.

[0937] Step 9:

[0938] The user checks the displayed summary information and warning message and decides whether to access the web page. At the same time, the emotion engine analyzes the user's real-time emotions and sends the results to the server. The input is the user's decision and real-time emotion data, and the output is whether to access the page and updated emotion data.

[0939] (Application example 2)

[0940] 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."

[0941] Currently, when browsing the Internet, there is a risk of accessing inappropriate advertisements or content. Furthermore, information is provided in a uniform manner without considering the user's emotional state, resulting in a suboptimal user experience. Furthermore, users may experience stress or anxiety when exposed to inappropriate content. There is a need to resolve these issues and provide a safer, more user-friendly Internet browsing environment.

[0942] 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.

[0943] In this invention, the server includes means for crawling multiple web pages on the Internet and analyzing the contents of those pages, means for identifying web pages containing inappropriate advertisements based on the analyzed contents and storing the URLs of those pages in a database, means for receiving the URL of a web page that a user attempts to access and determining whether the URL is stored in the database, means for summarizing the contents of the web page containing the inappropriate advertisement, means for analyzing the user's emotional state and generating a warning message or alternative information according to the emotional state, and means for transmitting the summarized contents and information according to the user's emotions to the user's terminal and displaying them as a pop-up window. This makes it possible to provide optimal information according to the user's emotional state while preventing the user from being exposed to inappropriate advertisements or content.

[0944] "Web pages on the Internet" refers to individual pages of publicly accessible websites connected to the Internet.

[0945] "Crawl" refers to the technical process of automatically visiting designated web pages and collecting their content.

[0946] "Content analysis" refers to the process of understanding collected information, such as text and images, from web pages and extracting specific patterns and features.

[0947] "Inappropriate Advertisements" refers to advertisements that may be offensive to users, especially those that contain sexual or violent content.

[0948] "Identifying web pages" refers to the process of finding and identifying web pages that meet certain criteria based on their analyzed content.

[0949] "Means for storing in a database" refers to the process of storing the URLs of the identified web pages and other related information in a dedicated database.

[0950] "Means for receiving the URL of the web page that the user wishes to access" refers to the process of obtaining the address information of the web page that the user has entered or selected.

[0951] "Means for determining" refers to the process of checking a received URL against a list of inappropriate web pages in a database to determine whether it is an inappropriate page.

[0952] "Means for summarizing content" refers to the process of providing a brief summary of the content of a particular Web page.

[0953] "Means for analyzing emotional state" refers to the technical process of recognizing and analyzing the user's current emotions from data such as facial expressions, voice, and input text.

[0954] "Warning Message" refers to a notification that alerts a user to the risk of a particular action or situation.

[0955] "Alternate Information" refers to safe and appropriate information that is provided in place of the page a user is attempting to access.

[0956] "Pop-up window" refers to a small window that appears floating on a user's screen and displays information or a message.

[0957] "User's Device" refers to the electronic device used by the User, such as a smartphone, tablet, or PC.

[0958] "System" refers to a collection of devices and programs that combine all of the above means and have the overall function.

[0959] This invention is a system for safely removing inappropriate advertisements on web pages and providing users with information optimized for them. This system functions in cooperation with a server, a terminal, and a user.

[0960] Server-side processing

[0961] The server uses a crawling engine and a natural language processing engine to crawl multiple web pages on the Internet and analyze their content. It uses web scraping tools such as Python's requests library and BeautifulSoup library for the analysis. It also uses machine learning models to identify web pages that contain inappropriate ads based on the analyzed content. This information is stored in a database for later use.

[0962] The server then uses generative AI models such as Google's Text-To-Text Transfer Transformer (T5) and Bidirectional Encoder Representations from Transformers (BERT) to summarize the content of the webpage, and then uses Microsoft's Azure Emotion API and DeepFace library to apply facial and speech recognition technologies to analyze the user's emotional state.

[0963] Terminal side processing

[0964] When a user attempts to access a web page, the browser extension on the device monitors the URL and sends it to a server. The server checks the URL against a database to determine if it contains inappropriate content. If it does, a server-generated summary is sent to the device and displayed in a pop-up window.

[0965] The device also uses a camera and microphone to analyze the user's emotions in real time, and generates warning messages and alternative information based on the results of this emotion analysis.

[0966] User processing

[0967] Users use their browsers to view web page summaries and warning messages, and use this information to decide whether to access a particular web page. User emotion data is stored over the long term, and this data can be used to provide information optimized for each individual user.

[0968] Specific examples

[0969] For example, consider a situation where a user attempts to access the URL "https: / / example-adultcontent.com." The browser extension sends this URL to a server, which then consults a database. If the URL is determined to be a page containing inappropriate advertising, the AI ​​summarization model summarizes the page's content and generates a summary that reads, "This website contains inappropriate advertising, but its content is a review of image editing software."

[0970] In addition, the emotion engine analyzes the user's facial expressions, and if a negative emotion is detected, a warning message is generated stating, "This page currently contains inappropriate advertisements and is not recommended for viewing." This information and warning message are sent to the device and displayed as a pop-up window, allowing the user to obtain the desired information without being exposed to unpleasant advertisements.

[0971] Use the following as an example prompt:

[0972] test_url = "https: / / example-adultcontent.com"

[0973] test_image_path = "path / to / user_image.jpg"

[0974] main(test_url, test_image_path)

[0975] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0976] Step 1:

[0977] The server crawls multiple web pages on the Internet. During this process, it uses a specific algorithm to generate a list of web page URLs, and then uses Python's requests library and BeautifulSoup library to retrieve the content of each web page. The HTML data of the retrieved web pages is used as input, and the analysis results are output.

[0978] Step 2:

[0979] The server analyzes the content of the retrieved web pages. This analysis uses a natural language processing engine and an image recognition engine to process the text and image data of each web page. A machine learning model is used to determine whether the web page contains inappropriate advertising. The input data is the text and images of the web page, and the output is a determination of whether the web page contains inappropriate advertising.

[0980] Step 3:

[0981] The server identifies web pages containing inappropriate ads based on the analysis results and stores the URLs of these pages in a dedicated database. The stored information includes the URL of the web page and the reason why it was determined to be inappropriate. The output is to store a list of identified URLs in the database.

[0982] Step 4:

[0983] When a user accesses a web page using a browser on their device, a browser extension on the device monitors the URL. It captures the URL the user types in the address bar and sends it to a server. This URL is the input data, and the sent URL is the output.

