System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting news articles; means for summarizing the collected news articles according to age; and means for providing the summarized news articles to a terminal.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Today's news often contains highly specialized content, and there is a lack of information summarized in an easy-to-understand format, especially for children. This limits opportunities for children to learn about current events and social conditions. It also requires a great deal of effort for parents and teachers to provide news in a format that is easy for children to understand. Conventional news distribution systems have not been able to adequately resolve these issues. Furthermore, they lack read-record management and feedback functions, making it difficult to fully grasp children's interests and learning progress. [Means for solving the problem]

[0005] The present invention provides a system including a means for collecting news articles, a means for summarizing the collected news articles according to age, and a means for providing the summarized news articles to a terminal. The system further includes a means for displaying news articles selected by a user on the terminal, a means for marking the displayed news articles as read, and a means for collecting user feedback. The system also includes a means for categorizing news articles by related topic, a means for displaying related news articles together, and a means for storing the news articles in a knowledge base. The system also includes a means for sharing the news articles on a social networking site, a means for optimizing the shared news articles for a target age group, a means for analyzing feedback data on the news articles to improve a model, and a means for using the improved model to generate subsequent summaries. In this way, it is possible to provide news articles summarized in an easy-to-understand manner for children and efficiently manage learning progress and feedback.

[0006] A "news article" is a piece of writing about the latest current events or social issues, which can be obtained from newspapers, websites, news apps, etc.

[0007] An "aggregator" is a program or process for obtaining news articles from external news sources.

[0008] An "age-appropriate summarization tool" is a program or process that uses a generative AI model to summarize news articles in a simple, easy-to-understand format appropriate for a specific age group (e.g., elementary school students, middle school students).

[0009] A "means for providing" is a program or process for transmitting and displaying summarized news articles to a terminal.

[0010] A "displaying means" is a program or process that visually displays the summarized news article on the screen of a terminal.

[0011] A "mark as read means" is a program or process that records a news article as read on the system after a user has finished viewing the article.

[0012] A "feedback collection means" is a program or process that receives user opinions and requests as input, transfers them to a server, and stores them in a database.

[0013] A "topical categorization means" is a program or process that automatically categorizes news articles according to particular themes or topics.

[0014] A "collaborative display means" is a program or process that displays related news articles together in a list format.

[0015] A "means for storing in a knowledge base" is a program or process that stores classified news articles and their associated information in a database and manages them for future reference.

[0016] "Means for sharing on SNS" refers to a program or process that generates and sends links and summaries that allow users to share news articles on SNS (social networking services).

[0017] "Age-optimization" is a program or process that converts news articles shared on social media into simple, easy-to-understand language tailored to the age group of potential viewers.

[0018] A "means for analyzing feedback data" is a program or process that analyzes user opinions and requests as data and uses this data to improve the system or summary generation model.

[0019] "Means for improving the model" refers to a program or process that uses feedback data and new training data to tune the generative AI model and reflect it in the next news summary generation. [Brief explanation of the drawings]

[0020] [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

[0021] 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.

[0022] First, the terms used in the following description will be explained.

[0023] 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).

[0024] 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.

[0025] 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.

[0026] 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.

[0027] 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."

[0028] [First embodiment]

[0029] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0030] 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.

[0031] 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).

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0037] 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.

[0038] 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.

[0039] 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.

[0040] 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."

[0041] The present invention relates to a system for collecting news articles, summarizing them in an easy-to-understand manner for children, and providing them to a terminal. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes embodiments of the present invention with specific examples.

[0042] News article collection

[0043] server:

[0044] The server periodically sends requests to a news source (e.g., a news API) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[0045] News article preprocessing and classification

[0046] server:

[0047] The collected news articles are first preprocessed using a text analysis library, which includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The preprocessed news articles are then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[0048] Summary Generation

[0049] server:

[0050] The classified news articles are then fed into a generative AI model, which then summarizes the article in an easy-to-understand way based on the user's (parent's or child's) age preference. For example, for an 8-year-old child, the article might be simplified to something like, "The earth's temperature is rising. We're all working together to stop this."

[0051] Providing news summaries

[0052] Device:

[0053] When a user opens the application, the device sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the device's user interface.

[0054] News read status management

[0055] User:

[0056] Users (usually children) can tap on a displayed news article to view details, then press a button to mark it as "read."

[0057] Device:

[0058] The "read" mark data is sent from the device to the server and stored in a database on the server side, allowing the user's browsing history to be managed.

[0059] View related news

[0060] server:

[0061] To display related news articles in one place, the server categorizes the summarized news articles by topic and stores them in a database, so that when a user reads a particular article, other related articles are also displayed.

[0062] Feedback collection

[0063] User:

[0064] Users can fill out a feedback form for a news article and press the "Submit" button. For example, they can enter a request such as "I'd like this article to be more detailed."

[0065] Device:

[0066] The feedback data is sent from the device to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[0067] SNS sharing function

[0068] User:

[0069] Users can click the "Share" button on the details screen of a news article to share it on social media.

[0070] Device:

[0071] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[0072] server:

[0073] Shared news articles will be displayed to other users in an age-optimized format.

[0074] Model Improvement

[0075] server:

[0076] The collected feedback data is analyzed and the generative AI model is tuned based on the results. The improved model is used to generate the next news summary.

[0077] Specific examples

[0078] As a concrete example, consider a 10-year-old child using an app. When the app is opened, a summary article titled "New Report on Climate Change" ("The temperature of the earth has been rising recently. We are all working hard to prevent this.") is retrieved from the server. The child then reads the article and presses the "Read" button. A related news item, "The Importance of Recycling," is then displayed, which the child can read.

[0079] In this way, the present invention summarizes news articles in an easy-to-understand manner and provides them to children, thereby increasing children's opportunities to come into contact with current events and social situations and promoting communication with parents and teachers.

[0080] The processing flow will be explained below.

[0081] Step 1:

[0082] Server: Sends requests to news sources (e.g., news APIs) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[0083] Step 2:

[0084] Server: The retrieved news articles are preprocessed using a text analysis library, which includes removing noise data (advertisements, links, special characters, etc.) and normalizing character encoding.

[0085] Step 3:

[0086] Server: Preprocessed news articles are input into a topic classification model. The model classifies news articles into specific topics (e.g., environment, politics, science and technology). The classification results are stored in a temporary data store.

[0087] Step 4:

[0088] Server: The classified news articles are fed into a generative AI model to be summarized for children. Based on the age setting (e.g., elementary school, middle school), a summary text is generated in an easy-to-understand format.

[0089] Step 5:

[0090] Server: Stores summarized news articles and their metadata (topic, target demographic, etc.) in a database. The stored data is linked to other related news articles.

[0091] Step 6:

[0092] Device: When a user (parent or child) opens the application, the device sends a request to the server to retrieve the latest summary news articles.

[0093] Step 7:

[0094] Terminal: The retrieved news summaries are displayed in a user interface in the form of a list, with each news item showing its title and summary text.

[0095] Step 8:

[0096] User: Taps on a displayed news article to view details, then presses a button to mark it as "read."

[0097] Step 9:

[0098] Terminal: Detects user actions and sends the "read" mark data to the server.

[0099] Step 10:

[0100] Server: The received "read" data is saved in a database and managed as the user's browsing history.

[0101] Step 11:

[0102] Server: Based on the related information of the news article, selects news articles on related topics and prepares them for provision to users.

[0103] Step 12:

[0104] On-device: When a user reads a particular article, other related articles are displayed along with it. Related news is displayed in a list format for easy access.

[0105] Step 13:

[0106] User: Can fill out a form to provide feedback on a news article and press the "Submit" button.

[0107] Step 14:

[0108] Terminal: Sends feedback data to the server.

[0109] Step 15:

[0110] Server: The received feedback data is stored in a database and used as reference data for model improvement.

[0111] Step 16:

[0112] User: Press the "Share" button on the news article details screen to share the news on social media.

[0113] Step 17:

[0114] Device: The SNS selection screen is displayed, and the article summary and sharing link are sent to the SNS selected by the user.

[0115] Step 18:

[0116] Server: News articles shared on social media are displayed to other users in a format optimized for the target age group.

[0117] Step 19:

[0118] Server: Analyzes the collected feedback data and tunes the generative AI model. The improved model is used to generate the next news summary.

[0119] Example 1

[0120] 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."

[0121] Conventional news article delivery systems do not provide summaries based on specific age settings, making it difficult for users to understand news articles in a way that is appropriate for them, especially children. Furthermore, they lack the ability to manage read statuses and collect feedback when users view articles, making it difficult to provide content that is tailored to the user's usage. Furthermore, they lack the ability to efficiently display related news articles, and do not provide sufficient support for deepening understanding.

[0122] 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.

[0123] In this invention, the server includes means for collecting news articles, means for preprocessing the collected news articles using a text analysis library, means for inputting the preprocessed news articles into a topic classification model and classifying them into specific categories, means for inputting the classified news articles into a generative AI model and summarizing them based on user settings, means for providing the summarized news articles to a terminal, and means for displaying them on a user interface. This enables the provision of easy-to-understand summarized news articles tailored to the user's age. Additionally, the server can mark news articles selected by the user as read, send and store the read data on the server, and collect feedback, enabling the provision of content based on more specific user usage patterns. Furthermore, the automatic classification and display of related news articles enables deeper understanding and information acquisition.

[0124] "News article gathering means" means hardware or software for periodically retrieving current news articles from online news sources.

[0125] A "text analysis library" is a software tool used to preprocess collected news articles, such as removing noise data and normalizing character encoding.

[0126] "Topic classification model" refers to a machine learning model for classifying preprocessed news articles into specific categories.

[0127] A "generative AI model" refers to an artificial intelligence model that takes categorized news articles as input and provides an easy-to-understand summary based on the user's age settings.

[0128] "Summarization method" refers to the process and software that uses a generative AI model to summarize news articles in a format that is easy for users to understand.

[0129] "Means for providing to the terminal" refers to the processes and software for transmitting summarized news articles from the server to the terminal and for the terminal to receive and display them.

[0130] "User Interface" refers to the on-screen interface through which a user can view news articles, provide feedback, and access other features.

[0131] "Means to mark as read" refers to the functionality and processes that allow a user to mark a news article as "read" and save that information.

[0132] "Feedback collection means" refers to the process and software for collecting user-entered feedback on news articles and transmitting and storing that data on a server.

[0133] "Means for categorizing by relevant topic" refers to the process and software for re-categorizing summarized news articles by topic and storing them in a database.

[0134] "Means for displaying related news articles together" refers to processes and software for displaying other articles related to a particular article at the same time as the user reads that article.

[0135] "Means for storing in knowledge base" refers to the functions and processes by which processed news articles and related information are stored in a database for later retrieval.

[0136] The present invention relates to a system for collecting news articles, summarizing them in an easy-to-understand manner for children, and providing them to a terminal. Specific embodiments for carrying out the present invention will be described below.

[0137] News article collection

[0138] server:

[0139] The server uses the requests library to periodically send HTTP requests to news sources (e.g., news APIs) to retrieve new news articles, which are then stored in a temporary data store (e.g., an SQL database) on the server, for example in the form of a pandas dataframe.

[0140] News article preprocessing and classification

[0141] server:

[0142] The collected news articles are first preprocessed using text analysis libraries such as NLTK and spaCy, which include removing noise data (such as advertisements and links) using regular expressions and standardizing character encoding.

[0143] The preprocessed news articles are then fed into a topic classification model (e.g., BERT, GPT-3) to classify them into specific categories (e.g., environment, politics, science and technology).

[0144] Summary Generation

[0145] server:

[0146] The classified news articles are fed into a generative AI model (e.g., OpenAI's GPT-3), which then summarizes the article in an easy-to-understand way based on the user's (parent or child's) age preference. Examples of prompts include:

[0147] "Summarize the following news article in a way that would be understandable for a 10-year-old.

[0148] News Article:

[0149] As the Earth's temperature rises, various measures are being taken around the world...

[0150] "

[0151] The generative AI model uses this prompt to summarize the news article as something like, "The temperature of the earth has been rising recently. We are all working hard to prevent this."

[0152] Providing news summaries

[0153] Device:

[0154] When a user opens the application, the device sends an HTTP request to the server to retrieve the latest news summary articles, which are then displayed in a list format in the device's user interface using RecyclerView and ListView.

[0155] News read status management

[0156] User:

[0157] Users (mainly children) can tap on the displayed news article to view details, and then press the "read" button after viewing.

[0158] Device:

[0159] The "read" mark data is sent from the device to the server by sending an HTTP POST request, and the server stores this data in a database and manages it as the user's browsing history.

[0160] View related news

[0161] server:

[0162] The server recategorizes the summarized news articles by topic and stores them in a database, allowing users to simultaneously view other related articles as they read a particular article.

[0163] Feedback collection

[0164] User:

[0165] Users can enter their feedback on the news article in the form and press the "Submit" button. They can enter requests such as "I wish this article had more details."

[0166] Device:

[0167] The device sends the feedback data to the server, which stores it in a database and uses it to improve the generative AI model.

[0168] SNS sharing function

[0169] User:

[0170] Users can click the "Share" button on the details screen of a news article to share it on social media.

[0171] Device:

[0172] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[0173] Model Improvement

[0174] server:

[0175] The server analyzes the collected feedback data and uses the results to tune the generative AI model, which is then used to generate the next news summary.

[0176] Specific examples

[0177] For example, if a 10-year-old child were to use this system, the process would go something like this:

[0178] 1. When a child opens the app, the device retrieves the latest summary news article (e.g., "The temperature of the earth has been rising recently. We are all working hard to prevent this.") from the server.

[0179] 2. The child reads the article and then presses the "read" button, which saves the read information on the server.

[0180] 3. Additionally, a related news item, "The Importance of Recycling," is displayed, which children can also read.

[0181] Thus, the present invention is a system that provides children with easy-to-understand summaries of news articles, helps them understand current affairs and social situations, and promotes communication with parents and teachers.

[0182] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0183] Step 1:

[0184] News article collection

[0185] The server sends HTTP GET requests to a news source (e.g., a news API). The input is the API endpoint and authentication information, and the server executes this periodically. The output is the JSON data of the retrieved news articles. The server converts this data into a pandas dataframe and stores it in a temporary data store (e.g., a SQL database).

[0186] Step 2:

[0187] News article preprocessing

[0188] The server retrieves news articles from the temporary data store and preprocesses them using a text analysis library (e.g., NLTK, spaCy). The input is the news article data saved in the previous step, and the output is the preprocessed clean text data. Specific operations include removing noise data (advertisements, links, etc.) using regular expressions and standardizing character encoding.

[0189] Step 3:

[0190] Topic Classification

[0191] The server inputs preprocessed news articles into a topic classification model (e.g., BERT, GPT-3). The input is clean text data, and the output is data classified into specific categories (e.g., environment, politics, science and technology). Specifically, the text data is converted into feature vectors, and classification results are obtained using a pre-trained model.

[0192] Step 4:

[0193] Summary Generation

[0194] The server inputs the classified news article into a generative AI model (e.g., OpenAI's GPT-3). The input is the classified news article and a prompt, and the output is a summarized news article based on the user's age preference. Specifically, the prompt is generated in the following format:

[0195] "Summarize the following news article in a way that is understandable for a 10-year-old.

[0196] News Article:

[0197] As the Earth's temperature rises, various measures are being taken around the world...

[0198] "

[0199] The generative AI model uses this information to summarize the news, and the server stores the summarized article in a database.

[0200] Step 5:

[0201] Providing news summaries

[0202] When a user opens the application on a device, the device sends an HTTP GET request to the server to request a summary news article. The input is the user's request, and the output is the latest summary news article data. The server sends this in JSON format to the device, which then displays it using RecyclerView or ListView.

[0203] Step 6:

[0204] News read status management

[0205] Users can tap on a news article to view details and then press the "mark as read" button after viewing. The input is the user's action, and the output is the updated read data. The device sends this data as an HTTP POST request to the server, which stores it in a database.

[0206] Step 7:

[0207] View related news

[0208] The server reclassifies summarized news articles by topic and stores them in a database. The input is the summarized news article data, and the output is the data classified by topic. When a user reads a particular article, the device sends a request to the server to display related news as well. The server sends the corresponding related news in JSON format to the device, and the device displays it.

[0209] Step 8:

[0210] Feedback collection

[0211] The user enters their feedback on the news article into the form and presses the "Submit" button. The input is the user's feedback content, and the output is the feedback data. The terminal sends the feedback data to the server via an HTTP POST request, and the server stores it in a database.

[0212] Step 9:

[0213] SNS sharing function

[0214] The user presses the "Share" button on the news article details screen, which takes them to the SNS sharing selection screen. The input is the user's operation, and the output is a sharing link and article summary. The device uses the SNS API to send the article summary and sharing link.

[0215] Step 10:

[0216] Model Improvement

[0217] The server analyzes the collected feedback data and tunes the generative AI model. The input is the feedback data and the output is the improved model. The server applies the new model to be used in generating the next news summary.

[0218] (Application example 1)

[0219] 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."

[0220] In today's world, it is important for children to have opportunities to read news articles, but the content of news is often too technical and difficult to understand. Therefore, there is a need for an easy-to-understand news summary system that helps children understand and be interested in the news.

[0221] 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.

[0222] In this invention, the server includes means for collecting news content, means for summarizing the collected news content according to age, means for providing the summarized news content to the information terminal, means for using a generative AI model to generate a summary, means for displaying news content selected by a user on the information terminal, means for marking the displayed news content as read, means for collecting user feedback, means for instructing the generative AI model to generate a summary using a prompt sentence, means for classifying the news content by related topic, means for displaying related news content together, means for saving the news content in a knowledge base, and means for providing summarized news for each related topic to the information terminal. This enables children to understand the news in an easy-to-understand manner and continue reading with interest.

[0223] "News content" refers to information and articles about current events and happenings obtained from a variety of sources.

[0224] "Means of collection" refers to the methods and devices used to obtain news content from news APIs and other information services and store it on a server.

[0225] "Age-appropriate summarization methods" are algorithms or systems that simplify collected news content for specific age groups, making it easier to understand.

[0226] The "means for providing to an information terminal" refers to a method or device for transmitting summarized news content from a server to an information terminal (such as a smartphone or tablet) and displaying it.

[0227] A "generative AI model" is a machine learning model that uses artificial intelligence technology to analyze input text and perform tasks such as summarizing and classifying it.

[0228] A "prompt sentence" is input text used to instruct a generative AI model to perform a specific operation or process.

[0229] The "means for displaying news content selected by the user" refers to an interface or method for displaying news content selected by the user on an information terminal.

[0230] A "means for marking as read" is a method or device for marking news content viewed by a user as read.

[0231] A "means for collecting feedback" is a method or system for obtaining opinions and thoughts from users and transmitting them to a server for storage.

[0232] A "topical classification method" is an algorithm or system that categorizes collected news content into specific themes or topics.

[0233] The "means for displaying related news content together" refers to a method or device for simultaneously displaying other news content related to the news that the user is viewing.

[0234] A "means for storing in a knowledge base" is a method or apparatus for storing categorized news content in a database or other storage system for future reference.

[0235] The present invention relates to a system that collects news content, summarizes it for specific age groups, and provides it to information terminals. This system includes functions such as news collection, preprocessing, classification, summary generation, provision, read management, display of related news, feedback collection, summarization using a generative AI model, and instructions using prompt sentences. Specific embodiments for implementing this invention are described below.

[0236] Gathering news content

[0237] The server periodically sends requests to information providers such as news APIs to retrieve the latest news content, which is then stored in a temporary data store.

[0238] News content preprocessing and classification

[0239] The server preprocesses the retrieved news content using a text analysis library. This preprocessing includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The preprocessed news content is then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[0240] Summary Generation

[0241] The server inputs the classified news content into a generative AI model. The generative AI model then summarizes the news content in an easy-to-understand format based on the user's (parent's or child's) age setting. For example, for an 8-year-old child, the summary might be simplified to something like, "The temperature of the earth is rising. We are all working together to stop this."

[0242] Providing news summaries

[0243] When the device opens the application, it sends a request to the server to retrieve the latest news summary, which is then displayed in a list format on the device's user interface.

[0244] News read status management

[0245] Users can tap on the displayed news content to view the details, then press a button to mark it as "read." The device then sends the "read" mark data to the server, which then stores it in a database. This allows the user's browsing history to be managed.

[0246] View related news

[0247] The server categorizes the summarized news content by topic and stores it in a database so that related news content can be displayed all at once. When a user reads a particular news item, other related news content is also displayed.

[0248] Feedback collection

[0249] Users can fill out a feedback form about news content and press the "Submit" button. For example, they can enter a request such as "I would like this article to be more detailed." The device then sends the feedback data to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[0250] SNS sharing function

[0251] Users can share news content on social media by clicking the "Share" button on the news content details screen. The device displays a social media selection screen and sends a news summary and a sharing link to the social media platform selected by the user. The server then displays the shared news content to other users in a format optimized for the target age group.

[0252] Model Improvement

[0253] The server analyzes the collected feedback data and uses the results to tune the generative AI model, which is then used to generate the next news summary.

[0254] Specific examples

[0255] As a concrete example, consider a 10-year-old child using an app. When the app is opened, a summary article titled "New Discoveries in Science and Technology" ("A new star has been discovered, which may reveal more about the secrets of the universe") is retrieved from the server. The child then reads the article and presses the "Read" button. Related news items, such as "Advances in Science and Technology," are then displayed, which the child can read. The child can also share the article they have read on social media and send feedback. An example of this prompt is as follows:

[0256] Example of input prompt for generative AI model

[0257] Send a request to the API to get the latest news articles and summarize them for kids aged 8-12.