[0984] Step 5:

[0985] The server checks the received URL against a database to determine whether it contains inappropriate content. Database search technology is used to check the existence of the URL. The input is the URL sent by the user, and the output is the result of whether the URL is inappropriate.

[0986] Step 6:

[0987] If the server determines that the webpage contains inappropriate ads, it uses an AI generative model such as Google's T5 or BERT to summarize the content of the webpage. The generative AI model takes the text data of the webpage as input and outputs a concise summary text.

[0988] Step 7:

[0989] The server receives real-time captured images of the user's facial expressions or voice data to analyze the user's emotional state. Using this as input data, it determines the user's emotional state using an emotion analysis engine (DeepFace or Microsoft Azure Emotion API). The analysis result of the emotional state is obtained as output.

[0990] Step 8:

[0991] The server generates a warning message or alternative information according to the user's emotional state based on the emotion analysis results. The input data is the emotion analysis results, and the output is an appropriate warning message or alternative information.

[0992] Step 9:

[0993] The server sends the summarized content and emotion-based information to the user's terminal and displays them as a pop-up window on the terminal. The input data are the summary information and the emotion-based message, and the output is transmission to the user's terminal and display of the pop-up window.

[0994] 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.

[0995] 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.

[0996] 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.

[0997] [Fourth embodiment]

[0998] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0999] 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.

[1000] 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).

[1001] 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.

[1002] 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.

[1003] 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).

[1004] 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. 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.

[1005] 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.

[1006] 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.

[1007] 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.

[1008] 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.

[1009] 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.

[1010] 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."

[1011] Overall program overview

[1012] This invention relates to a system for securely acquiring the contents of web pages on the Internet. This system involves the cooperation of a server, a terminal, and a user, and aims to restrict access to web pages that contain sexually explicit advertisements.

[1013] Server-side processing

[1014] The server first crawls multiple web pages on the Internet. It analyzes the content of the crawled web pages to identify whether they contain sexually explicit advertisements. The URLs of the identified web pages are stored in a database. When analyzing this content, the server uses natural language processing and image recognition technologies to evaluate the content of the advertisements.

[1015] The server then uses the massive dataset to train a natural language processing model. The trained AI model is then used to summarize the content of the web page. This summarization model is designed to quickly generate a summary of the web page in question. The summarized information is also stored in a database, ready to be served immediately if needed.

[1016] Terminal side processing

[1017] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. The extension sends the URL to a server, which receives the URL and checks it against a list of problem pages stored in a database. If a problem is detected, a summary is sent from the server to the device.

[1018] The device displays a pop-up window to provide summary information to the user. The pop-up window displays, "This page contains inappropriate advertisements. Summary of page contents: (Summary text)." Based on this summary information, the user can obtain the necessary information without being exposed to unpleasant advertisements.

[1019] User processing

[1020] Users can view web page summary information in their browsers and use this information to decide whether to visit a particular web page. This allows users to easily access the information they want while avoiding pages that contain sexually explicit advertisements.

[1021] Specific examples

[1022] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." A browser extension sends this URL to a server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model summarizes the content of this page and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..."

[1023] This summary information is sent to the device and displayed as a pop-up window by the browser extension. The user reads the summary and feels that they have obtained the desired information without being exposed to annoying advertisements. In this way, the system provides users with access to the information they need while preventing them from being exposed to unwanted advertisements.

[1024] This system configuration promotes the soundness of Internet advertising and provides users with a safe and comfortable web browsing environment.

[1025] The processing flow will be explained below.

[1026] Step 1:

[1027] Server: Crawl web pages

[1028] The server automatically collects pages on the Internet, crawls new pages based on a periodically updated URL list, obtains the HTML source of the crawled pages, and stores it for content analysis.

[1029] Step 2:

[1030] Server: Content analysis and identification

[1031] The server analyzes the content of the crawled pages, using natural language processing and image recognition technology to examine the ad banners and metadata on the pages, determining whether they contain sexually explicit content and storing the URLs of problematic pages in a database.

[1032] Step 3:

[1033] Server: Training the natural language processing model

[1034] The server trains a natural language processing model (such as BERT or GPT-3) using a large amount of text data. The training data includes high-quality summaries and feedback. This model is then used to summarize the content of web pages.

[1035] Step 4:

[1036] Terminal: Monitoring web page requests

[1037] A browser extension installed on the user's device monitors the URLs the user attempts to access and sends this URL information to a server.

[1038] Step 5:

[1039] Server: URL rating

[1040] The server checks the received URL against a database to see if the URL exists and whether it is a page containing sexual advertisements.

[1041] Step 6:

[1042] Server: Generate summary information

[1043] If the page is determined to be relevant, the server uses a natural language processing model to summarize the content of the page, generates this summary information, and sends it to the terminal.

[1044] Step 7:

[1045] Terminal: Display summary information

[1046] The browser extension displays the received summary information in a popup window, displaying the message "This page contains inappropriate advertising. Summary of page contents: (summary text)" to the user.

[1047] Step 8:

[1048] User: Review summary information and make a decision

[1049] The user reviews the summary information provided in the pop-up window, allowing them to decide whether to obtain the information they need without being exposed to the intrusive advertisement.

[1050] Step 9:

[1051] User: Safe Web Browsing

[1052] Users can safely obtain the information they are looking for while avoiding inappropriate ads, improving the user experience and promoting the integrity of internet advertising.

[1053] Example 1

[1054] 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."

[1055] On the Internet, there are web pages containing inappropriate content, and accessing them can lead to unpleasant experiences or expose users to harmful information. Web pages containing sexually explicit advertisements are particularly undesirable for users. To address this, it is necessary to provide an environment in which users can browse the web with peace of mind. However, existing filtering systems have the problem of being unable to completely filter out harmful content.

[1056] 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.

[1057] In this invention, the server includes means for crawling multiple web pages on the Internet and analyzing the contents of those pages, means for identifying sexual advertisements based on the analyzed contents and saving the addresses of those pages in data storage means, means for receiving the addresses of web pages that a user wishes to access and determining whether the addresses are saved in the data storage means, means for summarizing the contents of web pages that include sexual advertisements using natural language processing, and means for transmitting the summarized contents to the user's information processing device and displaying them as a pop-up window. This allows users to efficiently access necessary information while reducing the risk of accessing web pages that include sexual advertisements.

[1058] "Crawling" is the process of automatically visiting many web pages on the Internet to collect information.

[1059] "Content analysis" refers to the act of extracting text and image information from collected data and evaluating and classifying it for a specific purpose.

[1060] "Sexual advertising" refers to advertising displays that contain sexual content and are considered offensive or harmful to users.

[1061] An "address" is a Uniform Resource Locator (URL) that identifies a specific web page on the Internet.

[1062] "Data storage means" refers to a database or storage device that stores analyzed data and allows the information to be quickly referenced later.

[1063] "Receiving" is the act of receiving data or information sent from an external source.

[1064] "Natural language processing" refers to the technology and methods that allow computers to understand and generate human language.

[1065] "Summarizing" is the act of concisely summarizing long text or complex information and extracting only the important points.

[1066] An "information processing device" is a device such as a computer or smartphone used by a user.

[1067] A "pop-up window" is a small window that appears within a user interface and is used to display information such as notifications or alerts.

[1068] An "extension" is a software module that adds functionality to a browser or other application.

[1069] MODE FOR CARRYING OUT THE INVENTION

[1070] This invention relates to a system for securely retrieving the contents of web pages on the Internet. In particular, it aims to restrict access to web pages containing sexually explicit advertisements, thereby preventing users from viewing inappropriate content. This system operates in cooperation with a server, a terminal, and a user.

[1071] Server-side processing

[1072] The server uses existing web crawling tools, such as Python's BeautifulSoup and Scrapy, to crawl the web. It collects the HTML content of each crawled web page and stores it in temporary storage. The server then applies natural language processing (NLP) techniques using TensorFlow and PyTorch to analyze the text data within the web page. At the same time, it also uses OpenCV and TensorFlow for image recognition to determine whether the web page contains sexually explicit advertisements.

[1073] The server trains a natural language processing model using a huge dataset (e.g., Wikipedia, Common Crawl). The training process is streamlined by using a high-performance GPU (e.g., Nvidia Tesla). This trained AI model is then used to summarize the content of the web page in question. The analysis results, summary information, and URLs are stored in a database. Databases such as MySQL and PostgreSQL are used.

[1074] Terminal side processing

[1075] When a user attempts to access a specific web page, the device utilizes a browser extension (e.g., Google Chrome Extension). This extension is implemented in JavaScript and captures the user's navigation events to obtain the URL. This URL is then sent to the server as an HTTPS request. The server checks the received URL against a list of problem pages in a database, and if a problem is detected, it sends a summary of the URL to the device.

[1076] The device displays a pop-up window based on the summary information received from the server. The pop-up window displays the message "This page contains inappropriate advertisements. Summary of page content: (Summary text)." This allows users to avoid inappropriate content.

[1077] User processing

[1078] Users can decide whether to access a particular web page based on the summary information displayed in their browser. This information allows them to safely and efficiently obtain the information they need. This also allows users to use the Internet safely, avoiding anxiety and unpleasant experiences.

[1079] Specific examples

[1080] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." The browser extension sends this URL to the server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model then summarizes the page's contents and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..." This summary is then sent to the device and displayed by the browser extension in a pop-up window. The user can read the summary and obtain the desired information without being exposed to any offensive advertisements.

[1081] Prompt Sentence Examples

[1082] "Summarize the content of a given web page URL, determine whether it contains sexual advertisements, and notify the user if necessary."

[1083] This system will promote the integrity of Internet advertising and provide users with a safe and comfortable web browsing environment.

[1084] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1085] Processing step flow

[1086] Step 1: Crawl the web page (server)

[1087] The server crawls multiple web pages on the Internet using a web crawler tool such as Python's BeautifulSoup or Scrapy. The input is the URL of the web page to be crawled, and the output is data including the HTML content. This data is stored in temporary storage.

[1088] Specifically, the server runs a Python script, and the web crawler tool crawls through the specified URL list to collect HTML data, which is then stored in local storage.

[1089] Step 2: Analyzing the web page content (server)

[1090] The server analyzes the collected HTML data and extracts text and images. The analysis uses frameworks such as TensorFlow and PyTorch. The HTML content is input, and the extracted text data and image URLs are output.

[1091] Specifically, the server parses the HTML using BeautifulSoup to extract all text content, and also extracts and lists URLs for image tags. This extracted data is used in the next analysis step.

[1092] Step 3: Identifying Sexual Ads (Server)

[1093] The server analyzes the extracted text data and image URLs to identify sexually explicit advertisements. This is done using natural language processing and image recognition technology. The extracted text data and image URLs are used as input, and the server outputs the identification of web pages containing sexually explicit advertisements and their URLs.

[1094] Specifically, the server uses an NLP model to evaluate text data and detect sexually explicit terms and expressions. At the same time, it uses an image recognition model to evaluate images retrieved from extracted image URLs. Based on the identification results, the webpage URLs are added to a problem page list.

[1095] Step 4: Training the summary model (server)

[1096] The server trains a natural language processing model using a large text dataset, taking the text dataset as input and the trained summarization model as output, which is used to quickly and effectively summarize web page content.

[1097] Specifically, the server loads a text dataset and trains a model using a natural language processing framework (e.g., TensorFlow), with the training process being streamlined using high-performance GPU hardware.

[1098] Step 5: Monitor the URL (Device)

[1099] A browser extension installed on a user's device monitors the URLs of web pages accessed by the user. The monitored URLs are used as input, and a request to send the URL to a server is used as output.

[1100] Specifically, the browser extension uses JavaScript scripts to capture user navigation events and retrieve the URL to be visited, which is then sent to the server as an HTTPS request.

[1101] Step 6: Identify the URL and send the summary information (server)

[1102] The server compares the received URL with a list of problem pages in the database, and if it determines that there is a problem, it generates and sends summary information. The received URL is the input, and the summary information is the output.

[1103] Specifically, the server performs a database query to check whether the URL is included in the problem page list, and if so, summarizes the page content using a trained summarization model and sends the summary information in JSON format to the device.

[1104] Step 7: View summary information (terminal)

[1105] The terminal displays a pop-up window based on the summary information received from the server. The received summary information is used as input, and the display in the pop-up window is used as output.

[1106] Specifically, the browser extension uses HTML and CSS to generate a popup window and display summary information. JavaScript DOM manipulation causes the popup to appear on the user's screen. The popup displays the following message: "This page contains inappropriate advertisements. Summary of page contents: (Summary text)."

[1107] Step 8: User decision (User)

[1108] The user checks the summary information displayed in the pop-up window and decides whether to access the page. The summary information is the input, and the action that determines whether to access the page is the output.