[0258] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0259] Step 1:

[0260] The server periodically sends requests to the news API to retrieve the latest news content. The retrieved news content is stored in a temporary data store. The input is the news content from the news API, and the output is the data stored in the temporary data store. Specifically, it sends an HTTP request, parses the response, and stores it in a database.

[0261] Step 2:

[0262] The server preprocesses the news content retrieved from the temporary data store. Preprocessing includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The input is the news content from the temporary data store, and the output is the preprocessed news content. Specifically, it uses a text analysis library to remove unnecessary information and standardize the format.

[0263] Step 3:

[0264] The server inputs the preprocessed news content into a topic classification model and classifies it into specific categories (e.g., environment, politics, science and technology). The input is the preprocessed news content, and the output is the news content classified by category. The specific operation is to use a machine learning model to determine which category each news content belongs to.

[0265] Step 4:

[0266] The server inputs the classified news content into a generative AI model and summarizes the news content based on the age setting. The input is news content classified by category, and the output is summarized news content appropriate for the age. A prompt sentence is used to instruct the generative AI model to generate a summary. The specific operation is to input an appropriate prompt sentence to the generative AI model and retrieve the generated summary.

[0267] Step 5:

[0268] When the application is opened, the device sends a request to the server to retrieve the latest news summary. The input is the user request, and the output is the news summary for display. The specific operation is to send an HTTP request to retrieve data from the server and display it on the user interface.

[0269] Step 6:

[0270] The user can tap on the displayed news content to view details and press a button to mark it as "read." The input is the user's operation, and the output is the news content marked as read. The specific operation is to tap on the user interface and then press the "read" button.

[0271] Step 7:

[0272] The terminal sends the "read" mark data to the server, and the server stores it in a database. The input is the read mark data, and the output is the data stored in the database. The specific operation is to send the data to the server by sending an HTTP request, and the server stores the read information in the database.

[0273] Step 8:

[0274] The server categorizes the summarized news content by topic and stores it in a database in order to display related news content in one place. The input is the summarized news content, and the output is the news content categorized by topic. The specific operation is to use a classification algorithm to extract related news and store it in the database.

[0275] Step 9:

[0276] The user fills in the feedback form for the news content and presses the "Send" button. The input is the user's feedback, and the output is feedback data from the terminal. The specific operation is to fill in the feedback form on the user interface and press the send button.

[0277] Step 10:

[0278] The terminal sends feedback data to the server, which stores it in a database. The input is the feedback data, and the output is the data stored in the database. The specific operation is to send an HTTP request to the server to send data, and then store the feedback information on the server.

[0279] Step 11:

[0280] Users can share news content on social media by pressing the "Share" button on the news content details screen. The input is the user's share operation, and the output is the news content shared on the social media. The specific operation is that after pressing the share button, an SNS selection screen is displayed, and the news summary and sharing link are sent to the selected SNS.

[0281] Step 12:

[0282] The server analyzes the collected feedback data and tunes the generative AI model based on the results. The input is the feedback data, and the output is the tuned generative AI model. Specifically, the server analyzes the feedback data and reflects the analysis results in the parameters of the generative AI model.

[0283] 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.

[0284] The present invention relates to a system that collects news articles, summarizes them in an easy-to-understand manner for children, and combines them with an emotion engine that recognizes the user's emotions. Hereinafter, embodiments of the present invention will be described with reference to specific examples.

[0285] News article collection

[0286] server:

[0287] The server periodically sends requests to a news source (e.g., a news API) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[0288] News article preprocessing and classification

[0289] server:

[0290] The collected news articles are first preprocessed using a text analysis library, which includes removing noise data (such as advertisements, links, and special characters) and normalizing character encoding. The preprocessed news articles are then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[0291] Summary Generation

[0292] server:

[0293] The classified news articles are then fed into a generative AI model, which then summarizes the article in an easy-to-understand way based on the user's (parent's or child's) age preference. For example, for an 8-year-old child, the article might be simplified to something like, "The earth's temperature is rising. We're all working together to stop this."

[0294] Providing and displaying news summaries

[0295] Device:

[0296] When a user opens the application, the device sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the device's user interface.

[0297] News read status management

[0298] User:

[0299] Users (usually children) tap on a displayed news article to view details, then press a button to mark it as "read."

[0300] Device:

[0301] The "read" mark data is sent from the device to the server and stored in a database on the server side, allowing the user's browsing history to be managed.

[0302] View related news

[0303] server:

[0304] To display related news articles in one place, the server categorizes the summarized news articles by topic and stores them in a database, so that when a user reads a particular article, other related articles are also displayed.

[0305] Feedback collection

[0306] User:

[0307] Users can fill out a feedback form for a news article and press the "Submit" button. For example, they can enter a request such as "I'd like this article to be more detailed."

[0308] Device:

[0309] The feedback data is sent from the device to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[0310] SNS sharing function

[0311] User:

[0312] Users can click the "Share" button on the details screen of a news article to share it on social media.

[0313] Device:

[0314] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[0315] server:

[0316] Shared news articles will be displayed to other users in an age-optimized format.

[0317] Model Improvement

[0318] server:

[0319] The collected feedback data is analyzed and the generative AI model is tuned based on the results. The improved model is used to generate the next news summary.

[0320] Introducing the Emotion Engine

[0321] emotion recognition

[0322] Device:

[0323] While a user is browsing a news article, the device uses a built-in camera and microphone to analyze the user's emotions in real time. The emotion engine analyzes the user's facial expressions and tone of voice to generate emotion data.

[0324] Emotion data analysis and storage

[0325] server:

[0326] The emotion data sent from the emotion engine is analyzed and stored in a database, thereby accumulating a history of the user's emotions regarding news articles.

[0327] Selecting the next news story

[0328] server:

[0329] The next news article to be served is selected based on the accumulated emotional data and the browsing history of news articles. For example, if the user shows a "happy" expression, the system will serve up positive news.

[0330] Specific examples

[0331] As a concrete example, consider a 10-year-old child using the app. While the child is reading a summary article titled "New Report on Climate Change," the emotion engine analyzes the child's facial expressions and determines that the user is interested. A related news article about "The Importance of Recycling" is also displayed, which the child can read. After finishing the article, the child presses the "Read" button, and the emotion data is saved in the database.

[0332] In this way, by summarizing news articles in an easy-to-understand manner and combining it with emotion recognition, the present invention can provide appropriate news articles for children, enhance learning effectiveness, promote communication with parents and teachers, and encourage children to become interested in current events and social situations.

[0333] The processing flow will be explained below.

[0334] Step 1:

[0335] Server: Sends requests to news sources (e.g., news APIs) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[0336] Step 2:

[0337] Server: The retrieved news articles are preprocessed using a text analysis library, which includes removing noise data (advertisements, links, special characters, etc.) and normalizing character encoding.

[0338] Step 3:

[0339] Server: Preprocessed news articles are input into a topic classification model. The model classifies news articles into specific topics (e.g., environment, politics, science and technology). The classification results are stored in a temporary data store.

[0340] Step 4:

[0341] Server: The classified news articles are fed into a generative AI model to be summarized for children. Based on the age setting (e.g., elementary school, middle school), a summary text is generated in an easy-to-understand format.

[0342] Step 5:

[0343] Server: Stores summarized news articles and their metadata (topic, target demographic, etc.) in a database. The stored data is linked to other related news articles.

[0344] Step 6:

[0345] Device: When a user (parent or child) opens the application, the device sends a request to the server to retrieve the latest summary news articles.

[0346] Step 7:

[0347] Terminal: The retrieved news summaries are displayed in a user interface in the form of a list, with each news item showing its title and summary text.

[0348] Step 8:

[0349] User: Taps on a displayed news article to view details, then presses a button to mark it as "read."

[0350] Step 9:

[0351] Terminal: Detects user actions and sends the "read" mark data to the server.

[0352] Step 10:

[0353] Server: The received "read" data is saved in a database and managed as the user's browsing history.

[0354] Step 11:

[0355] Server: Based on the related information of the news article, selects news articles on related topics and prepares them for provision to users.

[0356] Step 12:

[0357] On-device: When a user reads a particular article, other related articles are displayed along with it. Related news is displayed in a list format for easy access.

[0358] Step 13:

[0359] User: Can fill out a form to provide feedback on a news article and press the "Submit" button.

[0360] Step 14:

[0361] Terminal: Sends feedback data to the server.

[0362] Step 15:

[0363] Server: The received feedback data is stored in a database and used as reference data for model improvement.

[0364] Step 16:

[0365] User: Press the "Share" button on the news article details screen to share the news on social media.

[0366] Step 17:

[0367] Device: The SNS selection screen is displayed, and the article summary and sharing link are sent to the SNS selected by the user.

[0368] Step 18:

[0369] Server: News articles shared on social media are displayed to other users in a format optimized for the target age group.

[0370] Step 19:

[0371] Server: Analyzes the collected feedback data and tunes the generative AI model. The improved model is used to generate the next news summary.

[0372] Step 20:

[0373] Device: While the user is viewing a news article, the device uses the built-in camera and microphone to analyze the user's facial expressions and tone of voice in real time.

[0374] Step 21:

[0375] Device: The emotion engine generates the user's emotion data and temporarily stores the analysis results on the device.

[0376] Step 22:

[0377] Terminal: Sends emotion data to the server in real time.

[0378] Step 23:

[0379] Server: Analyzes the emotion data sent from the emotion engine and stores it in a database, thereby accumulating a history of users' emotions toward news articles.

[0380] Step 24:

[0381] Server: Based on the accumulated emotional data and the browsing history of news articles, the server selects the next news article to be provided. For example, if the user shows a "happy" expression, the server will provide positive news.

[0382] Examples:

[0383] For example, if a 10-year-old child is using the app to read a new report on climate change, the emotion engine will analyze the child's facial expressions and determine that the user is interested. Related news about the importance of recycling will also be displayed. After finishing the article, the child can press the "read" button, and the emotion data will be stored in the database, and the next appropriate news item will be displayed.

[0384] Example 2

[0385] 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."

[0386] Existing news article delivery systems often do not adequately summarize articles for children, making them difficult to understand. Furthermore, they are unable to provide news that takes into account user feedback and emotions, making it difficult to sustain interest. Furthermore, they do not adequately provide relevant news articles, resulting in low learning outcomes for users.

[0387] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0388] In this invention, the server includes means for collecting news articles, means for preprocessing the collected news articles, means for classifying the preprocessed news articles using a topic classification model, means for inputting the classified news articles into a generative AI model using prompt sentences and summarizing them according to age, and means for providing the summarized news articles to a terminal. This makes it possible to provide news articles in a format that is easy for children to understand. Furthermore, by utilizing user feedback and emotional data, it is possible to provide more appropriate articles for each user and improve learning effectiveness. Furthermore, by displaying related news articles together, it is possible to deepen the user's understanding.

[0389] A "news article" is written information about a current event or topic obtained from online or offline sources.

[0390] "Preprocessing" is the process of performing data preparation tasks such as removing noise data from the collected text data of news articles and normalizing character encoding.

[0391] A "topic classification model" is a machine learning model used to automatically classify news articles into specific categories (e.g., environment, politics, science and technology, etc.).

[0392] A "prompt" is a text sentence that provides the information or instructions needed to input into a generative AI model.

[0393] A "generative AI model" is an artificial intelligence model that uses natural language processing techniques to summarize or generate information from presented input data.

[0394] "Summarization" refers to extracting the key information from an original news article and restating it in a shorter form.

[0395] A "terminal" is a computing device (e.g., smartphone, tablet, PC, etc.) that displays news articles and accepts user operations.

[0396] "User feedback" refers to information such as opinions, requests, and evaluations provided by users regarding news articles.

[0397] "Related news articles" are other news articles that are related in topic or content to a particular article.

[0398] A "knowledge base" is a database that systematically stores news articles and related data, enabling them to be searched and analyzed.

[0399] The present invention relates to a system for providing appropriate news articles to children and enhancing their learning effect. Specific embodiments of the present invention will be described below.

[0400] News article collection and preprocessing

[0401] server:

[0402] The server periodically sends HTTP requests to information sources, such as news APIs, to retrieve the latest news articles. The retrieved news articles are received as JSON-formatted data and stored in a temporary data store. The server then preprocesses the collected news articles using a text analysis library (e.g., NLTK, SpaCy, etc.). Preprocessing includes removing noise data such as advertisements, links, and special characters, and normalizing character encoding.

[0403] News article classification and summarization

[0404] server:

[0405] The preprocessed news articles are fed into a topic classification model (e.g., environment, politics, science and technology) to be classified into specific categories. The classified news articles are then fed into a generative AI model (e.g., GPT-3) with a prompt. The generative AI model summarizes the articles based on the user's specified age.

[0406] Example prompt sentence:

[0407] "Please provide a quick summary of the following article for an 8-year-old: {News article text}"

[0408] The generated summary articles are stored in a database.

[0409] Providing and displaying news summaries

[0410] Device:

[0411] When a user opens the application, the device sends an HTTP request to the server to retrieve the latest summarized news articles, which are then displayed in a user interface (e.g., a list view).

[0412] News read tracking and feedback collection

[0413] User:

[0414] Users can tap on a news article to view details and press the "mark as read" button after finishing reading. They can also fill out a feedback form for the news article and press the "submit" button. The feedback data includes requests and improvements for the article content.

[0415] Device:

[0416] The "read" mark data and feedback data are sent from the terminal to the server, where the data is stored in a database and the user's browsing history and feedback information are managed.

[0417] Displaying related news and managing knowledge base

[0418] server:

[0419] The server categorizes summarized news articles by topic and stores them in a database. When a user views a particular article, other related articles are also displayed. This display of related news can deepen the user's understanding. The server also provides a knowledge base for systematically storing news articles and related data.

[0420] Introducing an emotion engine and selecting the next news article

[0421] Device:

[0422] While a user is browsing a news article, the system uses the built-in camera and microphone to analyze the user's emotions in real time. The emotion engine analyzes the user's facial expressions and tone of voice to generate emotion data.

[0423] server:

[0424] The emotional data sent from the emotion engine is received and analyzed. The analyzed emotional data is stored in a database, and the user's emotional history regarding news articles is accumulated. The server selects the next news article to be provided based on the accumulated emotional data and the news article viewing history. For example, if the user shows a happy expression, positive news will be provided first.

[0425] Specific examples

[0426] As a concrete example, consider a 10-year-old child using the app. While the child is viewing a summary of a "new report on climate change," the emotion engine analyzes the child's facial expressions and determines that the user is interested. Related news articles about the importance of recycling are also displayed, which the child can read. After finishing the article, the child presses the "read" button, and the emotion data is stored in the database. In this way, by summarizing news articles in an easy-to-understand manner and combining it with emotion recognition, the app can provide appropriate news articles for children and improve their learning. It can also promote communication with parents and teachers, encouraging children to become interested in current events and social situations.

[0427] Thus, the present invention is a system that provides users with an innovative news article delivery experience by integrating a series of processes, including news article collection, preprocessing, classification, summarization, delivery, emotion recognition, and display of related news.

[0428] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0429] Step 1:

[0430] News article collection

[0431] The server periodically sends HTTP requests to the news API to retrieve the latest news article data in JSON format, which is then stored in a temporary data store on the server.

[0432] Input: A request to the News API

[0433] Data processing: Obtaining JSON data from the news API

[0434] Output: Retrieved news article data

[0435] Specifically, it sets up a regular scheduled job to access the news API every hour or every day.

[0436] Step 2:

[0437] News article preprocessing

[0438] The server uses a text analysis library (e.g., NLTK, SpaCy, etc.) to preprocess the news article data, removing noise data (e.g., advertisements, links, special characters, etc.) and normalizing character encoding.

[0439] Input: Acquired news article data

[0440] Data processing: Removal of noise data, normalization of character encoding

[0441] Output: Preprocessed news article data

[0442] Specifically, we remove noise data using regular expressions and split words using SpaCy's tokenizer.

[0443] Step 3:

[0444] News article classification

[0445] The server uses a topic classification model to classify the preprocessed news article data into specific categories (e.g., environment, politics, science and technology, etc.).

[0446] Input: Preprocessed news article data

[0447] Data processing: Category classification using topic classification model

[0448] Output: Categorized news article data

[0449] Specifically, the operation involves applying a text classification algorithm using a machine learning model.

[0450] Step 4:

[0451] News article summary generation

[0452] The server inputs the classified news article data into a generative AI model (e.g., GPT-3) and summarizes the article based on the user's specified age. Specific summarization instructions are given to the generative AI model using prompt sentences.

[0453] Input: Categorized news article data, prompt

[0454] Data Computation: Generative AI Models for Summarization

[0455] Output: Summarized news article data

[0456] Example prompt: "Please provide a brief summary of the following article for an 8-year-old: {news article text}"

[0457] Step 5:

[0458] Providing news summaries

[0459] The terminal sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the terminal's user interface.

[0460] Input: Summary news article data on the server

[0461] Data processing: None

[0462] Output: A summary news article displayed on your terminal

[0463] Specifically, news data in JSON format is retrieved via an HTTP request and displayed in a list view.

[0464] Step 6:

[0465] News read status management

[0466] Users tap on a news article to view the details, and when they are finished reading, they press the "read" button. The device sends the "read" mark data to the server, which stores it in a database.

[0467] Input: User presses the read button

[0468] Data processing: Saving read information to a database

[0469] Output: Updated user browsing history

[0470] Specifically, the read information is sent to the server in JSON format as a POST request.

[0471] Step 7:

[0472] Feedback collection

[0473] The user fills in the feedback form for the news article and presses the "Submit" button. The terminal sends the feedback data to the server, where it is stored in a database.

[0474] Input: User feedback data

[0475] Data processing: Saving feedback data to a database

[0476] Output: Updated feedback information

[0477] Specifically, the input contents of the feedback form are sent to the server in JSON format as a POST request.

[0478] Step 8:

[0479] View related news

[0480] The server categorizes summarized news articles by topic and stores them in a database. When a user views a particular article, other related articles are also retrieved and displayed.

[0481] Input: Summary news article data

[0482] Data processing: Acquisition of related news article information

[0483] Output: User interface with related news displayed

[0484] Specifically, news articles belonging to the same category are retrieved from the server as "related articles" via a query.

[0485] Step 9:

[0486] Emotion data analysis and storage

[0487] While a user is viewing a news article, their emotions are analyzed in real time using the device's built-in camera and microphone. The server receives the analyzed emotional data and stores it in a database.

[0488] Input: User facial and voice data

[0489] Data processing: Emotion data analysis using an emotion engine

[0490] Output: Emotion data stored in a database

[0491] Specifically, emotions are analyzed using facial recognition technology and voice emotion recognition technology and stored in a database.

[0492] Step 10:

[0493] Selecting the next news story

[0494] The server selects the next news article to serve based on the accumulated emotional data and the browsing history of news articles. For example, if the user shows a happy expression, it will prioritize positive news articles.

[0495] Input: Emotion data, news article browsing history

[0496] Data processing: Selecting the next news article to be served

[0497] Output: Selected news article data

[0498] Specifically, we apply a recommendation algorithm based on sentiment analysis.

[0499] (Application example 2)

[0500] 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."

[0501] Current news article distribution systems have difficulty providing appropriate and easy-to-understand summaries for children. Furthermore, they lack mechanisms for grasping the degree to which children understand a news article or the emotions it evokes. As a result, they are unable to fully stimulate children's interest or enhance their learning. Another issue is that they are unable to provide personalized news article content, resulting in a uniform distribution system.

[0502] 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.

[0503] In this invention, the server includes means for collecting news articles, means for summarizing the collected news articles according to age, means for providing the summarized news articles to the terminal, means for recognizing the user's emotions while the summarized news articles are displayed, and means for analyzing and saving the emotion data. This makes it possible to provide news articles in a format that is interesting and easy for children to understand, and to personalize the next news article to be provided based on the emotion data.

[0504] "News articles" are text data about current events and happenings distributed by news organizations and information providers.

[0505] "Means of collection" refers to the ability to obtain the latest news articles from news APIs and other information sources via the network.

[0506] "Age-appropriate summarization" refers to the function of making collected news articles easier to understand and concise for a specific age group.

[0507] "Means for providing" refers to the function of delivering summarized news articles to user terminals so that they can be viewed.

[0508] "Device" refers to a device such as a smartphone, tablet, or computer that a user uses to view news articles.

[0509] "Means for recognizing emotions" refers to a function that uses a camera or microphone to analyze facial expressions and tone of voice when a user is viewing a news article, in order to determine emotions.

[0510] "Emotional data" refers to digital data that indicates the emotional state of a user analyzed from facial expressions, tone of voice, etc.

[0511] "Means for analyzing and storing" refers to the function for processing the recognized emotional data and storing it in storage such as a database.

[0512] "Personalization" refers to optimizing and providing content and information to suit the interests and concerns of each individual user.

[0513] The present invention relates to a system that summarizes news articles in an easy-to-understand manner for children and combines it with an emotion engine that recognizes the user's emotions. To effectively implement this invention, the server, terminals, and users must work together to execute each step.

[0514] server

[0515] The server collects news articles, preprocesses them, generates summaries, and analyzes and stores sentiment data.