[1109] Specifically, users view the pop-up, examine its contents, and, if necessary, close the browser tab or visit another secure site.

[1110] In this way, a system is configured in which the server, terminal, and user work together in sequence to safely obtain the contents of web pages on the Internet.

[1111] (Application example 1)

[1112] 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."

[1113] The goal is to prevent users from being exposed to unintended sexual advertisements when browsing web pages on the Internet, while at the same time providing a means for them to quickly and safely access the information they need. Furthermore, by using dedicated applications, it is necessary to realize a comfortable browsing experience on a variety of devices, including smartphones and head-mounted displays.

[1114] 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.

[1115] In this invention, the server includes a means for crawling multiple web pages on the Internet and analyzing the content of those pages, a means for identifying web pages containing sexual advertisements based on the analyzed content and storing the URLs of those pages in a database, and a means for receiving the URL of a web page a user attempts to access and determining whether the URL is stored in the database. This allows users to enjoy safe and comfortable web browsing without being exposed to inappropriate advertisements. Furthermore, a dedicated application that can be installed on a smartphone or head-mounted display can monitor accessed URLs and provide summaries of the web page content using a generative AI model.

[1116] The "Internet" is a network that interconnects computers and networks around the world, enabling the exchange of information.

[1117] "Crawling" is a technique for automatically crawling through web pages on the Internet and collecting information.

[1118] "Means for analyzing the content of web pages" refers to technology that extracts data such as text and images from web pages and uses them to understand and evaluate their content.

[1119] "Sexual advertising" refers to advertisements that appear on web pages and contain sexual content.

[1120] A "database" is a system for systematically storing and managing data.

[1121] "Means for receiving the URL of the web page that the user wishes to access" refers to a technique for obtaining the URL of a specific web page when the user accesses that web page.

[1122] A "pop-up window" is a small window that temporarily appears on a user's device to provide important information or warnings.

[1123] A "browser extension" is a software component that adds or extends the functionality of a web browser.

[1124] A "generative AI model" is an artificial intelligence model that has been pre-trained on large datasets to perform tasks such as text generation and summarization.

[1125] A "smartphone" is a mobile phone equipped with internet connectivity and many applications.

[1126] A "head-mounted display" is a device worn on the head that displays visual information.

[1127] To implement this invention, the server, terminal, and user must work together. The details of how each part works are described below.

[1128] Server-side processing

[1129] The server first crawls multiple web pages on the Internet and analyzes their content. Specifically, it retrieves the content of the web pages using the requests library and analyzes the HTML content using BeautifulSoup. Based on the analyzed page content, it determines whether or not the pages contain sexual advertisements. The URLs of pages that contain sexual advertisements are stored in a database.

[1130] The server then trains a summarization model using natural language processing techniques, using the pipeline function in the transformers library to train a generative AI model on a large dataset, which is then used to summarize the content of web pages.

[1131] Terminal side processing

[1132] When a user attempts to access a specific web page in a browser, a browser extension on the device monitors the URL. This monitoring is performed using JavaScript and related browser extension APIs. The URL of the web page the user is attempting to access is sent to a server, which determines whether the URL is stored in its database. If the determination is that the page is problematic, a summary is sent from the server to the device, which then displays the summary to the user in a pop-up window.

[1133] User processing

[1134] Users can check the summary information displayed on their browser and decide whether to access the website based on the content. This allows users to quickly and safely access the information they need without being exposed to intrusive advertisements.

[1135] Specific examples

[1136] As a specific use case, consider the case where a user attempts to access the URL "http: / / example.com / sample-page." At this time, the browser extension sends this URL to the server. The server references the database and identifies this URL as a page containing sexually explicit advertisements. The generative AI model summarizes the page content and generates summary information such as, "This website provides reviews of image editing software. The key points of the reviews are that the user interface is easy to use and that a wide range of editing tools is available." This summary information is sent to the device and displayed as a pop-up window by the browser extension.

[1137] Prompt Sentence Examples

[1138] The following is an example of a prompt that should be given to the generative AI model regarding the URL the user attempted to access and a summary of its content:

[1139] URL: http: / / example.com / sample-page

[1140] Description: This website provides reviews of photo editing software. The main points of the reviews are that the user interface is easy to use and that there are a wide range of editing tools available.

[1141] In this way, the present invention provides users with access to necessary information while preventing them from being exposed to unintended sexual advertisements when browsing web pages on the Internet.By using devices such as smartphones and head-mounted displays, a comfortable browsing experience can be achieved on a variety of devices.

[1142] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1143] Step 1:

[1144] The server crawls multiple web pages on the Internet and retrieves their contents. Specifically, it retrieves the HTML content of each web page using the requests library and parses it using BeautifulSoup. The input is the URL of the web page, and the output is the content of the page, such as text and images.

[1145] Step 2:

[1146] The server determines whether the web page contains sexual advertisements based on the analyzed content. It uses specific keywords and image recognition algorithms to detect sexual content. The input is the web page content extracted in step 1, and the output is the result of determining whether the web page contains sexual advertisements.

[1147] Step 3:

[1148] The server stores the URLs of pages that are determined to contain sexually explicit ads in a database. The database records the URLs of problematic web pages and their verdicts. The input is the verdict and URL obtained in step 2, and the output is the records stored in the database.

[1149] Step 4:

[1150] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. This browser extension captures the URL entered by the user in real time and sends it to a server. The input is the URL the user is trying to access, and the output is the data sent to the server.

[1151] Step 5:

[1152] The server checks the received URL against a list of problem pages stored in a database. The input is the URL sent from the device, and the output is a decision as to whether the URL is a problematic page.

[1153] Step 6:

[1154] The server uses a generative AI model to create summaries for URLs that are determined to be problematic. Specifically, it summarizes the page content using the summarization function of the transformers library. The input is the determined web page content, and the output is a summary.

[1155] Step 7:

[1156] The server sends the summarized content to the terminal. The input is the generated summary sentence, and the output is the data sent to the terminal.

[1157] Step 8:

[1158] The terminal displays the received summary information to the user as a pop-up window. Specifically, it displays a pop-up on the browser using JavaScript. The input is the summary data received from the server, and the output is the pop-up window displayed to the user.

[1159] Step 9:

[1160] The user checks the displayed summary information and makes the final decision on whether to access it. The input is the summary information displayed in the popup window, and the output is the user's decision on whether to allow or deny access.

[1161] 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.