[0516] 1. Collect news articles periodically from a news API and store them in a temporary data store. Here, you can use an existing news API such as NewsAPI.

[0517] 2. Preprocess the collected news articles using a text analysis library such as TextBlob to remove noise data and normalize character encoding.

[0518] 3. Classify the preprocessed news articles using a topic classification model and generate age-appropriate summaries using a generative AI model (e.g., OpenAI GPT-3), with an example prompt such as "Summarize the news article for a 10-year-old: [insert news article here]."

[0519] 4. The summarized news articles are stored in a database.

[0520] 5. While the user is browsing a news article, the emotion engine analyzes the user's facial expressions and tone of voice through the camera and microphone to generate emotion data.

[0521] 6. Emotional data is analyzed and stored in a database along with browsing history and related news.

[0522] Terminal

[0523] The device displays news articles and collects and transmits emotion data through a user interface.

[0524] 1. When a user opens the application, it sends a request to the server to get the latest summary news articles.

[0525] 2. The retrieved news articles are displayed in a list format on the user interface.

[0526] 3. When a user selects a news article and views its details, the device's built-in camera and microphone are used to analyze the user's emotions in real time and send the data to the emotion engine.

[0527] 4. After the user reads a news article, a button is provided to mark it as "read" and the data is sent to the server.

[0528] User

[0529] Users (mainly children) operate the device to read news articles and cooperate in collecting emotion data.

[0530] 1. Select the news article that interests you from the list of news articles.

[0531] 2. Have the students read news articles and analyze emotional data via a camera or microphone.

[0532] 3. After reading the details of the news article, press the "read" button to mark the article and provide feedback.

[0533] As a concrete example, consider a 10-year-old child viewing a summary article titled "New Report on Climate Change." While the child is reading the article, the emotion engine analyzes the child's facial expressions and determines that the child is interested. It also displays a related news article about "The Importance of Recycling," which the child can continue reading. This emotion information and browsing history are stored in a database and used for future news distribution.

[0534] In this way, by operating this system in cooperation with the server, terminals, and users, news articles can be provided in a format that is easy for children to understand and that will interest them, thereby improving the effectiveness of their learning.

[0535] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0536] Step 1:

[0537] The server collects news articles by periodically sending requests to the news API to retrieve the latest news articles. This request requires an API key and a specific endpoint. The input is the API key and the endpoint, and the output is the retrieved news article. The news article is then stored in a temporary data store.

[0538] Step 2:

[0539] The server preprocesses the collected news articles. Specifically, it uses a text analysis library such as TextBlob to remove noise data from the news articles (such as advertisements, links, and special characters) and normalizes character encoding. The input is the collected news articles, and the output is the preprocessed news articles. These preprocessed news articles are then input into a topic classification model.

[0540] Step 3:

[0541] The server categorizes the preprocessed news articles by topic. It uses a topic classification model to classify the news articles into specific categories (e.g., environment, politics, science and technology). The input is the preprocessed news articles, and the output is the classified news articles. The classified news articles are stored in a database.

[0542] Step 4:

[0543] The server inputs the classified news article into a generative AI model to summarize it for children. The input is the classified news article and the user's (mainly child) age preference. The generative AI model (e.g., OpenAI GPT-3) generates a summary using the prompt "Summarize the news article for a 10-year-old child: [insert news article here]". The output is a summarized news article. This summarized news article is stored in a database.

[0544] Step 5:

[0545] The device displays a list of news articles through a user interface. When a user opens the application, it sends a request to the server to retrieve the latest summarized news articles. The input is the user request, and the output is the list of retrieved news articles. The news articles are displayed in list format on the device screen.

[0546] Step 6:

[0547] The device recognizes emotions when the user selects a news article and views its details. When the user selects a news article, the device's built-in camera and microphone are used to analyze the user's facial expression and tone of voice in real time. The input is the news article selection and the user's facial expression and tone of voice. The output is recognized emotion data, which is sent to the server.

[0548] Step 7:

[0549] The server analyzes the emotion data and stores it in a database. The input is the transmitted emotion data, and the output is the analyzed emotion data. This data is used to select the next news article.

[0550] Step 8:

[0551] The server selects the next news article to be served based on the analyzed emotion data and the browsing history of news articles. The input is emotion data and browsing history, and the output is the next news article to be served. This article will be served the next time the user opens the application.

[0552] 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.

[0553] 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.

[0554] 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.

[0555] [Second embodiment]

[0556] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0557] 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.

[0558] 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).

[0559] 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.

[0560] 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.

[0561] 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).

[0562] 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.

[0563] 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.

[0564] 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.

[0565] 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.

[0566] In the smart glasses 214, 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.

[0567] 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."

[0568] The present invention relates to a system for collecting news articles, summarizing them in an easy-to-understand manner for children, and providing them to a terminal. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes embodiments of the present invention with specific examples.

[0569] News article collection

[0570] server:

[0571] The server periodically sends requests to a news source (e.g., a news API) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[0572] News article preprocessing and classification

[0573] server:

[0574] The collected news articles are first preprocessed using a text analysis library, which includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The preprocessed news articles are then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[0575] Summary Generation

[0576] server:

[0577] The classified news articles are then fed into a generative AI model, which then summarizes the article in an easy-to-understand way based on the user's (parent's or child's) age preference. For example, for an 8-year-old child, the article might be simplified to something like, "The earth's temperature is rising. We're all working together to stop this."

[0578] Providing news summaries

[0579] Device:

[0580] When a user opens the application, the device sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the device's user interface.

[0581] News read status management

[0582] User:

[0583] Users (usually children) can tap on a displayed news article to view details, then press a button to mark it as "read."

[0584] Device:

[0585] The "read" mark data is sent from the device to the server and stored in a database on the server side, allowing the user's browsing history to be managed.

[0586] View related news

[0587] server:

[0588] To display related news articles in one place, the server categorizes the summarized news articles by topic and stores them in a database, so that when a user reads a particular article, other related articles are also displayed.

[0589] Feedback collection

[0590] User:

[0591] Users can fill out a feedback form for a news article and press the "Submit" button. For example, they can enter a request such as "I'd like this article to be more detailed."

[0592] Device:

[0593] The feedback data is sent from the device to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[0594] SNS sharing function

[0595] User:

[0596] Users can click the "Share" button on the details screen of a news article to share it on social media.

[0597] Device:

[0598] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[0599] server:

[0600] Shared news articles will be displayed to other users in an age-optimized format.

[0601] Model Improvement

[0602] server:

[0603] The collected feedback data is analyzed and the generative AI model is tuned based on the results. The improved model is used to generate the next news summary.

[0604] Specific examples

[0605] As a concrete example, consider a 10-year-old child using an app. When the app is opened, a summary article titled "New Report on Climate Change" ("The temperature of the earth has been rising recently. We are all working hard to prevent this.") is retrieved from the server. The child then reads the article and presses the "Read" button. A related news item, "The Importance of Recycling," is then displayed, which the child can read.

[0606] In this way, the present invention summarizes news articles in an easy-to-understand manner and provides them to children, thereby increasing children's opportunities to come into contact with current events and social situations and promoting communication with parents and teachers.

[0607] The processing flow will be explained below.

[0608] Step 1:

[0609] Server: Sends requests to news sources (e.g., news APIs) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[0610] Step 2:

[0611] Server: The retrieved news articles are preprocessed using a text analysis library, which includes removing noise data (advertisements, links, special characters, etc.) and normalizing character encoding.

[0612] Step 3:

[0613] Server: Preprocessed news articles are input into a topic classification model. The model classifies news articles into specific topics (e.g., environment, politics, science and technology). The classification results are stored in a temporary data store.

[0614] Step 4:

[0615] Server: The classified news articles are fed into a generative AI model to be summarized for children. Based on the age setting (e.g., elementary school, middle school), a summary text is generated in an easy-to-understand format.

[0616] Step 5:

[0617] Server: Stores summarized news articles and their metadata (topic, target demographic, etc.) in a database. The stored data is linked to other related news articles.

[0618] Step 6:

[0619] Device: When a user (parent or child) opens the application, the device sends a request to the server to retrieve the latest summary news articles.

[0620] Step 7:

[0621] Terminal: The retrieved news summaries are displayed in a user interface in the form of a list, with each news item showing its title and summary text.

[0622] Step 8:

[0623] User: Taps on a displayed news article to view details, then presses a button to mark it as "read."

[0624] Step 9:

[0625] Terminal: Detects user actions and sends the "read" mark data to the server.

[0626] Step 10:

[0627] Server: The received "read" data is saved in a database and managed as the user's browsing history.

[0628] Step 11:

[0629] Server: Based on the related information of the news article, selects news articles on related topics and prepares them for provision to users.

[0630] Step 12:

[0631] On-device: When a user reads a particular article, other related articles are displayed along with it. Related news is displayed in a list format for easy access.

[0632] Step 13:

[0633] User: Can fill out a form to provide feedback on a news article and press the "Submit" button.

[0634] Step 14:

[0635] Terminal: Sends feedback data to the server.

[0636] Step 15:

[0637] Server: The received feedback data is stored in a database and used as reference data for model improvement.

[0638] Step 16:

[0639] User: Press the "Share" button on the news article details screen to share the news on social media.

[0640] Step 17:

[0641] Device: The SNS selection screen is displayed, and the article summary and sharing link are sent to the SNS selected by the user.

[0642] Step 18:

[0643] Server: News articles shared on social media are displayed to other users in a format optimized for the target age group.

[0644] Step 19:

[0645] Server: Analyzes the collected feedback data and tunes the generative AI model. The improved model is used to generate the next news summary.

[0646] Example 1

[0647] 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."

[0648] Conventional news article delivery systems do not provide summaries based on specific age settings, making it difficult for users to understand news articles in a way that is appropriate for them, especially children. Furthermore, they lack the ability to manage read statuses and collect feedback when users view articles, making it difficult to provide content that is tailored to the user's usage. Furthermore, they lack the ability to efficiently display related news articles, and do not provide sufficient support for deepening understanding.

[0649] 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.

[0650] In this invention, the server includes means for collecting news articles, means for preprocessing the collected news articles using a text analysis library, means for inputting the preprocessed news articles into a topic classification model and classifying them into specific categories, means for inputting the classified news articles into a generative AI model and summarizing them based on user settings, means for providing the summarized news articles to a terminal, and means for displaying them on a user interface. This enables the provision of easy-to-understand summarized news articles tailored to the user's age. Additionally, the server can mark news articles selected by the user as read, send and store the read data on the server, and collect feedback, enabling the provision of content based on more specific user usage patterns. Furthermore, the automatic classification and display of related news articles enables deeper understanding and information acquisition.

[0651] "News article gathering means" means hardware or software for periodically retrieving current news articles from online news sources.

[0652] A "text analysis library" is a software tool used to preprocess collected news articles, such as removing noise data and normalizing character encoding.

[0653] "Topic classification model" refers to a machine learning model for classifying preprocessed news articles into specific categories.

[0654] A "generative AI model" refers to an artificial intelligence model that takes categorized news articles as input and provides an easy-to-understand summary based on the user's age settings.

[0655] "Summarization method" refers to the process and software that uses a generative AI model to summarize news articles in a format that is easy for users to understand.

[0656] "Means for providing to the terminal" refers to the processes and software for transmitting summarized news articles from the server to the terminal and for the terminal to receive and display them.

[0657] "User Interface" refers to the on-screen interface through which a user can view news articles, provide feedback, and access other features.

[0658] "Means to mark as read" refers to the functionality and processes that allow a user to mark a news article as "read" and save that information.

[0659] "Feedback collection means" refers to the process and software for collecting user-entered feedback on news articles and transmitting and storing that data on a server.

[0660] "Means for categorizing by relevant topic" refers to the process and software for re-categorizing summarized news articles by topic and storing them in a database.

[0661] "Means for displaying related news articles together" refers to processes and software that allow a user to simultaneously view other articles related to a particular article as the user reads it.

[0662] "Means for storing in knowledge base" refers to the functions and processes by which processed news articles and related information are stored in a database for later retrieval.

[0663] The present invention relates to a system for collecting news articles, summarizing them in an easy-to-understand manner for children, and providing them to a terminal. Specific embodiments for carrying out the present invention will be described below.

[0664] News article collection

[0665] server:

[0666] The server uses the requests library to periodically send HTTP requests to news sources (e.g., news APIs) to retrieve new news articles, which are then stored in a temporary data store (e.g., an SQL database) on the server, for example in the form of a pandas dataframe.

[0667] News article preprocessing and classification

[0668] server:

[0669] The collected news articles are first preprocessed using text analysis libraries such as NLTK and spaCy, which include removing noise data (such as advertisements and links) using regular expressions and standardizing character encoding.

[0670] The preprocessed news articles are then fed into a topic classification model (e.g., BERT, GPT-3) to classify them into specific categories (e.g., environment, politics, science and technology).

[0671] Summary Generation

[0672] server:

[0673] The classified news articles are fed into a generative AI model (e.g., OpenAI's GPT-3), which then summarizes the article in an easy-to-understand way based on the user's (parent or child's) age preference. Examples of prompts include:

[0674] "Summarize the following news article in a way that would be understandable for a 10-year-old.

[0675] News Article:

[0676] As the Earth's temperature rises, various measures are being taken around the world...

[0677] "

[0678] The generative AI model uses this prompt to summarize the news article as something like, "The temperature of the earth has been rising recently. We are all working hard to prevent this."

[0679] Providing news summaries

[0680] Device:

[0681] When a user opens the application, the device sends an HTTP request to the server to retrieve the latest news summary articles, which are then displayed in a list format in the device's user interface using RecyclerView and ListView.

[0682] News read status management

[0683] User:

[0684] Users (mainly children) can tap on the displayed news article to view details, and then press the "read" button after viewing.

[0685] Device:

[0686] The "read" mark data is sent from the device to the server by sending an HTTP POST request, and the server stores this data in a database and manages it as the user's browsing history.

[0687] View related news

[0688] server:

[0689] The server recategorizes the summarized news articles by topic and stores them in a database, allowing users to simultaneously view other related articles as they read a particular article.

[0690] Feedback collection

[0691] User:

[0692] Users can enter their feedback on the news article in the form and press the "Submit" button. They can enter requests such as "I wish this article had more details."

[0693] Device:

[0694] The device sends the feedback data to the server, which stores it in a database and uses it to improve the generative AI model.

[0695] SNS sharing function

[0696] User:

[0697] Users can click the "Share" button on the details screen of a news article to share it on social media.

[0698] Device:

[0699] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[0700] Model Improvement

[0701] server:

[0702] The server analyzes the collected feedback data and uses the results to tune the generative AI model, which is then used to generate the next news summary.

[0703] Specific examples

[0704] For example, if a 10-year-old child were to use this system, the process would go something like this:

[0705] 1. When a child opens the app, the device retrieves the latest summary news article (e.g., "The temperature of the earth has been rising recently. We are all working hard to prevent this.") from the server.

[0706] 2. The child reads the article and then presses the "read" button, which saves the read information on the server.

[0707] 3. Additionally, a related news item, "The Importance of Recycling," is displayed, which children can also read.

[0708] Thus, the present invention is a system that provides children with easy-to-understand summaries of news articles, helps them understand current affairs and social situations, and promotes communication with parents and teachers.

[0709] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0710] Step 1:

[0711] News article collection

[0712] The server sends HTTP GET requests to a news source (e.g., a news API). The input is the API endpoint and authentication information, and the server executes this periodically. The output is the JSON data of the retrieved news articles. The server converts this data into a pandas dataframe and stores it in a temporary data store (e.g., a SQL database).

[0713] Step 2:

[0714] News article preprocessing

[0715] The server retrieves news articles from the temporary data store and preprocesses them using a text analysis library (e.g., NLTK, spaCy). The input is the news article data saved in the previous step, and the output is the preprocessed clean text data. Specific operations include removing noise data (advertisements, links, etc.) using regular expressions and standardizing character encoding.

[0716] Step 3:

[0717] Topic Classification

[0718] The server inputs preprocessed news articles into a topic classification model (e.g., BERT, GPT-3). The input is clean text data, and the output is data classified into specific categories (e.g., environment, politics, science and technology). Specifically, the text data is converted into feature vectors, and classification results are obtained using a pre-trained model.

[0719] Step 4:

[0720] Summary Generation

[0721] The server inputs the classified news article into a generative AI model (e.g., OpenAI's GPT-3). The input is the classified news article and a prompt, and the output is a summarized news article based on the user's age preference. Specifically, the prompt is generated in the following format:

[0722] "Summarize the following news article in a way that would be understandable for a 10-year-old.

[0723] News Article:

[0724] As the Earth's temperature rises, various measures are being taken around the world...

[0725] "

[0726] The generative AI model uses this information to summarize the news, and the server stores the summarized article in a database.

[0727] Step 5:

[0728] Providing news summaries

[0729] When a user opens the application on a device, the device sends an HTTP GET request to the server to request a summary news article. The input is the user's request, and the output is the latest summary news article data. The server sends this in JSON format to the device, which then displays it using RecyclerView or ListView.

[0730] Step 6:

[0731] News read status management

[0732] Users can tap on a news article to view details and then press the "mark as read" button after viewing. The input is the user's action, and the output is the updated read data. The device sends this data as an HTTP POST request to the server, which stores it in a database.

[0733] Step 7:

[0734] View related news

[0735] The server reclassifies summarized news articles by topic and stores them in a database. The input is the summarized news article data, and the output is the data classified by topic. When a user reads a particular article, the device sends a request to the server to display related news as well. The server sends the corresponding related news in JSON format to the device, and the device displays it.

[0736] Step 8:

[0737] Feedback collection

[0738] The user enters their feedback on the news article into the form and presses the "Submit" button. The input is the user's feedback content, and the output is the feedback data. The terminal sends the feedback data to the server via an HTTP POST request, and the server stores it in a database.

[0739] Step 9:

[0740] SNS sharing function

[0741] The user presses the "Share" button on the news article details screen, which takes them to the SNS sharing selection screen. The input is the user's operation, and the output is a sharing link and article summary. The device uses the SNS API to send the article summary and sharing link.

[0742] Step 10:

[0743] Model Improvement

[0744] The server analyzes the collected feedback data and tunes the generative AI model. The input is the feedback data and the output is the improved model. The server applies the new model to generate the next news summary.

[0745] (Application example 1)

[0746] 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."

[0747] In today's world, it is important for children to have opportunities to read news articles, but news content is often too technical and difficult to understand. Therefore, there is a need for an easy-to-understand news summary system that helps children understand and be interested in the news.

[0748] 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.

[0749] In this invention, the server includes means for collecting news content, means for summarizing the collected news content according to age, means for providing the summarized news content to the information terminal, means for using a generative AI model to generate a summary, means for displaying news content selected by a user on the information terminal, means for marking the displayed news content as read, means for collecting user feedback, means for instructing the generative AI model to generate a summary using a prompt sentence, means for classifying the news content by related topic, means for displaying related news content together, means for saving the news content in a knowledge base, and means for providing summarized news for each related topic to the information terminal. This enables children to understand the news in an easy-to-understand manner and continue reading with interest.

[0750] "News content" refers to information and articles about current events and happenings obtained from a variety of sources.

[0751] "Means of collection" refers to the methods and devices used to obtain news content from news APIs and other information services and store it on a server.

[0752] "Age-appropriate summarization methods" are algorithms or systems that simplify collected news content for specific age groups, making it easier to understand.

[0753] The "means for providing to an information terminal" refers to a method or device for transmitting summarized news content from a server to an information terminal (such as a smartphone or tablet) and displaying it.

[0754] A "generative AI model" is a machine learning model that uses artificial intelligence technology to analyze input text and perform tasks such as summarizing and classifying it.

[0755] A "prompt sentence" is input text used to instruct a generative AI model to perform a specific operation or process.

[0756] The "means for displaying news content selected by the user" refers to an interface or method for displaying news content selected by the user on an information terminal.

[0757] A "means for marking as read" is a method or device for marking news content viewed by a user as read.

[0758] A "means for collecting feedback" is a method or system for obtaining opinions and thoughts from users and transmitting them to a server for storage.

[0759] A "topical classification method" is an algorithm or system that categorizes collected news content into specific themes or topics.

[0760] The "means for displaying related news content together" refers to a method or device for simultaneously displaying other news content related to the news that the user is viewing.

[0761] A "means for storing in a knowledge base" is a method or apparatus for storing categorized news content in a database or other storage system for future reference.

[0762] The present invention relates to a system that collects news content, summarizes it for specific age groups, and provides it to information terminals. This system includes functions such as news collection, preprocessing, classification, summary generation, provision, read management, display of related news, feedback collection, summarization using a generative AI model, and instructions using prompt sentences. Specific embodiments for implementing this invention are described below.

[0763] Gathering news content

[0764] The server periodically sends requests to information providers such as news APIs to retrieve the latest news content, which is then stored in a temporary data store.

[0765] News content preprocessing and classification

[0766] The server preprocesses the retrieved news content using a text analysis library. This preprocessing includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The preprocessed news content is then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[0767] Summary Generation

[0768] The server inputs the classified news content into a generative AI model. The generative AI model then summarizes the news content in an easy-to-understand format based on the user's (parent's or child's) age setting. For example, for an 8-year-old child, the summary might be simplified to something like, "The temperature of the earth is rising. We are all working together to stop this."

[0769] Providing news summaries

[0770] When the device opens the application, it sends a request to the server to retrieve the latest news summary, which is then displayed in a list format on the device's user interface.

[0771] News read status management

[0772] Users can tap on the displayed news content to view the details, then press a button to mark the news as "read." The device then sends the "read" mark data to the server, which then stores it in a database. This allows the user's browsing history to be managed.

[0773] View related news

[0774] The server categorizes the summarized news content by topic and stores it in a database so that related news content can be displayed all at once. When a user reads a particular news item, other related news content is also displayed.

[0775] Feedback collection

[0776] Users can fill out a feedback form about news content and press the "Submit" button. For example, they can enter a request such as "I would like this article to be more detailed." The device then sends the feedback data to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[0777] SNS sharing function

[0778] Users can share news content on social media by clicking the "Share" button on the news content details screen. The device displays a social media selection screen and sends a news summary and a sharing link to the social media platform selected by the user. The server then displays the shared news content to other users in a format optimized for the target age group.

[0779] Model Improvement

[0780] The server analyzes the collected feedback data and uses the results to tune the generative AI model, which is then used to generate the next news summary.

[0781] Specific examples

[0782] As a concrete example, consider a 10-year-old child using an app. When the app is opened, a summary article titled "New Discoveries in Science and Technology" ("A new star has been discovered, which may reveal more about the secrets of the universe") is retrieved from the server. The child then reads the article and presses the "Read" button. Related news items, such as "Advances in Science and Technology," are then displayed, which the child can read. The child can also share the article they have read on social media and send feedback. An example of this prompt is as follows:

[0783] Example of input prompt for generative AI model

[0784] Send a request to the API to get the latest news articles and summarize them for kids aged 8-12.

[0785] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0786] Step 1:

[0787] The server periodically sends requests to the news API to retrieve the latest news content. The retrieved news content is stored in a temporary data store. The input is the news content from the news API, and the output is the data stored in the temporary data store. Specifically, it sends an HTTP request, parses the response, and stores it in a database.

[0788] Step 2:

[0789] The server preprocesses the news content retrieved from the temporary data store. Preprocessing includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The input is the news content from the temporary data store, and the output is the preprocessed news content. Specifically, it uses a text analysis library to remove unnecessary information and standardize the format.

[0790] Step 3:

[0791] The server inputs the preprocessed news content into a topic classification model and classifies it into specific categories (e.g., environment, politics, science and technology). The input is the preprocessed news content, and the output is the news content classified by category. The specific operation is to use a machine learning model to determine which category each news content belongs to.

[0792] Step 4:

[0793] The server inputs the classified news content into a generative AI model and summarizes the news content based on the age setting. The input is news content classified by category, and the output is summarized news content appropriate for the age. A prompt sentence is used to instruct the generative AI model to generate a summary. The specific operation is to input an appropriate prompt sentence to the generative AI model and retrieve the generated summary.

[0794] Step 5:

[0795] When the application is opened, the device sends a request to the server to retrieve the latest news summary. The input is the user request, and the output is the news summary for display. The specific operation is to send an HTTP request to retrieve data from the server and display it on the user interface.

[0796] Step 6:

[0797] The user can tap on the displayed news content to view details and press a button to mark it as "read." The input is the user's operation, and the output is the news content marked as read. The specific operation is to tap on the user interface and then press the "read" button.

[0798] Step 7:

[0799] The terminal sends the "read" mark data to the server, and the server stores it in a database. The input is the read mark data, and the output is the data stored in the database. The specific operation is to send the data to the server by sending an HTTP request, and the server stores the read information in the database.

[0800] Step 8:

[0801] The server categorizes the summarized news content by topic and stores it in a database in order to display related news content in one place. The input is the summarized news content, and the output is the news content categorized by topic. The specific operation is to use a classification algorithm to extract related news and store it in the database.

[0802] Step 9:

[0803] The user fills in the feedback form for the news content and presses the "Send" button. The input is the user's feedback, and the output is feedback data from the terminal. The specific operation is to fill in the feedback form on the user interface and press the send button.

[0804] Step 10:

[0805] The terminal sends feedback data to the server, which stores it in a database. The input is the feedback data, and the output is the data stored in the database. The specific operation is to send an HTTP request to the server to send data, and then store the feedback information on the server.

[0806] Step 11:

[0807] Users can share news content on social media by pressing the "Share" button on the news content details screen. The input is the user's share operation, and the output is the news content shared on the social media. The specific operation is that after pressing the share button, an SNS selection screen is displayed, and the news summary and sharing link are sent to the selected SNS.

[0808] Step 12:

[0809] The server analyzes the collected feedback data and tunes the generative AI model based on the results. The input is the feedback data, and the output is the tuned generative AI model. Specifically, the server analyzes the feedback data and reflects the analysis results in the parameters of the generative AI model.

[0810] 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.

[0811] The present invention relates to a system that collects news articles, summarizes them in an easy-to-understand manner for children, and combines them with an emotion engine that recognizes the user's emotions. Hereinafter, embodiments of the present invention will be described with reference to specific examples.

[0812] News article collection

[0813] server:

[0814] The server periodically sends requests to a news source (e.g., a news API) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[0815] News article preprocessing and classification

[0816] server:

[0817] The collected news articles are first preprocessed using a text analysis library, which includes removing noise data (such as advertisements, links, and special characters) and normalizing character encoding. The preprocessed news articles are then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[0818] Summary Generation

[0819] server:

[0820] The classified news articles are then fed into a generative AI model, which then summarizes the article in an easy-to-understand way based on the user's (parent's or child's) age preference. For example, for an 8-year-old child, the article might be simplified to something like, "The earth's temperature is rising. We're all working together to stop this."

[0821] Providing and displaying news summaries

[0822] Device:

[0823] When a user opens the application, the device sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the device's user interface.

[0824] News read status management

[0825] User:

[0826] Users (usually children) tap on a displayed news article to view details, then press a button to mark it as "read."

[0827] Device:

[0828] The "read" mark data is sent from the device to the server and stored in a database on the server side, allowing the user's browsing history to be managed.

[0829] View related news

[0830] server:

[0831] To display related news articles in one place, the server categorizes the summarized news articles by topic and stores them in a database, so that when a user reads a particular article, other related articles are also displayed.

[0832] Feedback collection

[0833] User:

[0834] Users can fill out a feedback form for a news article and press the "Submit" button. For example, they can enter a request such as "I'd like this article to be more detailed."

[0835] Device:

[0836] The feedback data is sent from the device to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[0837] SNS sharing function

[0838] User:

[0839] Users can click the "Share" button on the details screen of a news article to share it on social media.

[0840] Device:

[0841] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[0842] server:

[0843] Shared news articles will be displayed to other users in an age-optimized format.

[0844] Model Improvement

[0845] server:

[0846] The collected feedback data is analyzed and the generative AI model is tuned based on the results. The improved model is used to generate the next news summary.

[0847] Introducing the Emotion Engine

[0848] emotion recognition

[0849] Device:

[0850] While a user is browsing a news article, the device uses a built-in camera and microphone to analyze the user's emotions in real time. The emotion engine analyzes the user's facial expressions and tone of voice to generate emotion data.

[0851] Emotion data analysis and storage

[0852] server:

[0853] The emotional data sent from the emotion engine is analyzed and stored in a database, thereby accumulating a history of the user's emotions regarding news articles.

[0854] Selecting the next news story

[0855] server:

[0856] The next news article to be served is selected based on the accumulated emotional data and the browsing history of news articles. For example, if the user shows a "happy" expression, positive news will be served.

[0857] Specific examples

[0858] As a concrete example, consider a 10-year-old child using the app. While the child is reading a summary article titled "New Report on Climate Change," the emotion engine analyzes the child's facial expressions and determines that the user is interested. A related news article about "The Importance of Recycling" is also displayed, which the child can read. After finishing the article, the child presses the "Read" button, and the emotion data is saved in the database.

[0859] In this way, by summarizing news articles in an easy-to-understand manner and combining it with emotion recognition, the present invention can provide appropriate news articles for children, enhance learning effectiveness, promote communication with parents and teachers, and encourage children to become interested in current events and social situations.

[0860] The processing flow will be explained below.

[0861] Step 1:

[0862] Server: Sends requests to news sources (e.g., news APIs) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[0863] Step 2:

[0864] Server: The retrieved news articles are preprocessed using a text analysis library, which includes removing noise data (advertisements, links, special characters, etc.) and normalizing character encoding.

[0865] Step 3:

[0866] Server: Preprocessed news articles are input into a topic classification model. The model classifies news articles into specific topics (e.g., environment, politics, science and technology). The classification results are stored in a temporary data store.

[0867] Step 4:

[0868] Server: The classified news articles are fed into a generative AI model to be summarized for children. Based on the age setting (e.g., elementary school, middle school), a summary text is generated in an easy-to-understand format.

[0869] Step 5:

[0870] Server: Stores summarized news articles and their metadata (topic, target demographic, etc.) in a database. The stored data is linked to other related news articles.

[0871] Step 6:

[0872] Device: When a user (parent or child) opens the application, the device sends a request to the server to retrieve the latest summary news articles.

[0873] Step 7:

[0874] Terminal: The retrieved news summaries are displayed in a user interface in the form of a list, with each news item showing its title and summary text.

[0875] Step 8:

[0876] User: Taps on a displayed news article to view details, then presses a button to mark it as "read."

[0877] Step 9:

[0878] Terminal: Detects user actions and sends the "read" mark data to the server.

[0879] Step 10:

[0880] Server: The received "read" data is saved in a database and managed as the user's browsing history.

[0881] Step 11:

[0882] Server: Based on the related information of the news article, selects news articles on related topics and prepares them for provision to users.

[0883] Step 12:

[0884] On-device: When a user reads a particular article, other related articles are displayed along with it. Related news is displayed in a list format for easy access.

[0885] Step 13:

[0886] User: Can fill out a form to provide feedback on a news article and press the "Submit" button.

[0887] Step 14:

[0888] Terminal: Sends feedback data to the server.

[0889] Step 15:

[0890] Server: The received feedback data is stored in a database and used as reference data for model improvement.

[0891] Step 16:

[0892] User: Press the "Share" button on the news article details screen to share the news on social media.

[0893] Step 17:

[0894] Device: The SNS selection screen is displayed, and the article summary and sharing link are sent to the SNS selected by the user.

[0895] Step 18:

[0896] Server: News articles shared on social media are displayed to other users in a format optimized for the target age group.

[0897] Step 19:

[0898] Server: Analyzes the collected feedback data and tunes the generative AI model. The improved model is used to generate the next news summary.

[0899] Step 20:

[0900] Device: While the user is viewing a news article, the device uses the built-in camera and microphone to analyze the user's facial expressions and tone of voice in real time.

[0901] Step 21:

[0902] Device: The emotion engine generates the user's emotion data and temporarily stores the analysis results on the device.

[0903] Step 22:

[0904] Terminal: Sends emotion data to the server in real time.

[0905] Step 23:

[0906] Server: Analyzes the emotion data sent from the emotion engine and stores it in a database, thereby accumulating a history of users' emotions toward news articles.

[0907] Step 24:

[0908] Server: Based on the accumulated emotional data and the browsing history of news articles, the server selects the next news article to be provided. For example, if the user shows a "happy" expression, the server will provide positive news.

[0909] Examples:

[0910] For example, if a 10-year-old child is using the app to read a new report on climate change, the emotion engine will analyze the child's facial expressions and determine that the user is interested. Related news about the importance of recycling will also be displayed. After finishing the article, the child can press the "read" button, and the emotion data will be stored in the database, and the next appropriate news item will be displayed.

[0911] Example 2

[0912] 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."

[0913] Existing news article delivery systems often do not adequately summarize articles for children, making them difficult to understand. Furthermore, they are unable to provide news that takes into account user feedback and emotions, making it difficult to sustain interest. Furthermore, they do not adequately provide relevant news articles, resulting in low learning outcomes for users.

[0914] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0915] In this invention, the server includes means for collecting news articles, means for preprocessing the collected news articles, means for classifying the preprocessed news articles using a topic classification model, means for inputting the classified news articles into a generative AI model using prompt sentences and summarizing them according to age, and means for providing the summarized news articles to a terminal. This makes it possible to provide news articles in a format that is easy for children to understand. Furthermore, by utilizing user feedback and emotional data, it is possible to provide more appropriate articles for each user and improve learning effectiveness. Furthermore, by displaying related news articles together, it is possible to deepen the user's understanding.

[0916] A "news article" is written information about a current event or topic obtained from online or offline sources.

[0917] "Preprocessing" is the process of performing data preparation tasks such as removing noise data from the collected text data of news articles and normalizing character encoding.

[0918] A "topic classification model" is a machine learning model used to automatically classify news articles into specific categories (e.g., environment, politics, science and technology, etc.).

[0919] A "prompt" is a text sentence that provides the information or instructions needed to input into a generative AI model.

[0920] A "generative AI model" is an artificial intelligence model that uses natural language processing techniques to summarize or generate information from presented input data.

[0921] "Summarization" refers to extracting the key information from an original news article and restating it in a shorter form.

[0922] A "terminal" is a computing device (e.g., smartphone, tablet, PC, etc.) that displays news articles and accepts user operations.

[0923] "User feedback" refers to information such as opinions, requests, and evaluations provided by users regarding news articles.

[0924] "Related news articles" are other news articles that are related in topic or content to a particular article.

[0925] A "knowledge base" is a database that systematically stores news articles and related data, enabling them to be searched and analyzed.

[0926] The present invention relates to a system for providing appropriate news articles to children and enhancing their learning effect. Specific embodiments of the present invention will be described below.

[0927] News article collection and preprocessing

[0928] server:

[0929] The server periodically sends HTTP requests to information sources, such as news APIs, to retrieve the latest news articles. The retrieved news articles are received as JSON-formatted data and stored in a temporary data store. The server then preprocesses the collected news articles using a text analysis library (e.g., NLTK, SpaCy, etc.). Preprocessing includes removing noise data such as advertisements, links, and special characters, and normalizing character encoding.

[0930] News article classification and summarization

[0931] server:

[0932] The preprocessed news articles are fed into a topic classification model (e.g., environment, politics, science and technology) to be classified into specific categories. The classified news articles are then fed into a generative AI model (e.g., GPT-3) with a prompt. The generative AI model summarizes the articles based on the user's specified age.

[0933] Example prompt sentence:

[0934] "Please provide a quick summary of the following article for an 8-year-old: {News article text}"

[0935] The generated summary articles are stored in a database.

[0936] Providing and displaying news summaries

[0937] Device:

[0938] When a user opens the application, the device sends an HTTP request to the server to retrieve the latest summarized news articles, which are then displayed in a user interface (e.g., a list view).

[0939] News read tracking and feedback collection

[0940] User:

[0941] Users can tap on a news article to view details and press the "mark as read" button after finishing reading. They can also fill out a feedback form for the news article and press the "submit" button. The feedback data includes requests and improvements for the article content.

[0942] Device:

[0943] The "read" mark data and feedback data are sent from the terminal to the server, where the data is stored in a database and the user's browsing history and feedback information are managed.

[0944] Displaying related news and managing knowledge base

[0945] server:

[0946] The server categorizes summarized news articles by topic and stores them in a database. When a user views a particular article, other related articles are also displayed. This display of related news can deepen the user's understanding. The server also provides a knowledge base for systematically storing news articles and related data.

[0947] Introducing an emotion engine and selecting the next news article

[0948] Device:

[0949] While a user is browsing a news article, the system uses the built-in camera and microphone to analyze the user's emotions in real time. The emotion engine analyzes the user's facial expressions and tone of voice to generate emotion data.

[0950] server:

[0951] The emotional data sent from the emotion engine is received and analyzed. The analyzed emotional data is stored in a database, and the user's emotional history regarding news articles is accumulated. The server selects the next news article to be provided based on the accumulated emotional data and the news article viewing history. For example, if the user shows a happy expression, positive news will be provided first.

[0952] Specific examples

[0953] As a concrete example, consider a 10-year-old child using the app. While the child is viewing a summary of a "new report on climate change," the emotion engine analyzes the child's facial expressions and determines that the user is interested. Related news articles about the importance of recycling are also displayed, which the child can read. After finishing the article, the child presses the "read" button, and the emotion data is stored in the database. In this way, by summarizing news articles in an easy-to-understand manner and combining it with emotion recognition, the app can provide appropriate news articles for children and improve their learning. It can also promote communication with parents and teachers, encouraging children to become interested in current events and social situations.

[0954] Thus, the present invention is a system that provides users with an innovative news article delivery experience by integrating a series of processes, including news article collection, preprocessing, classification, summarization, delivery, emotion recognition, and display of related news.

[0955] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0956] Step 1:

[0957] News article collection

[0958] The server periodically sends HTTP requests to the news API to retrieve the latest news article data in JSON format, which is then stored in a temporary data store on the server.

[0959] Input: A request to the News API

[0960] Data processing: Obtaining JSON data from the news API

[0961] Output: Retrieved news article data

[0962] Specifically, it sets up a regular scheduled job to access the news API every hour or every day.

[0963] Step 2:

[0964] News article preprocessing

[0965] The server uses a text analysis library (e.g., NLTK, SpaCy, etc.) to preprocess the news article data, removing noise data (e.g., advertisements, links, special characters, etc.) and normalizing character encoding.

[0966] Input: Acquired news article data

[0967] Data processing: Removal of noise data, normalization of character encoding

[0968] Output: Preprocessed news article data

[0969] Specifically, we remove noise data using regular expressions and split words using SpaCy's tokenizer.

[0970] Step 3:

[0971] News article classification

[0972] The server uses a topic classification model to classify the preprocessed news article data into specific categories (e.g., environment, politics, science and technology, etc.).

[0973] Input: Preprocessed news article data

[0974] Data processing: Category classification using topic classification model

[0975] Output: Categorized news article data

[0976] Specifically, the operation involves applying a text classification algorithm using a machine learning model.

[0977] Step 4:

[0978] News article summary generation

[0979] The server inputs the classified news article data into a generative AI model (e.g., GPT-3) and summarizes the article based on the user's specified age. Specific summarization instructions are given to the generative AI model using prompt sentences.

[0980] Input: Categorized news article data, prompt

[0981] Data Computation: Generative AI Models for Summarization

[0982] Output: Summarized news article data

[0983] Example prompt: "Please provide a brief summary of the following article for an 8-year-old: {news article text}"

[0984] Step 5:

[0985] Providing news summaries

[0986] The terminal sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the terminal's user interface.

[0987] Input: Summary news article data on the server

[0988] Data processing: None

[0989] Output: A summary news article displayed on your terminal

[0990] Specifically, news data in JSON format is retrieved via an HTTP request and displayed in a list view.

[0991] Step 6:

[0992] News read status management

[0993] Users tap on a news article to view the details, and when they are finished reading, they press the "read" button. The device then sends the "read" mark data to the server, which then stores it in a database.

[0994] Input: User presses the read button

[0995] Data processing: Saving read information to a database

[0996] Output: Updated user browsing history

[0997] Specifically, the read information is sent to the server in JSON format as a POST request.

[0998] Step 7:

[0999] Feedback collection

[1000] The user fills in the feedback form for the news article and presses the "Submit" button. The terminal sends the feedback data to the server, where it is stored in a database.

[1001] Input: User feedback data

[1002] Data processing: Saving feedback data to a database

[1003] Output: Updated feedback information

[1004] Specifically, the input contents of the feedback form are sent to the server in JSON format as a POST request.

[1005] Step 8:

[1006] View related news

[1007] The server categorizes summarized news articles by topic and stores them in a database. When a user views a particular article, other related articles are also retrieved and displayed.

[1008] Input: Summary news article data

[1009] Data processing: Acquisition of related news article information

[1010] Output: User interface with related news displayed

[1011] Specifically, news articles belonging to the same category are retrieved from the server as "related articles" via a query.

[1012] Step 9:

[1013] Emotion data analysis and storage

[1014] While a user is viewing a news article, their emotions are analyzed in real time using the device's built-in camera and microphone. The server receives the analyzed emotional data and stores it in a database.

[1015] Input: User facial and voice data

[1016] Data processing: Emotion data analysis using an emotion engine

[1017] Output: Emotion data stored in a database

[1018] Specifically, emotions are analyzed using facial recognition technology and voice emotion recognition technology and stored in a database.

[1019] Step 10:

[1020] Selecting the next news story

[1021] The server selects the next news article to serve based on the accumulated emotional data and the browsing history of news articles. For example, if the user shows a happy expression, it will prioritize positive news articles.

[1022] Input: Emotion data, news article browsing history

[1023] Data processing: Selecting the next news article to be served

[1024] Output: Selected news article data

[1025] Specifically, we apply a recommendation algorithm based on sentiment analysis.

[1026] (Application example 2)

[1027] 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."

[1028] Current news article distribution systems have difficulty providing appropriate and easy-to-understand summaries for children. Furthermore, they lack mechanisms for grasping the degree to which children understand a news article or the emotions it evokes. As a result, they are unable to fully stimulate children's interest or enhance their learning. Another issue is that they are unable to provide personalized news article content, resulting in a uniform distribution system.

[1029] 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.

[1030] In this invention, the server includes means for collecting news articles, means for summarizing the collected news articles according to age, means for providing the summarized news articles to the terminal, means for recognizing the user's emotions while the summarized news articles are displayed, and means for analyzing and saving the emotion data. This makes it possible to provide news articles in a format that is interesting and easy for children to understand, and to personalize the next news article to be provided based on the emotion data.

[1031] "News articles" are text data about current events and happenings distributed by news organizations and information providers.