[1162] Overall program overview

[1163] This invention relates to a system that safely retrieves the contents of web pages on the Internet and presents appropriate information according to the user's emotional state. This system operates in cooperation with a server, a terminal, and a user, and provides information that takes into consideration the user's emotions, while restricting access to web pages that contain sexually explicit advertisements.

[1164] Server-side processing

[1165] The server first crawls multiple web pages on the Internet and analyzes their content. It uses natural language processing and image recognition technologies to identify whether they contain sexually explicit advertisements. It stores the URLs of web pages determined to contain sexually explicit advertisements in a database. The server then uses a large amount of data to train a natural language processing model. This model is then used to summarize the content of the web pages.

[1166] Furthermore, the server is equipped with an emotion engine that recognizes the user's emotions. The emotion engine analyzes emotions from information such as the user's facial expressions, voice, and input text. This emotion data is saved for each user and accumulates over time.

[1167] Terminal side processing

[1168] When a user tries to access a specific web page in their browser, a browser extension on the device monitors the URL. The extension sends the URL to a server. The server receives the URL and checks it against a list of problem pages stored in a database. If a problem is detected, a summary is sent from the server to the device.

[1169] The device displays this summary information in a pop-up window. Furthermore, the emotion engine analyzes the user's real-time emotions and adjusts the way information is presented according to the user's emotional state. For example, if the user is in a negative emotional state, an appropriate warning message can be displayed.

[1170] User processing

[1171] Users can view summary information about web pages in their browsers. Based on this information, they can decide whether to access a particular web page. Furthermore, information provided by the emotion engine allows them to learn the best way to respond to their own emotional state. User emotion data is stored over the long term, and an algorithm is run to provide the best display method for each individual user.

[1172] Specific examples

[1173] As a concrete example, consider a situation where a user attempts to access the URL "example-adultcontent.com." A browser extension sends this URL to a server. The server references a database and identifies the URL as a page containing sexually explicit advertisements. The AI ​​summarization model summarizes the content of this page and generates a summary such as, "This website provides reviews of image editing software. The main points of the review are..."

[1174] Furthermore, the emotion engine analyzes the user's emotions, and if the user is expressing negative emotions such as stress or anger, a warning message is generated stating, "This page currently contains inappropriate advertisements and is not recommended for viewing." This summary information and warning message are sent to the device and displayed by the browser extension as a pop-up window. Users feel that they have read this summary and obtained the desired information without being exposed to unpleasant advertisements.

[1175] This system configuration can provide a better user experience by preventing users from being exposed to unnecessary advertisements and by taking into consideration their emotional state. It is also expected to contribute to the soundness of internet advertising.

[1176] The processing flow will be explained below.

[1177] Step 1:

[1178] Server: Crawl web pages

[1179] The server periodically crawls pages on the Internet and retrieves the HTML source of new pages based on the specified URL list. This HTML source is saved for content analysis.

[1180] Step 2:

[1181] Server: Analyze content and identify problem pages

[1182] The server analyzes the content of the crawled pages. It uses natural language processing and image recognition technology to examine the ad banners and metadata on the pages. If any sexually explicit ads are found, the URL of the page is saved in a database as a problem page.

[1183] Step 3:

[1184] Server: Training the natural language processing model

[1185] The server trains a natural language processing model (such as BERT or GPT-3) using large amounts of text data, which is then used to summarize the content of web pages.

[1186] Step 4:

[1187] Server: Preparing the emotion engine

[1188] The server prepares the emotion engine and trains a model to analyze the user's facial expression data, voice data, input text, etc. This uses machine learning techniques to enable high-accuracy recognition of the user's emotional state.

[1189] Step 5:

[1190] Terminal: Monitoring web page requests

[1191] A browser extension installed on the user's device monitors the URLs the user attempts to access and sends this URL information to a server.

[1192] Step 6:

[1193] Server: URL rating

[1194] The server checks the received URL against its database, determines whether the URL is included in the list of problem pages, and returns the result.

[1195] Step 7:

[1196] Server: Generate summary information

[1197] If a page is determined to be problematic, the server uses a natural language processing model to summarize the content of the page, and then generates this summary information and sends it to the terminal.

[1198] Step 8:

[1199] Terminal: Acquisition and analysis of emotion data

[1200] The device's browser extension collects the user's facial expression data, voice data, input text, etc. in real time and sends it to the emotion engine, which analyzes this data and determines the user's emotional state.

[1201] Step 9:

[1202] Server: Sends summary information and generates warning messages

[1203] The server generates an appropriate warning message along with summary information based on the user's emotional state determined by the emotion engine, and transmits this information to the terminal.

[1204] Step 10:

[1205] Terminal: Display summary information and warning messages

[1206] The browser extension displays the received summary information and a warning message in a pop-up window, such as "This page contains inappropriate advertising. Page content summary: (summary text) Viewing is not recommended based on your current emotional state."

[1207] Step 11:

[1208] User: Review summary information and emotional feedback

[1209] Users can review the summary information and warning messages provided in the pop-up window, and with the necessary information, they can safely browse the web without being exposed to annoying ads.

[1210] Step 12:

[1211] User: Safe Web Browsing

[1212] Users can enjoy a better user experience by being protected from unnecessary advertisements and receiving information tailored to their emotional state.

[1213] Example 2

[1214] 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."

[1215] In today's internet usage environment, users often encounter inappropriate sexual advertisements when browsing web pages, significantly degrading the user experience. Furthermore, the lack of appropriate information presented in response to the user's emotional state can potentially increase the user's psychological burden. Conventional technologies have not provided a means to effectively solve these problems simultaneously, making it difficult to provide a comfortable browsing environment for users.

[1216] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes a means for crawling multiple web pages on the Internet and analyzing the contents of those pages, a means for identifying web pages containing sexual advertisements based on the analyzed contents and saving the URLs of those pages in a recording device, and a means for receiving the URL of a web page that a user attempts to access and determining whether the URL is saved in the recording device. This makes it possible to determine in advance whether a web page that a user attempts to view contains inappropriate advertisements and provide the user with appropriate information.

[1217] The server further includes means for summarizing the content of web pages containing sexual advertisements, means for transmitting the summarized content to the user's information processing terminal and displaying it as a pop-up window, means for recognizing the user's emotions and storing the emotion data in a recording device, and means for adjusting the information presentation method according to the user's emotional state, thereby realizing information provision that takes into consideration the user's emotional state and improving the user experience.

[1218] Furthermore, by including a means for training a natural language processing model for summarization and an extension function for the information processing terminal that monitors the URLs of web pages that users are attempting to access, it becomes possible to provide more accurate summary information and monitor in real time, thereby realizing a safe and comfortable internet experience for users without being exposed to inappropriate advertisements.

[1219] A "web page" is a document that displays information that is publicly available on the Internet and is written in a language such as HTML.

[1220] "Crawling" is the process of automatically visiting specific web pages and retrieving their content.

[1221] "Analysis" refers to the processing of acquired data using computer algorithms to extract or evaluate necessary information.

[1222] "Sexual advertising" means commercial information containing sexual content that may have an inappropriate effect on users.

[1223] A "storage device" is a physical or virtual device for storing data, such as a database.

[1224] An "information processing terminal" is an electronic device that can process information, such as a computer, smartphone, or tablet.

[1225] A "pop-up window" is a small window that suddenly appears on a user's screen and is used to present specific information.

[1226] "Emotion" refers to a psychological state that is expressed through a person's facial expression, voice, input text, etc.

[1227] "Emotional data" refers to information about the user's psychological state, including facial expressions and voice analysis results.

[1228] A "natural language processing model" is a machine learning model for understanding and processing human language, and is used for tasks such as summarizing and translating text.

[1229] An "extension" is an additional program that extends the functionality of existing software and is used to enhance the functionality of a browser.

[1230] This invention is a system in which a server, terminals, and users work together to provide a safe and comfortable Internet browsing environment. Specifically, the server crawls multiple web pages on the Internet and analyzes the content of those pages. It also monitors the URLs of web pages that users attempt to access, summarizes the content of problematic pages, and displays a warning to the user in real time.

[1231] Server-side processing

[1232] The server first uses a web crawler (e.g., BeautifulSoup, Scrapy) to crawl multiple web pages on the Internet. It then obtains the HTML content of each page and analyzes it using natural language processing technology (e.g., spaCy) or image recognition technology (e.g., OpenCV). This analysis allows it to extract text and images from the web pages.

[1233] The extracted content is then evaluated using an AI model (e.g., a model using TensorFlow) to determine whether it contains sexually explicit content. Based on the results of the evaluation, the URLs of web pages that are found to contain sexually explicit content are stored in a storage device (e.g., a MySQL database).

[1234] Additionally, the server summarizes the content of the web page using a natural language processing model (e.g., BERT), which is pre-trained using a large amount of text data.

[1235] In addition, the server uses an emotion engine (e.g., Microsoft Azure Cognitive Services) to recognize the user's emotions. Emotion data is stored in a recording device for each user.

[1236] Terminal side processing

[1237] A browser extension (e.g., Chrome Extension) is installed on the device, and when a user attempts to access a specific web page, it sends the URL to the server. The server receives the URL and compares it with a list of problem pages stored in a recording device. For URLs determined to be problematic, the server generates summary information and a warning message and sends them to the device.

[1238] The terminal displays the received summary information and warning message in a pop-up window, allowing the user to obtain information without being exposed to inappropriate content.

[1239] User processing

[1240] The user checks the displayed summary information and warning message and decides whether to access the web page. The emotion engine also analyzes the user's real-time emotions and sends the results to the server, which then further optimizes the way information is presented in the future.

[1241] Specific examples

[1242] For example, if a user attempts to access the URL "example-adultcontent.com," the device's browser extension sends the URL to the server. The server compares it with information in the database and determines that the URL contains inappropriate advertising. It generates a summary such as "This website provides reviews of image editing software. The gist of the review is..." Furthermore, the emotion engine analyzes the user's emotions, and if the user is feeling stressed or anxious, it generates a warning message saying, "This page contains inappropriate advertising. Viewing it is not recommended."

[1243] An example of the prompt that the user sees:

[1244] "Users submit the URL of the website they are trying to access. Additionally, they enter their current emotional state. For example, the URL 'example-adultcontent.com' and the emotion 'I'm stressed.'"

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

[1246] Step 1:

[1247] The server uses a web crawler to crawl multiple web pages on the Internet. The web crawler (e.g., BeautifulSoup, Scrapy) is configured to retrieve the HTML content of all pages in a specified domain. The input to this process is a list of URLs, and the output is the HTML code for each web page.

[1248] Step 2:

[1249] The server analyzes the acquired HTML code and extracts text and images. This analysis uses natural language processing technology (e.g., spaCy) and image recognition technology (e.g., OpenCV). The input is the HTML code, and the output is text data and image data. The extracted text and images are evaluated in the next step.

[1250] Step 3:

[1251] The server inputs the extracted text data and image data into an AI model (e.g., a model built using TensorFlow) to determine whether or not the data contains sexual advertisements. The input is text data and image data, and the output is a determination result of "contains / does not contain sexual advertisements." Based on this result, the determined URLs are stored in a recording device (e.g., a MySQL database) as those containing sexual advertisements.

[1252] Step 4:

[1253] The server uses a natural language processing model (e.g., BERT) to summarize the content of a web page. This model has been trained on a large amount of text data in advance. The input is the text data of the web page, and the output is the summary text. This summary text is saved for presentation to the user.

[1254] Step 5:

[1255] The server uses an emotion engine (e.g., Microsoft Azure Cognitive Services) to analyze the user's emotions. Inputs include the user's facial expressions, voice, and input text, and the output is the user's emotional data (e.g., "feeling stressed," "relaxed," etc.). This emotional data is stored in a recording device for each user.

[1256] Step 6:

[1257] A browser extension (e.g., Chrome Extension) is installed on the device, and when a user tries to access a specific web page, it monitors the URL and sends it to a server. The input is the URL that the user is trying to access, and the output is that the URL is sent to the server.

[1258] Step 7:

[1259] The server checks the received URL against the problem page list stored in the recording device. For URLs that are determined to be problematic, the server generates a warning message based on the previously generated summary text and the user's emotional data. The input is the URL to be accessed and the user's emotional data, and the output is the summary text and the warning message.

[1260] Step 8:

[1261] The terminal displays the summary information and warning message received from the server as a pop-up window. If a user attempts to access "example-adultcontent.com", it displays the message "This site contains inappropriate advertisements. Viewing is not recommended." The input is the summary text and warning message sent from the server, and the output is the display in the pop-up window.

[1262] Step 9:

[1263] The user checks the displayed summary information and warning message and decides whether to access the web page. At the same time, the emotion engine analyzes the user's real-time emotions and sends the results to the server. The input is the user's decision and real-time emotion data, and the output is whether to access the page and updated emotion data.