[1032] "Means of collection" refers to the ability to obtain the latest news articles from news APIs and other information sources via the network.

[1033] "Age-appropriate summarization" refers to the function of making collected news articles easier to understand and concise for a specific age group.

[1034] "Means for providing" refers to the function of delivering summarized news articles to user terminals so that they can be viewed.

[1035] "Device" refers to a device such as a smartphone, tablet, or computer that a user uses to view news articles.

[1036] "Means for recognizing emotions" refers to a function that uses a camera or microphone to analyze facial expressions and tone of voice when a user is viewing a news article, in order to determine emotions.

[1037] "Emotional data" refers to digital data that indicates the emotional state of a user analyzed from facial expressions, tone of voice, etc.

[1038] "Means for analyzing and storing" refers to the function for processing the recognized emotional data and storing it in storage such as a database.

[1039] "Personalization" refers to optimizing and providing content and information to suit the interests and concerns of each individual user.

[1040] The present invention relates to a system that summarizes news articles in an easy-to-understand manner for children and combines it with an emotion engine that recognizes the user's emotions. To effectively implement this invention, the server, terminals, and users must work together to execute each step.

[1041] server

[1042] The server collects news articles, preprocesses them, generates summaries, and analyzes and stores sentiment data.

[1043] 1. Periodically collect news articles from a news API and store them in a temporary data store. Here, you can use an existing news API, such as NewsAPI.

[1044] 2. Preprocess the collected news articles using a text analysis library such as TextBlob to remove noise data and normalize character encoding.

[1045] 3. Classify the preprocessed news articles using a topic classification model and generate age-appropriate summaries using a generative AI model (e.g., OpenAI GPT-3), with an example prompt such as "Summarize the news article for a 10-year-old: [insert news article here]."

[1046] 4. The summarized news articles are stored in a database.

[1047] 5. While the user is browsing a news article, the emotion engine analyzes the user's facial expressions and tone of voice through the camera and microphone to generate emotion data.

[1048] 6. Emotional data is analyzed and stored in a database along with browsing history and related news.

[1049] Terminal

[1050] The device displays news articles and collects and transmits emotion data through a user interface.

[1051] 1. When a user opens the application, it sends a request to the server to get the latest summary news articles.

[1052] 2. The retrieved news articles are displayed in a list format on the user interface.

[1053] 3. When a user selects a news article and views its details, the device's built-in camera and microphone are used to analyze the user's emotions in real time and send the data to the emotion engine.

[1054] 4. After the user reads a news article, a button is provided to mark it as "read" and the data is sent to the server.

[1055] User

[1056] Users (mainly children) operate the device to read news articles and cooperate in collecting emotion data.

[1057] 1. Select the news article that interests you from the list of news articles.

[1058] 2. Have the students read news articles and analyze emotional data via a camera or microphone.

[1059] 3. After reading the details of the news article, press the "read" button to mark the article and provide feedback.

[1060] As a concrete example, consider a 10-year-old child viewing a summary article titled "New Report on Climate Change." While the child is reading the article, the emotion engine analyzes the child's facial expressions and determines that the child is interested. It also displays a related news article about "The Importance of Recycling," which the child can continue reading. This emotion information and browsing history are stored in a database and used for future news distribution.

[1061] In this way, by operating this system in cooperation with the server, terminals, and users, news articles can be provided in a format that is easy for children to understand and that will interest them, thereby improving the effectiveness of their learning.

[1062] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1063] Step 1:

[1064] The server collects news articles by periodically sending requests to the news API to retrieve the latest news articles. This request requires an API key and a specific endpoint. The input is the API key and the endpoint, and the output is the retrieved news article. The news article is then stored in a temporary data store.

[1065] Step 2:

[1066] The server preprocesses the collected news articles. Specifically, it uses a text analysis library such as TextBlob to remove noise data from the news articles (such as advertisements, links, and special characters) and normalizes character encoding. The input is the collected news articles, and the output is the preprocessed news articles. These preprocessed news articles are then input into a topic classification model.

[1067] Step 3:

[1068] The server categorizes the preprocessed news articles by topic. It uses a topic classification model to classify the news articles into specific categories (e.g., environment, politics, science and technology). The input is the preprocessed news articles, and the output is the classified news articles. The classified news articles are stored in a database.

[1069] Step 4:

[1070] The server inputs the classified news article into a generative AI model to summarize it for children. The input is the classified news article and the user's (mainly child) age preference. The generative AI model (e.g., OpenAI GPT-3) generates a summary using the prompt "Summarize the news article for a 10-year-old child: [insert news article here]". The output is a summarized news article. This summarized news article is stored in a database.

[1071] Step 5:

[1072] The device displays a list of news articles through a user interface. When a user opens the application, it sends a request to the server to retrieve the latest summarized news articles. The input is the user request, and the output is the list of retrieved news articles. The news articles are displayed in list format on the device screen.

[1073] Step 6:

[1074] The device recognizes emotions when the user selects a news article and views its details. When the user selects a news article, the device's built-in camera and microphone are used to analyze the user's facial expression and tone of voice in real time. The input is the news article selection and the user's facial expression and tone of voice. The output is recognized emotion data, which is sent to the server.

[1075] Step 7:

[1076] The server analyzes the emotion data and stores it in a database. The input is the transmitted emotion data, and the output is the analyzed emotion data. This data is used to select the next news article.

[1077] Step 8:

[1078] The server selects the next news article to be served based on the analyzed emotion data and the browsing history of news articles. The input is emotion data and browsing history, and the output is the next news article to be served. This article will be served the next time the user opens the application.

[1079] 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.

[1080] 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.

[1081] 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.

[1082] [Third embodiment]

[1083] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1084] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[1085] 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).

[1086] 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.

[1087] 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.

[1088] 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).

[1089] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1090] 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.

[1091] 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.

[1092] 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.

[1093] 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.

[1094] 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."

[1095] The present invention relates to a system for collecting news articles, summarizing them in an easy-to-understand manner for children, and providing the summaries to a terminal. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes embodiments of the present invention with specific examples.

[1096] News article collection

[1097] server:

[1098] The server periodically sends requests to a news source (e.g., a news API) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[1099] News article preprocessing and classification

[1100] server:

[1101] The collected news articles are first preprocessed using a text analysis library, which includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The preprocessed news articles are then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[1102] Summary Generation

[1103] server:

[1104] The classified news articles are then fed into a generative AI model, which then summarizes the article in an easy-to-understand way based on the user's (parent's or child's) age preference. For example, for an 8-year-old child, the article might be simplified to something like, "The earth's temperature is rising. We're all working together to stop this."

[1105] Providing news summaries

[1106] Device:

[1107] When a user opens the application, the device sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the device's user interface.

[1108] News read status management

[1109] User:

[1110] Users (usually children) can tap on a displayed news article to view details, then press a button to mark it as "read."

[1111] Device:

[1112] The "read" mark data is sent from the device to the server and stored in a database on the server side, allowing the user's browsing history to be managed.

[1113] View related news

[1114] server:

[1115] To display related news articles in one place, the server categorizes the summarized news articles by topic and stores them in a database, so that when a user reads a particular article, other related articles are also displayed.

[1116] Feedback collection

[1117] User:

[1118] Users can fill out a feedback form for a news article and press the "Submit" button. For example, they can enter a request such as "I'd like this article to be more detailed."

[1119] Device:

[1120] The feedback data is sent from the device to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[1121] SNS sharing function

[1122] User:

[1123] Users can click the "Share" button on the details screen of a news article to share it on social media.

[1124] Device:

[1125] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[1126] server:

[1127] Shared news articles will be displayed to other users in an age-optimized format.

[1128] Model Improvement

[1129] server:

[1130] The collected feedback data is analyzed and the generative AI model is tuned based on the results. The improved model is used to generate the next news summary.

[1131] Specific examples

[1132] As a concrete example, consider a 10-year-old child using an app. When the app is opened, a summary article titled "New Report on Climate Change" ("The temperature of the earth has been rising recently. We are all working hard to prevent this.") is retrieved from the server. The child then reads the article and presses the "Read" button. A related news item, "The Importance of Recycling," is then displayed, which the child can read.

[1133] In this way, the present invention summarizes news articles in an easy-to-understand manner and provides them to children, thereby increasing children's opportunities to come into contact with current events and social situations and promoting communication with parents and teachers.

[1134] The processing flow will be explained below.

[1135] Step 1:

[1136] Server: Sends requests to news sources (e.g., news APIs) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[1137] Step 2:

[1138] Server: The retrieved news articles are preprocessed using a text analysis library, which includes removing noise data (advertisements, links, special characters, etc.) and normalizing character encoding.

[1139] Step 3:

[1140] Server: Preprocessed news articles are input into a topic classification model. The model classifies news articles into specific topics (e.g., environment, politics, science and technology). The classification results are stored in a temporary data store.

[1141] Step 4:

[1142] Server: The classified news articles are fed into a generative AI model to be summarized for children. Based on the age setting (e.g., elementary school, middle school), a summary text is generated in an easy-to-understand format.

[1143] Step 5:

[1144] Server: Stores summarized news articles and their metadata (topic, target demographic, etc.) in a database. The stored data is linked to other related news articles.

[1145] Step 6:

[1146] Device: When a user (parent or child) opens the application, the device sends a request to the server to retrieve the latest summary news articles.

[1147] Step 7:

[1148] Terminal: The retrieved news summaries are displayed in a user interface in the form of a list, with each news item showing its title and summary text.

[1149] Step 8:

[1150] User: Taps on a displayed news article to view details, then presses a button to mark it as "read."

[1151] Step 9:

[1152] Terminal: Detects user actions and sends the "read" mark data to the server.

[1153] Step 10:

[1154] Server: The received "read" data is saved in a database and managed as the user's browsing history.

[1155] Step 11:

[1156] Server: Based on the related information of the news article, selects news articles on related topics and prepares them for provision to users.

[1157] Step 12:

[1158] On-device: When a user reads a particular article, other related articles are displayed along with it. Related news is displayed in a list format for easy access.

[1159] Step 13:

[1160] User: Can fill out a form to provide feedback on a news article and press the "Submit" button.

[1161] Step 14:

[1162] Terminal: Sends feedback data to the server.

[1163] Step 15:

[1164] Server: The received feedback data is stored in a database and used as reference data for model improvement.

[1165] Step 16:

[1166] User: Press the "Share" button on the news article details screen to share the news on social media.

[1167] Step 17:

[1168] Device: The SNS selection screen is displayed, and the article summary and sharing link are sent to the SNS selected by the user.

[1169] Step 18:

[1170] Server: News articles shared on social media are displayed to other users in a format optimized for the target age group.

[1171] Step 19:

[1172] Server: Analyzes the collected feedback data and tunes the generative AI model. The improved model is used to generate the next news summary.

[1173] Example 1

[1174] 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."

[1175] Conventional news article delivery systems do not provide summaries based on specific age settings, making it difficult for users to understand news articles in a way that is appropriate for them, especially children. Furthermore, they lack the ability to manage read statuses and collect feedback when users view articles, making it difficult to provide content that is tailored to the user's usage. Furthermore, they lack the ability to efficiently display related news articles, and do not provide sufficient support for deepening understanding.

[1176] 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.

[1177] In this invention, the server includes means for collecting news articles, means for preprocessing the collected news articles using a text analysis library, means for inputting the preprocessed news articles into a topic classification model and classifying them into specific categories, means for inputting the classified news articles into a generative AI model and summarizing them based on user settings, means for providing the summarized news articles to a terminal, and means for displaying them on a user interface. This enables the provision of easy-to-understand summarized news articles tailored to the user's age. Additionally, the server can mark news articles selected by the user as read, send and store the read data on the server, and collect feedback, enabling the provision of content based on more specific user usage patterns. Furthermore, the automatic classification and display of related news articles enables deeper understanding and information acquisition.

[1178] "News article gathering means" means hardware or software for periodically retrieving current news articles from online news sources.

[1179] A "text analysis library" is a software tool used to preprocess collected news articles, such as removing noise data and normalizing character encoding.

[1180] "Topic classification model" refers to a machine learning model for classifying preprocessed news articles into specific categories.

[1181] A "generative AI model" refers to an artificial intelligence model that takes categorized news articles as input and provides an easy-to-understand summary based on the user's age settings.

[1182] "Summarization method" refers to the process and software that uses a generative AI model to summarize news articles in a format that is easy for users to understand.

[1183] "Means for providing to the terminal" refers to the processes and software for transmitting summarized news articles from the server to the terminal and for the terminal to receive and display them.

[1184] "User Interface" refers to the on-screen interface through which a user can view news articles, provide feedback, and access other features.

[1185] "Means to mark as read" refers to the functionality and processes that allow a user to mark a news article as "read" and save that information.

[1186] "Feedback collection means" refers to the process and software for collecting user-entered feedback on news articles and transmitting and storing that data on a server.

[1187] "Means for categorizing by relevant topic" refers to the process and software for re-categorizing summarized news articles by topic and storing them in a database.

[1188] "Means for displaying related news articles together" refers to processes and software for displaying other articles related to a particular article at the same time as the user reads that article.

[1189] "Means for storing in knowledge base" refers to the functions and processes by which processed news articles and related information are stored in a database for later retrieval.

[1190] The present invention relates to a system for collecting news articles, summarizing them in an easy-to-understand manner for children, and providing them to a terminal. Specific embodiments for carrying out the present invention will be described below.

[1191] News article collection

[1192] server:

[1193] The server uses the requests library to periodically send HTTP requests to news sources (e.g., news APIs) to retrieve new news articles, which are then stored in a temporary data store (e.g., an SQL database) on the server, for example in the form of a pandas dataframe.

[1194] News article preprocessing and classification

[1195] server:

[1196] The collected news articles are first preprocessed using text analysis libraries such as NLTK and spaCy, which include removing noise data (such as advertisements and links) using regular expressions and standardizing character encoding.

[1197] The preprocessed news articles are then fed into a topic classification model (e.g., BERT, GPT-3) to classify them into specific categories (e.g., environment, politics, science and technology).

[1198] Summary Generation

[1199] server:

[1200] The classified news articles are fed into a generative AI model (e.g., OpenAI's GPT-3), which then summarizes the article in an easy-to-understand way based on the user's (parent or child's) age preference. Examples of prompts include:

[1201] "Summarize the following news article in a way that is understandable for a 10-year-old.

[1202] News Article:

[1203] As the Earth's temperature rises, various measures are being taken around the world...

[1204] "

[1205] The generative AI model uses this prompt to summarize the news article as something like, "The temperature of the earth has been rising recently. We are all working hard to prevent this."

[1206] Providing news summaries

[1207] Device:

[1208] When a user opens the application, the device sends an HTTP request to the server to retrieve the latest summarized news articles, which are then displayed in a list format in the device's user interface using RecyclerView and ListView.

[1209] News read status management

[1210] User:

[1211] Users (mainly children) can tap on the displayed news article to view details, and then press the "read" button after viewing.

[1212] Device:

[1213] The "read" mark data is sent from the device to the server by sending an HTTP POST request, and the server stores this data in a database and manages it as the user's browsing history.

[1214] View related news

[1215] server:

[1216] The server recategorizes the summarized news articles by topic and stores them in a database, allowing users to simultaneously view other related articles as they read a particular article.

[1217] Feedback collection

[1218] User:

[1219] Users can enter their feedback on the news article in the form and press the "Submit" button. They can enter requests such as "I wish this article had more details."

[1220] Device:

[1221] The device sends the feedback data to the server, which stores it in a database and uses it to improve the generative AI model.

[1222] SNS sharing function

[1223] User:

[1224] Users can click the "Share" button on the details screen of a news article to share it on social media.

[1225] Device:

[1226] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[1227] Model Improvement

[1228] server:

[1229] The server analyzes the collected feedback data and uses the results to tune the generative AI model, which is then used to generate the next news summary.

[1230] Specific examples

[1231] For example, if a 10-year-old child were to use this system, the process would go something like this:

[1232] 1. When a child opens the app, the device retrieves the latest summary news article (e.g., "The temperature of the earth has been rising recently. We are all working hard to prevent this.") from the server.

[1233] 2. The child reads the article and then presses the "read" button, which saves the read information on the server.

[1234] 3. Additionally, a related news item, "The Importance of Recycling," is displayed, which children can also read.

[1235] Thus, the present invention is a system that provides easy-to-understand summaries of news articles to children, helps them understand current affairs and social situations, and promotes communication with parents and teachers.

[1236] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1237] Step 1:

[1238] News article collection

[1239] The server sends HTTP GET requests to a news source (e.g., a news API). The input is the API endpoint and authentication information, and the server executes this periodically. The output is the JSON data of the retrieved news articles. The server converts this data into a pandas dataframe and stores it in a temporary data store (e.g., a SQL database).

[1240] Step 2:

[1241] News article preprocessing

[1242] The server retrieves news articles from the temporary data store and preprocesses them using a text analysis library (e.g., NLTK, spaCy). The input is the news article data saved in the previous step, and the output is the preprocessed clean text data. Specific operations include removing noise data (advertisements, links, etc.) using regular expressions and standardizing character encoding.

[1243] Step 3:

[1244] Topic Classification

[1245] The server inputs preprocessed news articles into a topic classification model (e.g., BERT, GPT-3). The input is clean text data, and the output is data classified into specific categories (e.g., environment, politics, science and technology). Specifically, the text data is converted into feature vectors, and classification results are obtained using a pre-trained model.

[1246] Step 4:

[1247] Summary Generation

[1248] The server inputs the classified news article into a generative AI model (e.g., OpenAI's GPT-3). The input is the classified news article and a prompt, and the output is a summarized news article based on the user's age preference. Specifically, the prompt is generated in the following format:

[1249] "Summarize the following news article in a way that is understandable for a 10-year-old.

[1250] News Article:

[1251] As the Earth's temperature rises, various measures are being taken around the world...

[1252] "

[1253] The generative AI model uses this information to summarize the news, and the server stores the summarized article in a database.

[1254] Step 5:

[1255] Providing news summaries

[1256] When a user opens the application on a device, the device sends an HTTP GET request to the server to request a summary news article. The input is the user's request, and the output is the latest summary news article data. The server sends this in JSON format to the device, which then displays it using RecyclerView or ListView.

[1257] Step 6:

[1258] News read status management

[1259] Users can tap on a news article to view details and then press the "mark as read" button after viewing. The input is the user's action, and the output is the updated read data. The device sends this data as an HTTP POST request to the server, which stores it in a database.

[1260] Step 7:

[1261] View related news

[1262] The server reclassifies summarized news articles by topic and stores them in a database. The input is the summarized news article data, and the output is the data classified by topic. When a user reads a particular article, the device sends a request to the server to display related news as well. The server sends the corresponding related news in JSON format to the device, and the device displays it.

[1263] Step 8:

[1264] Feedback collection

[1265] The user enters their feedback on the news article into the form and presses the "Submit" button. The input is the user's feedback content, and the output is the feedback data. The terminal sends the feedback data to the server via an HTTP POST request, and the server stores it in a database.

[1266] Step 9:

[1267] SNS sharing function

[1268] The user presses the "Share" button on the news article details screen, which takes them to the SNS sharing selection screen. The input is the user's operation, and the output is a sharing link and article summary. The device uses the SNS API to send the article summary and sharing link.

[1269] Step 10:

[1270] Model Improvement

[1271] The server analyzes the collected feedback data and tunes the generative AI model. The input is the feedback data and the output is the improved model. The server applies the new model to be used in generating the next news summary.

[1272] (Application example 1)

[1273] 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."

[1274] In today's world, it is important for children to have opportunities to read news articles, but the content of news is often too technical and difficult to understand. Therefore, there is a need for an easy-to-understand news summary system that helps children understand and be interested in the news.

[1275] 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.

[1276] In this invention, the server includes means for collecting news content, means for summarizing the collected news content according to age, means for providing the summarized news content to the information terminal, means for using a generative AI model to generate a summary, means for displaying news content selected by a user on the information terminal, means for marking the displayed news content as read, means for collecting user feedback, means for instructing the generative AI model to generate a summary using a prompt sentence, means for classifying the news content by related topic, means for displaying related news content together, means for saving the news content in a knowledge base, and means for providing summarized news for each related topic to the information terminal. This enables children to understand the news in an easy-to-understand manner and continue reading with interest.

[1277] "News content" refers to information and articles about current events and happenings obtained from a variety of sources.

[1278] "Means of collection" refers to the methods and devices used to obtain news content from news APIs and other information services and store it on a server.

[1279] "Age-appropriate summarization methods" are algorithms or systems that simplify collected news content for specific age groups, making it easier to understand.

[1280] The "means for providing to an information terminal" refers to a method or device for transmitting summarized news content from a server to an information terminal (such as a smartphone or tablet) and displaying it.

[1281] A "generative AI model" is a machine learning model that uses artificial intelligence technology to analyze input text and perform tasks such as summarizing and classifying it.

[1282] A "prompt sentence" is input text used to instruct a generative AI model to perform a specific operation or process.

[1283] The "means for displaying news content selected by the user" refers to an interface or method for displaying news content selected by the user on an information terminal.

[1284] A "means for marking as read" is a method or device for marking news content viewed by a user as read.

[1285] A "means for collecting feedback" is a method or system for obtaining opinions and thoughts from users and transmitting them to a server for storage.

[1286] A "topical classification method" is an algorithm or system that categorizes collected news content into specific themes or topics.

[1287] The "means for displaying related news content together" refers to a method or device for simultaneously displaying other news content related to the news that the user is viewing.

[1288] A "means for storing in a knowledge base" is a method or apparatus for storing categorized news content in a database or other storage system for future reference.

[1289] The present invention relates to a system that collects news content, summarizes it for specific age groups, and provides it to information terminals. This system includes functions such as news collection, preprocessing, classification, summary generation, provision, read management, display of related news, feedback collection, summarization using a generative AI model, and instructions using prompt sentences. Specific embodiments for implementing this invention are described below.

[1290] Gathering news content

[1291] The server periodically sends requests to information providers such as news APIs to retrieve the latest news content, which is then stored in a temporary data store.

[1292] News content preprocessing and classification

[1293] The server preprocesses the retrieved news content using a text analysis library. This preprocessing includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The preprocessed news content is then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[1294] Summary Generation

[1295] The server inputs the classified news content into a generative AI model. The generative AI model then summarizes the news content in an easy-to-understand format based on the user's (parent's or child's) age setting. For example, for an 8-year-old child, the summary might be simplified to something like, "The temperature of the earth is rising. We are all working together to stop this."

[1296] Providing news summaries

[1297] When the device opens the application, it sends a request to the server to retrieve the latest news summary, which is then displayed in a list format on the device's user interface.

[1298] News read status management

[1299] Users can tap on the displayed news content to view the details, then press a button to mark it as "read." The device then sends the "read" mark data to the server, which then stores it in a database. This allows the user's browsing history to be managed.

[1300] View related news

[1301] The server categorizes the summarized news content by topic and stores it in a database so that related news content can be displayed all at once. When a user reads a particular news item, other related news content is also displayed.

[1302] Feedback collection

[1303] Users can fill out a feedback form about news content and press the "Submit" button. For example, they can enter a request such as "I would like this article to be more detailed." The device then sends the feedback data to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[1304] SNS sharing function

[1305] Users can share news content on social media by clicking the "Share" button on the news content details screen. The device displays a social media selection screen and sends a news summary and a sharing link to the social media platform selected by the user. The server then displays the shared news content to other users in a format optimized for the target age group.

[1306] Model Improvement

[1307] The server analyzes the collected feedback data and uses the results to tune the generative AI model, which is then used to generate the next news summary.

[1308] Specific examples

[1309] As a concrete example, consider a 10-year-old child using an app. When the app is opened, a summary article titled "New Discoveries in Science and Technology" ("A new star has been discovered, which may reveal more about the secrets of the universe") is retrieved from the server. The child then reads the article and presses the "Read" button. Related news items, such as "Advances in Science and Technology," are then displayed, which the child can read. The child can also share the article they have read on social media and send feedback. An example of this prompt is as follows:

[1310] Example of input prompt for generative AI model

[1311] Send a request to the API to get the latest news articles and summarize them for kids aged 8-12.

[1312] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1313] Step 1:

[1314] The server periodically sends requests to the news API to retrieve the latest news content. The retrieved news content is stored in a temporary data store. The input is the news content from the news API, and the output is the data stored in the temporary data store. Specifically, it sends an HTTP request, parses the response, and stores it in a database.

[1315] Step 2:

[1316] The server preprocesses the news content retrieved from the temporary data store. Preprocessing includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The input is the news content from the temporary data store, and the output is the preprocessed news content. Specifically, it uses a text analysis library to remove unnecessary information and standardize the format.

[1317] Step 3:

[1318] The server inputs the preprocessed news content into a topic classification model and classifies it into specific categories (e.g., environment, politics, science and technology). The input is the preprocessed news content, and the output is the news content classified by category. The specific operation is to use a machine learning model to determine which category each news content belongs to.

[1319] Step 4:

[1320] The server inputs the classified news content into a generative AI model and summarizes the news content based on the age setting. The input is news content classified by category, and the output is summarized news content appropriate for the age. A prompt sentence is used to instruct the generative AI model to generate a summary. The specific operation is to input an appropriate prompt sentence to the generative AI model and retrieve the generated summary.

[1321] Step 5:

[1322] When the application is opened, the device sends a request to the server to retrieve the latest news summary. The input is the user request, and the output is the news summary for display. The specific operation is to send an HTTP request to retrieve data from the server and display it on the user interface.

[1323] Step 6:

[1324] The user can tap on the displayed news content to view details and press a button to mark it as "read." The input is the user's operation, and the output is the news content marked as read. The specific operation is to tap on the user interface and then press the "read" button.

[1325] Step 7:

[1326] The terminal sends the "read" mark data to the server, and the server stores it in a database. The input is the read mark data, and the output is the data stored in the database. The specific operation is to send the data to the server by sending an HTTP request, and the server stores the read information in the database.

[1327] Step 8:

[1328] The server categorizes the summarized news content by topic and stores it in a database in order to display related news content in one place. The input is the summarized news content, and the output is the news content categorized by topic. The specific operation is to use a classification algorithm to extract related news and store it in the database.

[1329] Step 9:

[1330] The user fills in the feedback form for the news content and presses the "Send" button. The input is the user's feedback, and the output is feedback data from the terminal. The specific operation is to fill in the feedback form on the user interface and press the send button.

[1331] Step 10:

[1332] The terminal sends feedback data to the server, which stores it in a database. The input is the feedback data, and the output is the data stored in the database. The specific operation is to send an HTTP request to the server to send data, and then store the feedback information on the server.

[1333] Step 11:

[1334] Users can share news content on social media by pressing the "Share" button on the news content details screen. The input is the user's share operation, and the output is the news content shared on the social media. The specific operation is that after pressing the share button, an SNS selection screen is displayed, and the news summary and sharing link are sent to the selected SNS.

[1335] Step 12:

[1336] The server analyzes the collected feedback data and tunes the generative AI model based on the results. The input is the feedback data, and the output is the tuned generative AI model. Specifically, the server analyzes the feedback data and reflects the analysis results in the parameters of the generative AI model.

[1337] 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.

[1338] The present invention relates to a system that collects news articles, summarizes them in an easy-to-understand manner for children, and combines them with an emotion engine that recognizes the user's emotions. Hereinafter, embodiments of the present invention will be described with reference to specific examples.

[1339] News article collection

[1340] server:

[1341] The server periodically sends requests to a news source (e.g., a news API) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[1342] News article preprocessing and classification

[1343] server:

[1344] The collected news articles are first preprocessed using a text analysis library, which includes removing noise data (such as advertisements, links, and special characters) and normalizing character encoding. The preprocessed news articles are then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[1345] Summary Generation

[1346] server:

[1347] The classified news articles are then fed into a generative AI model, which then summarizes the article in an easy-to-understand way based on the user's (parent's or child's) age preference. For example, for an 8-year-old child, the article might be simplified to something like, "The earth's temperature is rising. We're all working together to stop this."

[1348] Providing and displaying news summaries

[1349] Device:

[1350] When a user opens the application, the device sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the device's user interface.

[1351] News read status management

[1352] User:

[1353] Users (usually children) tap on a displayed news article to view details, then press a button to mark it as "read."

[1354] Device:

[1355] The "read" mark data is sent from the device to the server and stored in a database on the server side, allowing the user's browsing history to be managed.

[1356] View related news

[1357] server:

[1358] To display related news articles in one place, the server categorizes the summarized news articles by topic and stores them in a database, so that when a user reads a particular article, other related articles are also displayed.

[1359] Feedback collection

[1360] User:

[1361] Users can fill out a feedback form for a news article and press the "Submit" button. For example, they can enter a request such as "I'd like this article to be more detailed."

[1362] Device:

[1363] The feedback data is sent from the device to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[1364] SNS sharing function

[1365] User:

[1366] Users can click the "Share" button on the details screen of a news article to share it on social media.

[1367] Device:

[1368] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[1369] server:

[1370] Shared news articles will be displayed to other users in an age-optimized format.

[1371] Model Improvement

[1372] server:

[1373] The collected feedback data is analyzed and the generative AI model is tuned based on the results. The improved model is used to generate the next news summary.

[1374] Introducing the Emotion Engine

[1375] emotion recognition

[1376] Device:

[1377] While a user is browsing a news article, the device uses a built-in camera and microphone to analyze the user's emotions in real time. The emotion engine analyzes the user's facial expressions and tone of voice to generate emotion data.

[1378] Emotion data analysis and storage

[1379] server:

[1380] The emotional data sent from the emotion engine is analyzed and stored in a database, thereby accumulating a history of the user's emotions regarding news articles.

[1381] Selecting the next news story

[1382] server:

[1383] The next news article to be served is selected based on the accumulated emotional data and the browsing history of news articles. For example, if the user shows a "happy" expression, positive news will be served.

[1384] Specific examples

[1385] As a concrete example, consider a 10-year-old child using the app. While the child is reading a summary article titled "New Report on Climate Change," the emotion engine analyzes the child's facial expressions and determines that the user is interested. A related news article about "The Importance of Recycling" is also displayed, which the child can read. After finishing the article, the child presses the "Read" button, and the emotion data is saved in the database.

[1386] In this way, by summarizing news articles in an easy-to-understand manner and combining it with emotion recognition, the present invention can provide appropriate news articles for children, enhance learning effectiveness, promote communication with parents and teachers, and encourage children to become interested in current events and social situations.

[1387] The processing flow will be explained below.

[1388] Step 1:

[1389] Server: Sends requests to news sources (e.g., news APIs) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[1390] Step 2:

[1391] Server: The retrieved news articles are preprocessed using a text analysis library, which includes removing noise data (advertisements, links, special characters, etc.) and normalizing character encoding.

[1392] Step 3:

[1393] Server: Preprocessed news articles are input into a topic classification model. The model classifies news articles into specific topics (e.g., environment, politics, science and technology). The classification results are stored in a temporary data store.

[1394] Step 4:

[1395] Server: The classified news articles are fed into a generative AI model to be summarized for children. Based on the age setting (e.g., elementary school, middle school), a summary text is generated in an easy-to-understand format.

[1396] Step 5:

[1397] Server: Stores summarized news articles and their metadata (topic, target demographic, etc.) in a database. The stored data is linked to other related news articles.

[1398] Step 6:

[1399] Device: When a user (parent or child) opens the application, the device sends a request to the server to retrieve the latest summary news articles.

[1400] Step 7:

[1401] Terminal: The retrieved news summaries are displayed in a user interface in the form of a list, with each news item showing its title and summary text.

[1402] Step 8:

[1403] User: Taps on a displayed news article to view details, then presses a button to mark it as "read."

[1404] Step 9:

[1405] Terminal: Detects user actions and sends the "read" mark data to the server.

[1406] Step 10:

[1407] Server: The received "read" data is saved in a database and managed as the user's browsing history.

[1408] Step 11:

[1409] Server: Based on the related information of the news article, selects news articles on related topics and prepares them for provision to users.

[1410] Step 12:

[1411] On-device: When a user reads a particular article, other related articles are displayed along with it. Related news is displayed in a list format for easy access.

[1412] Step 13:

[1413] User: Can fill out a form to provide feedback on a news article and press the "Submit" button.

[1414] Step 14:

[1415] Terminal: Sends feedback data to the server.

[1416] Step 15:

[1417] Server: The received feedback data is stored in a database and used as reference data for model improvement.

[1418] Step 16:

[1419] User: Press the "Share" button on the news article details screen to share the news on social media.

[1420] Step 17:

[1421] Device: The SNS selection screen is displayed, and the article summary and sharing link are sent to the SNS selected by the user.

[1422] Step 18:

[1423] Server: News articles shared on social media are displayed to other users in a format optimized for the target age group.

[1424] Step 19:

[1425] Server: Analyzes the collected feedback data and tunes the generative AI model. The improved model is used to generate the next news summary.

[1426] Step 20:

[1427] Device: While the user is viewing a news article, the device uses the built-in camera and microphone to analyze the user's facial expressions and tone of voice in real time.

[1428] Step 21:

[1429] Device: The emotion engine generates the user's emotion data and temporarily stores the analysis results on the device.

[1430] Step 22:

[1431] Terminal: Sends emotion data to the server in real time.

[1432] Step 23:

[1433] Server: Analyzes the emotion data sent from the emotion engine and stores it in a database, thereby accumulating a history of users' emotions toward news articles.

[1434] Step 24:

[1435] Server: Based on the accumulated emotional data and the browsing history of news articles, the server selects the next news article to be provided. For example, if the user shows a "happy" expression, the server will provide positive news.

[1436] Examples:

[1437] For example, if a 10-year-old child is using the app to read a new report on climate change, the emotion engine will analyze the child's facial expressions and determine that the user is interested. Related news about the importance of recycling will also be displayed. After finishing the article, the child can press the "read" button, and the emotion data will be stored in the database, and the next appropriate news item will be displayed.

[1438] Example 2

[1439] 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."

[1440] Existing news article delivery systems often do not adequately summarize articles for children, making them difficult to understand. Furthermore, they are unable to provide news that takes into account user feedback and emotions, making it difficult to sustain interest. Furthermore, they do not adequately provide relevant news articles, resulting in low learning outcomes for users.

[1441] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1442] In this invention, the server includes means for collecting news articles, means for preprocessing the collected news articles, means for classifying the preprocessed news articles using a topic classification model, means for inputting the classified news articles into a generative AI model using prompt sentences and summarizing them according to age, and means for providing the summarized news articles to a terminal. This makes it possible to provide news articles in a format that is easy for children to understand. Furthermore, by utilizing user feedback and emotional data, it is possible to provide more appropriate articles for each user and improve learning effectiveness. Furthermore, by displaying related news articles together, it is possible to deepen the user's understanding.

[1443] A "news article" is written information about a current event or topic obtained from online or offline sources.

[1444] "Preprocessing" is the process of performing data preparation tasks such as removing noise data from the collected text data of news articles and normalizing character encoding.

[1445] A "topic classification model" is a machine learning model used to automatically classify news articles into specific categories (e.g., environment, politics, science and technology, etc.).

[1446] A "prompt" is a text sentence that provides the information or instructions needed to input into a generative AI model.

[1447] A "generative AI model" is an artificial intelligence model that uses natural language processing techniques to summarize or generate information from presented input data.

[1448] "Summarization" refers to extracting the key information from an original news article and restating it in a shorter form.

[1449] A "terminal" is a computing device (e.g., smartphone, tablet, PC, etc.) that displays news articles and accepts user operations.

[1450] "User feedback" refers to information such as opinions, requests, and evaluations provided by users regarding news articles.

[1451] "Related news articles" are other news articles that are related in topic or content to a particular article.

[1452] A "knowledge base" is a database that systematically stores news articles and related data, enabling them to be searched and analyzed.

[1453] The present invention relates to a system for providing appropriate news articles to children and enhancing their learning effect. Specific embodiments of the present invention will be described below.

[1454] News article collection and preprocessing

[1455] server:

[1456] The server periodically sends HTTP requests to information sources, such as news APIs, to retrieve the latest news articles. The retrieved news articles are received as JSON-formatted data and stored in a temporary data store. The server then preprocesses the collected news articles using a text analysis library (e.g., NLTK, SpaCy, etc.). Preprocessing includes removing noise data such as advertisements, links, and special characters, and normalizing character encoding.

[1457] News article classification and summarization

[1458] server:

[1459] The preprocessed news articles are fed into a topic classification model (e.g., environment, politics, science and technology) to be classified into specific categories. The classified news articles are then fed into a generative AI model (e.g., GPT-3) with a prompt. The generative AI model summarizes the articles based on the user's specified age.

[1460] Example prompt sentence:

[1461] "Please provide a quick summary of the following article for an 8-year-old: {News article text}"

[1462] The generated summary articles are stored in a database.

[1463] Providing and displaying news summaries

[1464] Device:

[1465] When a user opens the application, the device sends an HTTP request to the server to retrieve the latest summarized news articles, which are then displayed in a user interface (e.g., a list view).

[1466] News read tracking and feedback collection

[1467] User:

[1468] Users can tap on a news article to view details and press the "mark as read" button after finishing reading. They can also fill out a feedback form for the news article and press the "submit" button. The feedback data includes requests and improvements for the article content.

[1469] Device:

[1470] The "read" mark data and feedback data are sent from the terminal to the server, where the data is stored in a database and the user's browsing history and feedback information are managed.

[1471] Displaying related news and managing knowledge base

[1472] server:

[1473] The server categorizes summarized news articles by topic and stores them in a database. When a user views a particular article, other related articles are also displayed. This display of related news can deepen the user's understanding. The server also provides a knowledge base for systematically storing news articles and related data.

[1474] Introducing an emotion engine and selecting the next news article

[1475] Device:

[1476] While a user is browsing a news article, the system uses the built-in camera and microphone to analyze the user's emotions in real time. The emotion engine analyzes the user's facial expressions and tone of voice to generate emotion data.

[1477] server:

[1478] The emotional data sent from the emotion engine is received and analyzed. The analyzed emotional data is stored in a database, and the user's emotional history regarding news articles is accumulated. The server selects the next news article to be provided based on the accumulated emotional data and the news article viewing history. For example, if the user shows a happy expression, positive news will be provided first.

[1479] Specific examples

[1480] As a concrete example, consider a 10-year-old child using the app. While the child is viewing a summary of a "new report on climate change," the emotion engine analyzes the child's facial expressions and determines that the user is interested. Related news articles about the importance of recycling are also displayed, which the child can read. After finishing the article, the child presses the "read" button, and the emotion data is stored in the database. In this way, by summarizing news articles in an easy-to-understand manner and combining it with emotion recognition, the app can provide appropriate news articles for children and improve their learning. It can also promote communication with parents and teachers, encouraging children to become interested in current events and social situations.

[1481] Thus, the present invention is a system that provides users with an innovative news article delivery experience by integrating a series of processes, including news article collection, preprocessing, classification, summarization, delivery, emotion recognition, and display of related news.

[1482] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1483] Step 1:

[1484] News article collection

[1485] The server periodically sends HTTP requests to the news API to retrieve the latest news article data in JSON format, which is then stored in a temporary data store on the server.

[1486] Input: A request to the News API

[1487] Data processing: Obtaining JSON data from the news API

[1488] Output: Retrieved news article data

[1489] Specifically, it sets up a regular scheduled job to access the news API every hour or every day.

[1490] Step 2:

[1491] News article preprocessing

[1492] The server uses a text analysis library (e.g., NLTK, SpaCy, etc.) to preprocess the news article data, removing noise data (e.g., advertisements, links, special characters, etc.) and normalizing character encoding.

[1493] Input: Acquired news article data

[1494] Data processing: Removal of noise data, normalization of character encoding

[1495] Output: Preprocessed news article data

[1496] Specifically, we remove noise data using regular expressions and split words using SpaCy's tokenizer.

[1497] Step 3:

[1498] News article classification

[1499] The server uses a topic classification model to classify the preprocessed news article data into specific categories (e.g., environment, politics, science and technology, etc.).

[1500] Input: Preprocessed news article data

[1501] Data processing: Category classification using topic classification model

[1502] Output: Categorized news article data

[1503] Specifically, the operation involves applying a text classification algorithm using a machine learning model.

[1504] Step 4:

[1505] News article summary generation

[1506] The server inputs the classified news article data into a generative AI model (e.g., GPT-3) and summarizes the article based on the user's specified age. Specific summarization instructions are given to the generative AI model using prompt sentences.

[1507] Input: Categorized news article data, prompt

[1508] Data Computation: Generative AI Models for Summarization

[1509] Output: Summarized news article data

[1510] Example prompt: "Please provide a brief summary of the following article for an 8-year-old: {news article text}"

[1511] Step 5:

[1512] Providing news summaries

[1513] The terminal sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the terminal's user interface.

[1514] Input: Summary news article data on the server

[1515] Data processing: None

[1516] Output: A summary news article displayed on your terminal

[1517] Specifically, news data in JSON format is retrieved via an HTTP request and displayed in a list view.

[1518] Step 6:

[1519] News read status management

[1520] Users tap on a news article to view the details, and when they are finished reading, they press the "read" button. The device sends the "read" mark data to the server, which stores it in a database.

[1521] Input: User presses the read button

[1522] Data processing: Saving read information to a database

[1523] Output: Updated user browsing history

[1524] Specifically, the read information is sent to the server in JSON format as a POST request.

[1525] Step 7:

[1526] Feedback collection

[1527] The user fills in the feedback form for the news article and presses the "Submit" button. The terminal sends the feedback data to the server, where it is stored in a database.

[1528] Input: User feedback data

[1529] Data processing: Saving feedback data to a database

[1530] Output: Updated feedback information

[1531] Specifically, the input contents of the feedback form are sent to the server in JSON format as a POST request.

[1532] Step 8:

[1533] View related news

[1534] The server categorizes summarized news articles by topic and stores them in a database. When a user views a particular article, other related articles are also retrieved and displayed.

[1535] Input: Summary news article data

[1536] Data processing: Acquisition of related news article information

[1537] Output: User interface with related news displayed

[1538] Specifically, news articles belonging to the same category are retrieved from the server as "related articles" via a query.

[1539] Step 9:

[1540] Emotion data analysis and storage

[1541] While a user is viewing a news article, their emotions are analyzed in real time using the device's built-in camera and microphone. The server receives the analyzed emotional data and stores it in a database.

[1542] Input: User facial and voice data

[1543] Data processing: Emotion data analysis using an emotion engine

[1544] Output: Emotion data stored in a database

[1545] Specifically, emotions are analyzed using facial recognition technology and voice emotion recognition technology and stored in a database.