[1264] (Application example 2)

[1265] 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."

[1266] Currently, when browsing the Internet, there is a risk of accessing inappropriate advertisements or content. Furthermore, information is provided in a uniform manner without considering the user's emotional state, resulting in a suboptimal user experience. Furthermore, users may experience stress or anxiety when exposed to inappropriate content. There is a need to resolve these issues and provide a safer, more user-friendly Internet browsing environment.

[1267] 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.

[1268] In this invention, the server includes means for crawling multiple web pages on the Internet and analyzing the contents of those pages, means for identifying web pages containing inappropriate advertisements based on the analyzed contents and storing the URLs of those pages in a database, means for receiving the URL of a web page that a user attempts to access and determining whether the URL is stored in the database, means for summarizing the contents of the web page containing the inappropriate advertisement, means for analyzing the user's emotional state and generating a warning message or alternative information according to the emotional state, and means for transmitting the summarized contents and information according to the user's emotions to the user's terminal and displaying them as a pop-up window. This makes it possible to provide optimal information according to the user's emotional state while preventing the user from being exposed to inappropriate advertisements or content.

[1269] "Web pages on the Internet" refers to individual pages of publicly accessible websites connected to the Internet.

[1270] "Crawl" refers to the technical process of automatically visiting designated web pages and collecting their content.

[1271] "Content analysis" refers to the process of understanding collected information, such as text and images, from web pages and extracting specific patterns and features.

[1272] "Inappropriate Advertisements" refers to advertisements that may be offensive to users, especially those that contain sexual or violent content.

[1273] "Identifying web pages" refers to the process of finding and identifying web pages that meet certain criteria based on their analyzed content.

[1274] "Means for storing in a database" refers to the process of storing the URLs of the identified web pages and other related information in a dedicated database.

[1275] "Means for receiving the URL of the web page that the user wishes to access" refers to the process of obtaining the address information of the web page that the user has entered or selected.

[1276] "Means for determining" refers to the process of checking a received URL against a list of inappropriate web pages in a database to determine whether it is an inappropriate page.

[1277] "Means for summarizing content" refers to the process of providing a brief summary of the content of a particular Web page.

[1278] "Means for analyzing emotional state" refers to the technical process of recognizing and analyzing the user's current emotions from data such as facial expressions, voice, and input text.

[1279] "Warning Message" refers to a notification that alerts a user to the risk of a particular action or situation.

[1280] "Alternate Information" refers to safe and appropriate information that is provided in place of the page a user is attempting to access.

[1281] "Pop-up window" refers to a small window that appears floating on a user's screen and displays information or a message.

[1282] "User's Device" refers to the electronic device used by the User, such as a smartphone, tablet, or PC.

[1283] "System" refers to a collection of devices and programs that combine all of the above means and have the overall function.

[1284] This invention is a system for safely removing inappropriate advertisements on web pages and providing users with information optimized for them. This system functions in cooperation with a server, a terminal, and a user.

[1285] Server-side processing

[1286] The server uses a crawling engine and a natural language processing engine to crawl multiple web pages on the Internet and analyze their content. It uses web scraping tools such as Python's requests library and BeautifulSoup library for the analysis. It also uses machine learning models to identify web pages that contain inappropriate ads based on the analyzed content. This information is stored in a database for later use.

[1287] The server then uses generative AI models such as Google's Text-To-Text Transfer Transformer (T5) and Bidirectional Encoder Representations from Transformers (BERT) to summarize the content of the webpage, and then uses Microsoft's Azure Emotion API and DeepFace library to apply facial and speech recognition technologies to analyze the user's emotional state.

[1288] Terminal side processing

[1289] When a user attempts to access a web page, the browser extension on the device monitors the URL and sends it to a server. The server checks the URL against a database to determine if it contains inappropriate content. If it does, a server-generated summary is sent to the device and displayed in a pop-up window.

[1290] The device also uses a camera and microphone to analyze the user's emotions in real time, and generates warning messages and alternative information based on the results of this emotion analysis.

[1291] User processing

[1292] Users use their browsers to view web page summaries and warning messages, and use this information to decide whether to access a particular web page. User emotion data is stored over the long term, and this data can be used to provide information optimized for each individual user.

[1293] Specific examples

[1294] For example, consider a situation where a user attempts to access the URL "https: / / example-adultcontent.com." The browser extension sends this URL to a server, which then consults a database. If the URL is determined to be a page containing inappropriate advertising, the AI ​​summarization model summarizes the page's content and generates a summary that reads, "This website contains inappropriate advertising, but its content is a review of image editing software."

[1295] In addition, the emotion engine analyzes the user's facial expressions, and if a negative emotion is detected, a warning message is generated stating, "This page currently contains inappropriate advertisements and is not recommended for viewing." This information and warning message are sent to the device and displayed as a pop-up window, allowing the user to obtain the desired information without being exposed to unpleasant advertisements.

[1296] Use the following as an example prompt:

[1297] test_url = "https: / / example-adultcontent.com"

[1298] test_image_path = "path / to / user_image.jpg"

[1299] main(test_url, test_image_path)

[1300] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1301] Step 1:

[1302] The server crawls multiple web pages on the Internet. During this process, it uses a specific algorithm to generate a list of web page URLs, and then uses Python's requests library and BeautifulSoup library to retrieve the content of each web page. The HTML data of the retrieved web pages is used as input, and the analysis results are output.

[1303] Step 2:

[1304] The server analyzes the content of the retrieved web pages. This analysis uses a natural language processing engine and an image recognition engine to process the text and image data of each web page. A machine learning model is used to determine whether the web page contains inappropriate advertising. The input data is the text and images of the web page, and the output is a determination of whether the web page contains inappropriate advertising.

[1305] Step 3:

[1306] The server identifies web pages containing inappropriate ads based on the analysis results and stores the URLs of these pages in a dedicated database. The stored information includes the URL of the web page and the reason why it was determined to be inappropriate. The output is to store a list of identified URLs in the database.

[1307] Step 4:

[1308] When a user accesses a web page using a browser on their device, a browser extension on the device monitors the URL. It captures the URL the user types in the address bar and sends it to a server. This URL is the input data, and the sent URL is the output.

[1309] Step 5:

[1310] The server checks the received URL against a database to determine whether it contains inappropriate content. Database search technology is used to check the existence of the URL. The input is the URL sent by the user, and the output is the result of whether the URL is inappropriate.