[1546] Step 10:

[1547] Selecting the next news story

[1548] The server selects the next news article to serve based on the accumulated emotional data and the browsing history of news articles. For example, if the user shows a happy expression, it will prioritize positive news articles.

[1549] Input: Emotion data, news article browsing history

[1550] Data processing: Selecting the next news article to be served

[1551] Output: Selected news article data

[1552] Specifically, we apply a recommendation algorithm based on sentiment analysis.

[1553] (Application example 2)

[1554] 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."

[1555] Current news article distribution systems have difficulty providing appropriate and easy-to-understand summaries for children. Furthermore, they lack mechanisms for grasping the degree to which children understand a news article or the emotions it evokes. As a result, they are unable to fully stimulate children's interest or enhance their learning. Another issue is that they are unable to provide personalized news article content, resulting in a uniform distribution system.

[1556] 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.

[1557] In this invention, the server includes means for collecting news articles, means for summarizing the collected news articles according to age, means for providing the summarized news articles to the terminal, means for recognizing the user's emotions while the summarized news articles are displayed, and means for analyzing and saving the emotion data. This makes it possible to provide news articles in a format that is interesting and easy for children to understand, and to personalize the next news article to be provided based on the emotion data.

[1558] "News articles" are text data about current events and happenings distributed by news organizations and information providers.

[1559] "Means of collection" refers to the ability to obtain the latest news articles from news APIs and other information sources via the network.

[1560] "Age-appropriate summarization" refers to the function of making collected news articles easier to understand and concise for a specific age group.

[1561] "Means for providing" refers to the function of delivering summarized news articles to user terminals so that they can be viewed.

[1562] "Device" refers to a device such as a smartphone, tablet, or computer that a user uses to view news articles.

[1563] "Means for recognizing emotions" refers to a function that uses a camera or microphone to analyze facial expressions and tone of voice when a user is viewing a news article, in order to determine emotions.

[1564] "Emotional data" refers to digital data that indicates the emotional state of a user analyzed from facial expressions, tone of voice, etc.

[1565] "Means for analyzing and storing" refers to the function for processing the recognized emotional data and storing it in storage such as a database.

[1566] "Personalization" refers to optimizing and providing content and information to suit the interests and concerns of each individual user.

[1567] The present invention relates to a system that summarizes news articles in an easy-to-understand manner for children and combines it with an emotion engine that recognizes the user's emotions. To effectively implement this invention, the server, terminals, and users must work together to execute each step.

[1568] server

[1569] The server collects news articles, preprocesses them, generates summaries, and analyzes and stores sentiment data.

[1570] 1. Periodically collect news articles from a news API and store them in a temporary data store. Here, you can use an existing news API, such as NewsAPI.

[1571] 2. Preprocess the collected news articles using a text analysis library such as TextBlob to remove noise data and normalize character encoding.

[1572] 3. Classify the preprocessed news articles using a topic classification model and generate age-appropriate summaries using a generative AI model (e.g., OpenAI GPT-3), with an example prompt such as "Summarize the news article for a 10-year-old: [insert news article here]."

[1573] 4. The summarized news articles are stored in a database.

[1574] 5. While the user is browsing a news article, the emotion engine analyzes the user's facial expressions and tone of voice through the camera and microphone to generate emotion data.

[1575] 6. Emotional data is analyzed and stored in a database along with browsing history and related news.

[1576] Terminal

[1577] The device displays news articles and collects and transmits emotion data through a user interface.

[1578] 1. When a user opens the application, it sends a request to the server to get the latest summary news articles.

[1579] 2. The retrieved news articles are displayed in a list format on the user interface.

[1580] 3. When a user selects a news article and views its details, the device's built-in camera and microphone are used to analyze the user's emotions in real time and send the data to the emotion engine.

[1581] 4. After the user reads a news article, a button is provided to mark it as "read" and the data is sent to the server.

[1582] User

[1583] Users (mainly children) operate the device to read news articles and cooperate in collecting emotion data.

[1584] 1. Select the news article that interests you from the list of news articles.

[1585] 2. Have the students read news articles and analyze emotional data via a camera or microphone.

[1586] 3. After reading the details of the news article, press the "read" button to mark the article and provide feedback.

[1587] As a concrete example, consider a 10-year-old child viewing a summary article titled "New Report on Climate Change." While the child is reading the article, the emotion engine analyzes the child's facial expressions and determines that the child is interested. It also displays a related news article about "The Importance of Recycling," which the child can continue reading. This emotion information and browsing history are stored in a database and used for future news distribution.

[1588] In this way, by operating this system in cooperation with the server, terminals, and users, news articles can be provided in a format that is easy for children to understand and that will interest them, thereby improving the effectiveness of their learning.

[1589] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1590] Step 1:

[1591] The server collects news articles by periodically sending requests to the news API to retrieve the latest news articles. This request requires an API key and a specific endpoint. The input is the API key and the endpoint, and the output is the retrieved news article. The news article is then stored in a temporary data store.

[1592] Step 2:

[1593] The server preprocesses the collected news articles. Specifically, it uses a text analysis library such as TextBlob to remove noise data from the news articles (such as advertisements, links, and special characters) and normalizes character encoding. The input is the collected news articles, and the output is the preprocessed news articles. These preprocessed news articles are then input into a topic classification model.

[1594] Step 3:

[1595] The server categorizes the preprocessed news articles by topic. It uses a topic classification model to classify the news articles into specific categories (e.g., environment, politics, science and technology). The input is the preprocessed news articles, and the output is the classified news articles. The classified news articles are stored in a database.

[1596] Step 4:

[1597] The server inputs the classified news article into a generative AI model to summarize it for children. The input is the classified news article and the user's (mainly child) age preference. The generative AI model (e.g., OpenAI GPT-3) generates a summary using the prompt "Summarize the news article for a 10-year-old child: [insert news article here]". The output is a summarized news article. This summarized news article is stored in a database.

[1598] Step 5:

[1599] The device displays a list of news articles through a user interface. When a user opens the application, it sends a request to the server to retrieve the latest summarized news articles. The input is the user request, and the output is the list of retrieved news articles. The news articles are displayed in list format on the device screen.

[1600] Step 6:

[1601] The device recognizes emotions when the user selects a news article and views its details. When the user selects a news article, the device's built-in camera and microphone are used to analyze the user's facial expression and tone of voice in real time. The input is the news article selection and the user's facial expression and tone of voice. The output is recognized emotion data, which is sent to the server.

[1602] Step 7:

[1603] The server analyzes the emotion data and stores it in a database. The input is the transmitted emotion data, and the output is the analyzed emotion data. This data is used to select the next news article.

[1604] Step 8:

[1605] The server selects the next news article to be served based on the analyzed emotion data and the browsing history of news articles. The input is emotion data and browsing history, and the output is the next news article to be served. This article will be served the next time the user opens the application.

[1606] 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.

[1607] 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.

[1608] 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.

[1609] [Fourth embodiment]

[1610] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1611] 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.

[1612] 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).

[1613] 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.

[1614] 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.

[1615] 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).

[1616] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1617] 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.

[1618] 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.

[1619] 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.

[1620] 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.

[1621] 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.

[1622] 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."

[1623] The present invention relates to a system for collecting news articles, summarizing them in an easy-to-understand manner for children, and providing the summaries to a terminal. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes embodiments of the present invention with specific examples.

[1624] News article collection

[1625] server:

[1626] The server periodically sends requests to a news source (e.g., a news API) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[1627] News article preprocessing and classification

[1628] server:

[1629] The collected news articles are first preprocessed using a text analysis library, which includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The preprocessed news articles are then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[1630] Summary Generation

[1631] server:

[1632] The classified news articles are then fed into a generative AI model, which then summarizes the article in an easy-to-understand way based on the user's (parent's or child's) age preference. For example, for an 8-year-old child, the article might be simplified to something like, "The earth's temperature is rising. We're all working together to stop this."

[1633] Providing news summaries

[1634] Device:

[1635] When a user opens the application, the device sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the device's user interface.

[1636] News read status management

[1637] User:

[1638] Users (usually children) can tap on a displayed news article to view details, then press a button to mark it as "read."

[1639] Device:

[1640] The "read" mark data is sent from the device to the server and stored in a database on the server side, allowing the user's browsing history to be managed.

[1641] View related news

[1642] server:

[1643] To display related news articles in one place, the server categorizes the summarized news articles by topic and stores them in a database, so that when a user reads a particular article, other related articles are also displayed.

[1644] Feedback collection

[1645] User:

[1646] Users can fill out a feedback form for a news article and press the "Submit" button. For example, they can enter a request such as "I'd like this article to be more detailed."

[1647] Device:

[1648] The feedback data is sent from the device to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[1649] SNS sharing function

[1650] User:

[1651] Users can click the "Share" button on the details screen of a news article to share it on social media.

[1652] Device:

[1653] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[1654] server:

[1655] Shared news articles will be displayed to other users in an age-optimized format.

[1656] Model Improvement

[1657] server:

[1658] The collected feedback data is analyzed and the generative AI model is tuned based on the results. The improved model is used to generate the next news summary.

[1659] Specific examples

[1660] As a concrete example, consider a 10-year-old child using an app. When the app is opened, a summary article titled "New Report on Climate Change" ("The temperature of the earth has been rising recently. We are all working hard to prevent this.") is retrieved from the server. The child then reads the article and presses the "Read" button. A related news item, "The Importance of Recycling," is then displayed, which the child can read.

[1661] In this way, the present invention summarizes news articles in an easy-to-understand manner and provides them to children, thereby increasing children's opportunities to come into contact with current events and social situations and promoting communication with parents and teachers.

[1662] The processing flow will be explained below.

[1663] Step 1:

[1664] Server: Sends requests to news sources (e.g., news APIs) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[1665] Step 2:

[1666] Server: The retrieved news articles are preprocessed using a text analysis library, which includes removing noise data (advertisements, links, special characters, etc.) and normalizing character encoding.

[1667] Step 3:

[1668] Server: Preprocessed news articles are input into a topic classification model. The model classifies news articles into specific topics (e.g., environment, politics, science and technology). The classification results are stored in a temporary data store.

[1669] Step 4:

[1670] Server: The classified news articles are fed into a generative AI model to be summarized for children. Based on the age setting (e.g., elementary school, middle school), a summary text is generated in an easy-to-understand format.

[1671] Step 5:

[1672] Server: Stores summarized news articles and their metadata (topic, target demographic, etc.) in a database. The stored data is linked to other related news articles.

[1673] Step 6:

[1674] Device: When a user (parent or child) opens the application, the device sends a request to the server to retrieve the latest summary news articles.

[1675] Step 7:

[1676] Terminal: The retrieved news summaries are displayed in a user interface in the form of a list, with each news item showing its title and summary text.

[1677] Step 8:

[1678] User: Taps on a displayed news article to view details, then presses a button to mark it as "read."

[1679] Step 9:

[1680] Terminal: Detects user actions and sends the "read" mark data to the server.

[1681] Step 10:

[1682] Server: The received "read" data is saved in a database and managed as the user's browsing history.

[1683] Step 11:

[1684] Server: Based on the related information of the news article, selects news articles on related topics and prepares them for provision to users.

[1685] Step 12:

[1686] On-device: When a user reads a particular article, other related articles are displayed along with it. Related news is displayed in a list format for easy access.

[1687] Step 13:

[1688] User: Can fill out a form to provide feedback on a news article and press the "Submit" button.

[1689] Step 14:

[1690] Terminal: Sends feedback data to the server.

[1691] Step 15:

[1692] Server: The received feedback data is stored in a database and used as reference data for model improvement.

[1693] Step 16:

[1694] User: Press the "Share" button on the news article details screen to share the news on social media.

[1695] Step 17:

[1696] Device: The SNS selection screen is displayed, and the article summary and sharing link are sent to the SNS selected by the user.

[1697] Step 18:

[1698] Server: News articles shared on social media are displayed to other users in a format optimized for the target age group.

[1699] Step 19:

[1700] Server: Analyzes the collected feedback data and tunes the generative AI model. The improved model is used to generate the next news summary.

[1701] Example 1

[1702] 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."

[1703] Conventional news article delivery systems do not provide summaries based on specific age settings, making it difficult for users to understand news articles in a way that is appropriate for them, especially children. Furthermore, they lack the ability to manage read statuses and collect feedback when users view articles, making it difficult to provide content that is tailored to the user's usage. Furthermore, they lack the ability to efficiently display related news articles, and do not provide sufficient support for deepening understanding.

[1704] 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.

[1705] In this invention, the server includes means for collecting news articles, means for preprocessing the collected news articles using a text analysis library, means for inputting the preprocessed news articles into a topic classification model and classifying them into specific categories, means for inputting the classified news articles into a generative AI model and summarizing them based on user settings, means for providing the summarized news articles to a terminal, and means for displaying them on a user interface. This enables the provision of easy-to-understand summarized news articles tailored to the user's age. Additionally, the server can mark news articles selected by the user as read, send and store the read data on the server, and collect feedback, enabling the provision of content based on more specific user usage patterns. Furthermore, the automatic classification and display of related news articles enables deeper understanding and information acquisition.

[1706] "News article gathering means" means hardware or software for periodically retrieving current news articles from online news sources.

[1707] A "text analysis library" is a software tool used to preprocess collected news articles, such as removing noise data and normalizing character encoding.

[1708] "Topic classification model" refers to a machine learning model for classifying preprocessed news articles into specific categories.

[1709] A "generative AI model" refers to an artificial intelligence model that takes categorized news articles as input and provides an easy-to-understand summary based on the user's age settings.

[1710] "Summarization method" refers to the process and software that uses a generative AI model to summarize news articles in a format that is easy for users to understand.

[1711] "Means for providing to the terminal" refers to the processes and software for transmitting summarized news articles from the server to the terminal and for the terminal to receive and display them.

[1712] "User Interface" refers to the on-screen interface through which a user can view news articles, provide feedback, and access other features.

[1713] "Means to mark as read" refers to the functionality and processes that allow a user to mark a news article as "read" and save that information.

[1714] "Feedback collection means" refers to the process and software for collecting user-entered feedback on news articles and transmitting and storing that data on a server.

[1715] "Means for categorizing by relevant topic" refers to the process and software for re-categorizing summarized news articles by topic and storing them in a database.

[1716] "Means for displaying related news articles together" refers to processes and software for displaying other articles related to a particular article at the same time as the user reads that article.

[1717] "Means for storing in knowledge base" refers to the functions and processes by which processed news articles and related information are stored in a database for later retrieval.

[1718] The present invention relates to a system for collecting news articles, summarizing them in an easy-to-understand manner for children, and providing them to a terminal. Specific embodiments for carrying out the present invention will be described below.

[1719] News article collection

[1720] server:

[1721] The server uses the requests library to periodically send HTTP requests to news sources (e.g., news APIs) to retrieve new news articles, which are then stored in a temporary data store (e.g., an SQL database) on the server, for example in the form of a pandas dataframe.

[1722] News article preprocessing and classification

[1723] server:

[1724] The collected news articles are first preprocessed using text analysis libraries such as NLTK and spaCy, which include removing noise data (such as advertisements and links) using regular expressions and standardizing character encoding.

[1725] The preprocessed news articles are then fed into a topic classification model (e.g., BERT, GPT-3) to classify them into specific categories (e.g., environment, politics, science and technology).

[1726] Summary Generation

[1727] server:

[1728] The classified news articles are fed into a generative AI model (e.g., OpenAI's GPT-3), which then summarizes the article in an easy-to-understand way based on the user's (parent or child's) age preference. Examples of prompts include:

[1729] "Summarize the following news article in a way that is understandable for a 10-year-old.

[1730] News Article:

[1731] As the Earth's temperature rises, various measures are being taken around the world...

[1732] "

[1733] The generative AI model uses this prompt to summarize the news article as something like, "The temperature of the earth has been rising recently. We are all working hard to prevent this."

[1734] Providing news summaries

[1735] Device:

[1736] When a user opens the application, the device sends an HTTP request to the server to retrieve the latest summarized news articles, which are then displayed in a list format in the device's user interface using RecyclerView and ListView.

[1737] News read status management

[1738] User:

[1739] Users (mainly children) can tap on the displayed news article to view details, and then press the "read" button after viewing.

[1740] Device:

[1741] The "read" mark data is sent from the device to the server by sending an HTTP POST request, and the server stores this data in a database and manages it as the user's browsing history.

[1742] View related news

[1743] server:

[1744] The server recategorizes the summarized news articles by topic and stores them in a database, allowing users to simultaneously view other related articles as they read a particular article.

[1745] Feedback collection

[1746] User:

[1747] Users can enter their feedback on the news article in the form and press the "Submit" button. They can enter requests such as "I wish this article had more details."

[1748] Device:

[1749] The device sends the feedback data to the server, which stores it in a database and uses it to improve the generative AI model.

[1750] SNS sharing function

[1751] User:

[1752] Users can click the "Share" button on the details screen of a news article to share it on social media.

[1753] Device:

[1754] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[1755] Model Improvement

[1756] server:

[1757] The server analyzes the collected feedback data and uses the results to tune the generative AI model, which is then used to generate the next news summary.

[1758] Specific examples

[1759] For example, if a 10-year-old child were to use this system, the process would go something like this:

[1760] 1. When a child opens the app, the device retrieves the latest summary news article (e.g., "The temperature of the earth has been rising recently. We are all working hard to prevent this.") from the server.

[1761] 2. The child reads the article and then presses the "read" button, which saves the read information on the server.

[1762] 3. Additionally, a related news item, "The Importance of Recycling," is displayed, which children can also read.

[1763] Thus, the present invention is a system that provides easy-to-understand summaries of news articles to children, helps them understand current affairs and social situations, and promotes communication with parents and teachers.

[1764] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1765] Step 1:

[1766] News article collection

[1767] The server sends HTTP GET requests to a news source (e.g., a news API). The input is the API endpoint and authentication information, and the server executes this periodically. The output is the JSON data of the retrieved news articles. The server converts this data into a pandas dataframe and stores it in a temporary data store (e.g., a SQL database).

[1768] Step 2:

[1769] News article preprocessing

[1770] The server retrieves news articles from the temporary data store and preprocesses them using a text analysis library (e.g., NLTK, spaCy). The input is the news article data saved in the previous step, and the output is the preprocessed clean text data. Specific operations include removing noise data (advertisements, links, etc.) using regular expressions and standardizing character encoding.

[1771] Step 3:

[1772] Topic Classification

[1773] The server inputs preprocessed news articles into a topic classification model (e.g., BERT, GPT-3). The input is clean text data, and the output is data classified into specific categories (e.g., environment, politics, science and technology). Specifically, the text data is converted into feature vectors, and classification results are obtained using a pre-trained model.

[1774] Step 4:

[1775] Summary Generation

[1776] The server inputs the classified news article into a generative AI model (e.g., OpenAI's GPT-3). The input is the classified news article and a prompt, and the output is a summarized news article based on the user's age preference. Specifically, the prompt is generated in the following format:

[1777] "Summarize the following news article in a way that is understandable for a 10-year-old.

[1778] News Article:

[1779] As the Earth's temperature rises, various measures are being taken around the world...

[1780] "

[1781] The generative AI model uses this information to summarize the news, and the server stores the summarized article in a database.

[1782] Step 5:

[1783] Providing news summaries

[1784] When a user opens the application on a device, the device sends an HTTP GET request to the server to request a summary news article. The input is the user's request, and the output is the latest summary news article data. The server sends this in JSON format to the device, which then displays it using RecyclerView or ListView.

[1785] Step 6:

[1786] News read status management

[1787] Users can tap on a news article to view details and then press the "mark as read" button after viewing. The input is the user's action, and the output is the updated read data. The device sends this data as an HTTP POST request to the server, which stores it in a database.

[1788] Step 7:

[1789] View related news

[1790] The server reclassifies summarized news articles by topic and stores them in a database. The input is the summarized news article data, and the output is the data classified by topic. When a user reads a particular article, the device sends a request to the server to display related news as well. The server sends the corresponding related news in JSON format to the device, and the device displays it.

[1791] Step 8:

[1792] Feedback collection

[1793] The user enters their feedback on the news article into the form and presses the "Submit" button. The input is the user's feedback content, and the output is the feedback data. The terminal sends the feedback data to the server via an HTTP POST request, and the server stores it in a database.

[1794] Step 9:

[1795] SNS sharing function

[1796] The user presses the "Share" button on the news article details screen, which takes them to the SNS sharing selection screen. The input is the user's operation, and the output is a sharing link and article summary. The device uses the SNS API to send the article summary and sharing link.

[1797] Step 10:

[1798] Model Improvement

[1799] The server analyzes the collected feedback data and tunes the generative AI model. The input is the feedback data and the output is the improved model. The server applies the new model to be used in generating the next news summary.

[1800] (Application example 1)

[1801] 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."

[1802] In today's world, it is important for children to have opportunities to read news articles, but the content of news is often too technical and difficult to understand. Therefore, there is a need for an easy-to-understand news summary system that helps children understand and be interested in the news.

[1803] 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.

[1804] In this invention, the server includes means for collecting news content, means for summarizing the collected news content according to age, means for providing the summarized news content to the information terminal, means for using a generative AI model to generate a summary, means for displaying news content selected by a user on the information terminal, means for marking the displayed news content as read, means for collecting user feedback, means for instructing the generative AI model to generate a summary using a prompt sentence, means for classifying the news content by related topic, means for displaying related news content together, means for saving the news content in a knowledge base, and means for providing summarized news for each related topic to the information terminal. This enables children to understand the news in an easy-to-understand manner and continue reading with interest.