[1311] Step 6:

[1312] If the server determines that the webpage contains inappropriate ads, it uses an AI generative model such as Google's T5 or BERT to summarize the content of the webpage. The generative AI model takes the text data of the webpage as input and outputs a concise summary text.

[1313] Step 7:

[1314] The server receives real-time captured images of the user's facial expressions or voice data to analyze the user's emotional state. Using this as input data, it determines the user's emotional state using an emotion analysis engine (DeepFace or Microsoft Azure Emotion API). The analysis result of the emotional state is obtained as output.

[1315] Step 8:

[1316] The server generates a warning message or alternative information according to the user's emotional state based on the emotion analysis results. The input data is the emotion analysis results, and the output is an appropriate warning message or alternative information.

[1317] Step 9:

[1318] The server sends the summarized content and emotion-based information to the user's terminal and displays them as a pop-up window on the terminal. The input data are the summary information and the emotion-based message, and the output is transmission to the user's terminal and display of the pop-up window.

[1319] 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.

[1320] 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.

[1321] 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.

[1322] 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.

[1323] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1324] 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.

[1325] 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).

[1326] 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.

[1327] 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."

[1328] 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.

[1329] 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).

[1330] 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.

[1331] 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.

[1332] 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.

[1333] 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.

[1334] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1335] 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.

[1336] 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.

[1337] 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.

[1338] 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.

[1339] 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.

[1340] The following is further disclosed regarding the above embodiment.

[1341] (Claim 1)

[1342] means for crawling a plurality of web pages on the Internet and analyzing the content of those pages;

[1343] means for identifying web pages containing sexually explicit advertisements based on the analyzed content and storing the URLs of these pages in a database;

[1344] means for receiving a URL of a web page that a user wishes to access and determining whether the URL is stored in a database;

[1345] means for summarizing the content of a web page containing sexually explicit advertisements;

[1346] means for transmitting the summarized content to a user's terminal and displaying it as a pop-up window;

[1347] A system including:

[1348] (Claim 2)

[1349] 10. The system of claim 1, further comprising means for training a natural language processing model for summarization.

[1350] (Claim 3)

[1351] 10. The system of claim 1, further comprising a browser extension that monitors the URLs of web pages that a user attempts to access.

[1352] "Example 1"

[1353] (Claim 1)

[1354] means for crawling a plurality of web pages on the Internet and analyzing the content of those pages;

[1355] means for identifying sexual advertisements based on the analyzed content and storing addresses of these pages in a data storage means;

[1356] means for receiving an address of a web page that a user wishes to access and determining whether the address is stored in a data storage means;

[1357] A means for summarizing the content of a web page containing sexual advertisements using natural language processing;

[1358] means for transmitting the summarized content to the user's information processing device and displaying it as a pop-up window;

[1359] A system including:

[1360] (Claim 2)

[1361] 10. The system of claim 1, further comprising means for training a natural language processing model for summarization.

[1362] (Claim 3)

[1363] 10. The system of claim 1, further comprising an extension that monitors addresses of web pages that users attempt to access.

[1364] "Application Example 1"

[1365] (Claim 1)

[1366] means for crawling a plurality of web pages on the Internet and analyzing the content of those pages;

[1367] means for identifying web pages containing sexually explicit advertisements based on the analyzed content and storing the URLs of these pages in a database;

[1368] means for receiving a URL of a web page that a user wishes to access and determining whether the URL is stored in a database;

[1369] means for summarizing the content of a web page containing sexually explicit advertisements;

[1370] means for transmitting the summarized content to a user's terminal and displaying it as a pop-up window;

[1371] A means to monitor the URLs accessed using a browser extension,

[1372] A means for generating summary content using a generative AI model;

[1373] A system including:

[1374] (Claim 2)

[1375] 10. The system of claim 1, further comprising means for training a natural language processing model for summarization.

[1376] (Claim 3)

[1377] 10. The system of claim 1, including an application that can be installed on a smartphone or a head-mounted display.

[1378] "Example 2: Combining Emotion Engines"

[1379] (Claim 1)

[1380] means for crawling a plurality of web pages on the Internet and analyzing the content of those pages;

[1381] means for identifying web pages containing sexual advertisements based on the analyzed content and storing the URLs of these pages in a storage device;

[1382] a means for receiving a URL of a web page that a user wishes to access and determining whether the URL is stored in a storage device;

[1383] means for summarizing the content of a web page containing sexually explicit advertisements;

[1384] means for transmitting the summarized content to the user's information processing terminal and displaying it as a pop-up window;

[1385] means for recognizing a user's emotion and storing emotion data in a recording device;

[1386] means for adjusting the presentation of information according to the emotional state of the user;

[1387] A system including:

[1388] (Claim 2)

[1389] 10. The system of claim 1, further comprising means for training a natural language processing model for summarization.

[1390] (Claim 3)

[1391] 10. The system of claim 1, further comprising an extension to the information processing terminal that monitors the URLs of web pages that users attempt to access.

[1392] "Application example 2 when combining emotion engines"

[1393] (Claim 1)

[1394] means for crawling a plurality of web pages on the Internet and analyzing the content of those pages;

[1395] means for identifying web pages containing inappropriate advertisements based on the analyzed content and storing the URLs of these pages in a database;

[1396] means for receiving a URL of a web page that a user wishes to access and determining whether the URL is stored in a database;

[1397] means for summarizing the content of a web page containing objectionable advertising;

[1398] means for analyzing the emotional state of a user and generating a warning message or alternative information according to the emotional state;

[1399] means for transmitting the summarized content and information according to the emotion to a terminal of the user and displaying the information as a pop-up window;

[1400] A system including:

[1401] (Claim 2)

[1402] 10. The system of claim 1, further comprising means for training a natural language processing model for summarization.

[1403] (Claim 3)

[1404] 10. The system of claim 1, further comprising a browser extension that monitors the URLs of web pages that a user attempts to access. [Explanation of symbols]

[1405] 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. means for crawling a plurality of web pages on the Internet and analyzing the content of those pages; means for identifying web pages containing sexually explicit advertisements based on the analyzed content and storing the URLs of these pages in a database; means for receiving a URL of a web page that a user wishes to access and determining whether the URL is stored in a database; means for summarizing the content of a web page containing sexually explicit advertisements; means for transmitting the summarized content to a user's terminal and displaying it as a pop-up window; A system including:

2. The system of claim 1 , further comprising means for training a natural language processing model for summarization.

3. 10. The system of claim 1, further comprising a browser extension that monitors the URLs of web pages that users attempt to access.

Images