[1805] "News content" refers to information and articles about current events and happenings obtained from a variety of sources.

[1806] "Means of collection" refers to the methods and devices used to obtain news content from news APIs and other information services and store it on a server.

[1807] "Age-appropriate summarization methods" are algorithms or systems that simplify collected news content for specific age groups, making it easier to understand.

[1808] The "means for providing to an information terminal" refers to a method or device for transmitting summarized news content from a server to an information terminal (such as a smartphone or tablet) and displaying it.

[1809] A "generative AI model" is a machine learning model that uses artificial intelligence technology to analyze input text and perform tasks such as summarizing and classifying it.

[1810] A "prompt sentence" is input text used to instruct a generative AI model to perform a specific operation or process.

[1811] The "means for displaying news content selected by the user" refers to an interface or method for displaying news content selected by the user on an information terminal.

[1812] A "means for marking as read" is a method or device for marking news content viewed by a user as read.

[1813] A "means for collecting feedback" is a method or system for obtaining opinions and thoughts from users and transmitting them to a server for storage.

[1814] A "topical classification method" is an algorithm or system that categorizes collected news content into specific themes or topics.

[1815] The "means for displaying related news content together" refers to a method or device for simultaneously displaying other news content related to the news that the user is viewing.

[1816] A "means for storing in a knowledge base" is a method or apparatus for storing categorized news content in a database or other storage system for future reference.

[1817] The present invention relates to a system that collects news content, summarizes it for specific age groups, and provides it to information terminals. This system includes functions such as news collection, preprocessing, classification, summary generation, provision, read management, display of related news, feedback collection, summarization using a generative AI model, and instructions using prompt sentences. Specific embodiments for implementing this invention are described below.

[1818] Gathering news content

[1819] The server periodically sends requests to information providers such as news APIs to retrieve the latest news content, which is then stored in a temporary data store.

[1820] News content preprocessing and classification

[1821] The server preprocesses the retrieved news content using a text analysis library. This preprocessing includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The preprocessed news content is then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[1822] Summary Generation

[1823] The server inputs the classified news content into a generative AI model. The generative AI model then summarizes the news content in an easy-to-understand format based on the user's (parent's or child's) age setting. For example, for an 8-year-old child, the summary might be simplified to something like, "The temperature of the earth is rising. We are all working together to stop this."

[1824] Providing news summaries

[1825] When the device opens the application, it sends a request to the server to retrieve the latest news summary, which is then displayed in a list format on the device's user interface.

[1826] News read status management

[1827] Users can tap on the displayed news content to view the details, then press a button to mark it as "read." The device then sends the "read" mark data to the server, which then stores it in a database. This allows the user's browsing history to be managed.

[1828] View related news

[1829] The server categorizes the summarized news content by topic and stores it in a database so that related news content can be displayed all at once. When a user reads a particular news item, other related news content is also displayed.

[1830] Feedback collection

[1831] Users can fill out a feedback form about news content and press the "Submit" button. For example, they can enter a request such as "I would like this article to be more detailed." The device then sends the feedback data to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[1832] SNS sharing function

[1833] Users can share news content on social media by clicking the "Share" button on the news content details screen. The device displays a social media selection screen and sends a news summary and a sharing link to the social media platform selected by the user. The server then displays the shared news content to other users in a format optimized for the target age group.

[1834] Model Improvement

[1835] The server analyzes the collected feedback data and uses the results to tune the generative AI model, which is then used to generate the next news summary.

[1836] Specific examples

[1837] As a concrete example, consider a 10-year-old child using an app. When the app is opened, a summary article titled "New Discoveries in Science and Technology" ("A new star has been discovered, which may reveal more about the secrets of the universe") is retrieved from the server. The child then reads the article and presses the "Read" button. Related news items, such as "Advances in Science and Technology," are then displayed, which the child can read. The child can also share the article they have read on social media and send feedback. An example of this prompt is as follows:

[1838] Example of input prompt for generative AI model

[1839] Send a request to the API to get the latest news articles and summarize them for kids aged 8-12.

[1840] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1841] Step 1:

[1842] The server periodically sends requests to the news API to retrieve the latest news content. The retrieved news content is stored in a temporary data store. The input is the news content from the news API, and the output is the data stored in the temporary data store. Specifically, it sends an HTTP request, parses the response, and stores it in a database.

[1843] Step 2:

[1844] The server preprocesses the news content retrieved from the temporary data store. Preprocessing includes removing noise data (advertisements, links, etc.) and normalizing character encoding. The input is the news content from the temporary data store, and the output is the preprocessed news content. Specifically, it uses a text analysis library to remove unnecessary information and standardize the format.

[1845] Step 3:

[1846] The server inputs the preprocessed news content into a topic classification model and classifies it into specific categories (e.g., environment, politics, science and technology). The input is the preprocessed news content, and the output is the news content classified by category. The specific operation is to use a machine learning model to determine which category each news content belongs to.

[1847] Step 4:

[1848] The server inputs the classified news content into a generative AI model and summarizes the news content based on the age setting. The input is news content classified by category, and the output is summarized news content appropriate for the age. A prompt sentence is used to instruct the generative AI model to generate a summary. The specific operation is to input an appropriate prompt sentence to the generative AI model and retrieve the generated summary.

[1849] Step 5:

[1850] When the application is opened, the device sends a request to the server to retrieve the latest news summary. The input is the user request, and the output is the news summary for display. The specific operation is to send an HTTP request to retrieve data from the server and display it on the user interface.

[1851] Step 6:

[1852] The user can tap on the displayed news content to view details and press a button to mark it as "read." The input is the user's operation, and the output is the news content marked as read. The specific operation is to tap on the user interface and then press the "read" button.

[1853] Step 7:

[1854] The terminal sends the "read" mark data to the server, and the server stores it in a database. The input is the read mark data, and the output is the data stored in the database. The specific operation is to send the data to the server by sending an HTTP request, and the server stores the read information in the database.

[1855] Step 8:

[1856] The server categorizes the summarized news content by topic and stores it in a database in order to display related news content in one place. The input is the summarized news content, and the output is the news content categorized by topic. The specific operation is to use a classification algorithm to extract related news and store it in the database.

[1857] Step 9:

[1858] The user fills in the feedback form for the news content and presses the "Send" button. The input is the user's feedback, and the output is feedback data from the terminal. The specific operation is to fill in the feedback form on the user interface and press the send button.

[1859] Step 10:

[1860] The terminal sends feedback data to the server, which stores it in a database. The input is the feedback data, and the output is the data stored in the database. The specific operation is to send an HTTP request to the server to send data, and then store the feedback information on the server.

[1861] Step 11:

[1862] Users can share news content on social media by pressing the "Share" button on the news content details screen. The input is the user's share operation, and the output is the news content shared on the social media. The specific operation is that after pressing the share button, an SNS selection screen is displayed, and the news summary and sharing link are sent to the selected SNS.

[1863] Step 12:

[1864] The server analyzes the collected feedback data and tunes the generative AI model based on the results. The input is the feedback data, and the output is the tuned generative AI model. Specifically, the server analyzes the feedback data and reflects the analysis results in the parameters of the generative AI model.

[1865] 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.

[1866] The present invention relates to a system that collects news articles, summarizes them in an easy-to-understand manner for children, and combines them with an emotion engine that recognizes the user's emotions. Hereinafter, embodiments of the present invention will be described with reference to specific examples.

[1867] News article collection

[1868] server:

[1869] The server periodically sends requests to a news source (e.g., a news API) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[1870] News article preprocessing and classification

[1871] server:

[1872] The collected news articles are first preprocessed using a text analysis library, which includes removing noise data (such as advertisements, links, and special characters) and normalizing character encoding. The preprocessed news articles are then fed into a topic classification model to be classified into specific categories (e.g., environment, politics, science and technology).

[1873] Summary Generation

[1874] server:

[1875] The classified news articles are then fed into a generative AI model, which then summarizes the article in an easy-to-understand way based on the user's (parent's or child's) age preference. For example, for an 8-year-old child, the article might be simplified to something like, "The earth's temperature is rising. We're all working together to stop this."

[1876] Providing and displaying news summaries

[1877] Device:

[1878] When a user opens the application, the device sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the device's user interface.

[1879] News read status management

[1880] User:

[1881] Users (usually children) tap on a displayed news article to view details, then press a button to mark it as "read."

[1882] Device:

[1883] The "read" mark data is sent from the device to the server and stored in a database on the server side, allowing the user's browsing history to be managed.

[1884] View related news

[1885] server:

[1886] To display related news articles in one place, the server categorizes the summarized news articles by topic and stores them in a database, so that when a user reads a particular article, other related articles are also displayed.

[1887] Feedback collection

[1888] User:

[1889] Users can fill out a feedback form for a news article and press the "Submit" button. For example, they can enter a request such as "I'd like this article to be more detailed."

[1890] Device:

[1891] The feedback data is sent from the device to the server, where it is stored in a database. The stored feedback data is used to improve the model.

[1892] SNS sharing function

[1893] User:

[1894] Users can click the "Share" button on the details screen of a news article to share it on social media.

[1895] Device:

[1896] A social networking site selection screen will be displayed, and the article summary and sharing link will be sent to the social networking site selected by the user.

[1897] server:

[1898] Shared news articles will be displayed to other users in an age-optimized format.

[1899] Model Improvement

[1900] server:

[1901] The collected feedback data is analyzed and the generative AI model is tuned based on the results. The improved model is used to generate the next news summary.

[1902] Introducing the Emotion Engine

[1903] emotion recognition

[1904] Device:

[1905] While a user is browsing a news article, the device uses a built-in camera and microphone to analyze the user's emotions in real time. The emotion engine analyzes the user's facial expressions and tone of voice to generate emotion data.

[1906] Emotion data analysis and storage

[1907] server:

[1908] The emotion data sent from the emotion engine is analyzed and stored in a database, thereby accumulating a history of the user's emotions regarding news articles.

[1909] Selecting the next news story

[1910] server:

[1911] The next news article to be served is selected based on the accumulated emotional data and the browsing history of news articles. For example, if the user shows a "happy" expression, the system will serve up positive news.

[1912] Specific examples

[1913] As a concrete example, consider a 10-year-old child using the app. While the child is reading a summary article titled "New Report on Climate Change," the emotion engine analyzes the child's facial expressions and determines that the user is interested. A related news article about "The Importance of Recycling" is also displayed, which the child can read. After finishing the article, the child presses the "Read" button, and the emotion data is saved in the database.

[1914] In this way, by summarizing news articles in an easy-to-understand manner and combining it with emotion recognition, the present invention can provide appropriate news articles for children, enhance learning effectiveness, promote communication with parents and teachers, and encourage children to become interested in current events and social situations.

[1915] The processing flow will be explained below.

[1916] Step 1:

[1917] Server: Sends requests to news sources (e.g., news APIs) to retrieve the latest news articles, which are then stored in a temporary data store on the server.

[1918] Step 2:

[1919] Server: The retrieved news articles are preprocessed using a text analysis library, which includes removing noise data (advertisements, links, special characters, etc.) and normalizing character encoding.

[1920] Step 3:

[1921] Server: Preprocessed news articles are input into a topic classification model. The model classifies news articles into specific topics (e.g., environment, politics, science and technology). The classification results are stored in a temporary data store.

[1922] Step 4:

[1923] Server: The classified news articles are fed into a generative AI model to be summarized for children. Based on the age setting (e.g., elementary school, middle school), a summary text is generated in an easy-to-understand format.

[1924] Step 5:

[1925] Server: Stores summarized news articles and their metadata (topic, target demographic, etc.) in a database. The stored data is linked to other related news articles.

[1926] Step 6:

[1927] Device: When a user (parent or child) opens the application, the device sends a request to the server to retrieve the latest summary news articles.

[1928] Step 7:

[1929] Terminal: The retrieved news summaries are displayed in a user interface in the form of a list, with each news item showing its title and summary text.

[1930] Step 8:

[1931] User: Taps on a displayed news article to view details, then presses a button to mark it as "read."

[1932] Step 9:

[1933] Terminal: Detects user actions and sends the "read" mark data to the server.

[1934] Step 10:

[1935] Server: The received "read" data is saved in a database and managed as the user's browsing history.

[1936] Step 11:

[1937] Server: Based on the related information of the news article, selects news articles on related topics and prepares them for provision to users.

[1938] Step 12:

[1939] On-device: When a user reads a particular article, other related articles are displayed along with it. Related news is displayed in a list format for easy access.

[1940] Step 13:

[1941] User: Can fill out a form to provide feedback on a news article and press the "Submit" button.

[1942] Step 14:

[1943] Terminal: Sends feedback data to the server.

[1944] Step 15:

[1945] Server: The received feedback data is stored in a database and used as reference data for model improvement.

[1946] Step 16:

[1947] User: Press the "Share" button on the news article details screen to share the news on social media.

[1948] Step 17:

[1949] Device: The SNS selection screen is displayed, and the article summary and sharing link are sent to the SNS selected by the user.

[1950] Step 18:

[1951] Server: News articles shared on social media are displayed to other users in a format optimized for the target age group.

[1952] Step 19:

[1953] Server: Analyzes the collected feedback data and tunes the generative AI model. The improved model is used to generate the next news summary.

[1954] Step 20:

[1955] Device: While the user is viewing a news article, the device uses the built-in camera and microphone to analyze the user's facial expressions and tone of voice in real time.

[1956] Step 21:

[1957] Device: The emotion engine generates the user's emotion data and temporarily stores the analysis results on the device.

[1958] Step 22:

[1959] Terminal: Sends emotion data to the server in real time.

[1960] Step 23:

[1961] Server: Analyzes the emotion data sent from the emotion engine and stores it in a database, thereby accumulating a history of users' emotions toward news articles.

[1962] Step 24:

[1963] Server: Based on the accumulated emotional data and the browsing history of news articles, the server selects the next news article to be provided. For example, if the user shows a "happy" expression, the server will provide positive news.

[1964] Examples:

[1965] For example, if a 10-year-old child is using the app to read a new report on climate change, the emotion engine will analyze the child's facial expressions and determine that the user is interested. Related news about the importance of recycling will also be displayed. After finishing the article, the child can press the "read" button, and the emotion data will be stored in the database, and the next appropriate news item will be displayed.

[1966] Example 2

[1967] 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."

[1968] Existing news article delivery systems often do not adequately summarize articles for children, making them difficult to understand. Furthermore, they are unable to provide news that takes into account user feedback and emotions, making it difficult to sustain interest. Furthermore, they do not adequately provide relevant news articles, resulting in low learning outcomes for users.

[1969] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1970] In this invention, the server includes means for collecting news articles, means for preprocessing the collected news articles, means for classifying the preprocessed news articles using a topic classification model, means for inputting the classified news articles into a generative AI model using prompt sentences and summarizing them according to age, and means for providing the summarized news articles to a terminal. This makes it possible to provide news articles in a format that is easy for children to understand. Furthermore, by utilizing user feedback and emotional data, it is possible to provide more appropriate articles for each user and improve learning effectiveness. Furthermore, by displaying related news articles together, it is possible to deepen the user's understanding.

[1971] A "news article" is written information about a current event or topic obtained from online or offline sources.

[1972] "Preprocessing" is the process of performing data preparation tasks such as removing noise data from the collected text data of news articles and normalizing character encoding.

[1973] A "topic classification model" is a machine learning model used to automatically classify news articles into specific categories (e.g., environment, politics, science and technology, etc.).

[1974] A "prompt" is a text sentence that provides the information or instructions needed to input into a generative AI model.

[1975] A "generative AI model" is an artificial intelligence model that uses natural language processing techniques to summarize or generate information from presented input data.

[1976] "Summarization" refers to extracting the key information from an original news article and restating it in a shorter form.

[1977] A "terminal" is a computing device (e.g., smartphone, tablet, PC, etc.) that displays news articles and accepts user operations.

[1978] "User feedback" refers to information such as opinions, requests, and evaluations provided by users regarding news articles.

[1979] "Related news articles" are other news articles that are related in topic or content to a particular article.

[1980] A "knowledge base" is a database that systematically stores news articles and related data, enabling them to be searched and analyzed.

[1981] The present invention relates to a system for providing appropriate news articles to children and enhancing their learning effect. Specific embodiments of the present invention will be described below.

[1982] News article collection and preprocessing

[1983] server:

[1984] The server periodically sends HTTP requests to information sources, such as news APIs, to retrieve the latest news articles. The retrieved news articles are received as JSON-formatted data and stored in a temporary data store. The server then preprocesses the collected news articles using a text analysis library (e.g., NLTK, SpaCy, etc.). Preprocessing includes removing noise data such as advertisements, links, and special characters, and normalizing character encoding.

[1985] News article classification and summarization

[1986] server:

[1987] The preprocessed news articles are fed into a topic classification model (e.g., environment, politics, science and technology) to be classified into specific categories. The classified news articles are then fed into a generative AI model (e.g., GPT-3) with a prompt. The generative AI model summarizes the articles based on the user's specified age.

[1988] Example prompt sentence:

[1989] "Please provide a quick summary of the following article for an 8-year-old: {News article text}"

[1990] The generated summary articles are stored in a database.

[1991] Providing and displaying news summaries

[1992] Device:

[1993] When a user opens the application, the device sends an HTTP request to the server to retrieve the latest summarized news articles, which are then displayed in a user interface (e.g., a list view).

[1994] News read tracking and feedback collection

[1995] User:

[1996] Users can tap on a news article to view details and press the "mark as read" button after finishing reading. They can also fill out a feedback form for the news article and press the "submit" button. The feedback data includes requests and improvements for the article content.

[1997] Device:

[1998] The "read" mark data and feedback data are sent from the terminal to the server, where the data is stored in a database and the user's browsing history and feedback information are managed.

[1999] Displaying related news and managing knowledge base

[2000] server:

[2001] The server categorizes summarized news articles by topic and stores them in a database. When a user views a particular article, other related articles are also displayed. This display of related news can deepen the user's understanding. The server also provides a knowledge base for systematically storing news articles and related data.

[2002] Introducing an emotion engine and selecting the next news article

[2003] Device:

[2004] While a user is browsing a news article, the system uses the built-in camera and microphone to analyze the user's emotions in real time. The emotion engine analyzes the user's facial expressions and tone of voice to generate emotion data.

[2005] server:

[2006] The emotional data sent from the emotion engine is received and analyzed. The analyzed emotional data is stored in a database, and the user's emotional history regarding news articles is accumulated. The server selects the next news article to be provided based on the accumulated emotional data and the news article viewing history. For example, if the user shows a happy expression, positive news will be provided first.

[2007] Specific examples

[2008] As a concrete example, consider a 10-year-old child using the app. While the child is viewing a summary of a "new report on climate change," the emotion engine analyzes the child's facial expressions and determines that the user is interested. Related news articles about the importance of recycling are also displayed, which the child can read. After finishing the article, the child presses the "read" button, and the emotion data is stored in the database. In this way, by summarizing news articles in an easy-to-understand manner and combining it with emotion recognition, the app can provide appropriate news articles for children and improve their learning. It can also promote communication with parents and teachers, encouraging children to become interested in current events and social situations.

[2009] Thus, the present invention is a system that provides users with an innovative news article delivery experience by integrating a series of processes, including news article collection, preprocessing, classification, summarization, delivery, emotion recognition, and display of related news.

[2010] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2011] Step 1:

[2012] News article collection

[2013] The server periodically sends HTTP requests to the news API to retrieve the latest news article data in JSON format, which is then stored in a temporary data store on the server.

[2014] Input: A request to the News API

[2015] Data processing: Obtaining JSON data from the news API

[2016] Output: Retrieved news article data

[2017] Specifically, it sets up a regular scheduled job to access the news API every hour or every day.

[2018] Step 2:

[2019] News article preprocessing

[2020] The server uses a text analysis library (e.g., NLTK, SpaCy, etc.) to preprocess the news article data, removing noise data (e.g., advertisements, links, special characters, etc.) and normalizing character encoding.

[2021] Input: Acquired news article data

[2022] Data processing: Removal of noise data, normalization of character encoding

[2023] Output: Preprocessed news article data

[2024] Specifically, we remove noise data using regular expressions and split words using SpaCy's tokenizer.

[2025] Step 3:

[2026] News article classification

[2027] The server uses a topic classification model to classify the preprocessed news article data into specific categories (e.g., environment, politics, science and technology, etc.).

[2028] Input: Preprocessed news article data

[2029] Data processing: Category classification using topic classification model

[2030] Output: Categorized news article data

[2031] Specifically, the operation involves applying a text classification algorithm using a machine learning model.

[2032] Step 4:

[2033] News article summary generation

[2034] The server inputs the classified news article data into a generative AI model (e.g., GPT-3) and summarizes the article based on the user's specified age. Specific summarization instructions are given to the generative AI model using prompt sentences.

[2035] Input: Categorized news article data, prompt

[2036] Data Computation: Generative AI Models for Summarization

[2037] Output: Summarized news article data

[2038] Example prompt: "Please provide a brief summary of the following article for an 8-year-old: {news article text}"

[2039] Step 5:

[2040] Providing news summaries

[2041] The terminal sends a request to the server to retrieve the latest summarized news articles, which are then displayed in a list format on the terminal's user interface.

[2042] Input: Summary news article data on the server

[2043] Data processing: None

[2044] Output: A summary news article displayed on your terminal

[2045] Specifically, news data in JSON format is retrieved via an HTTP request and displayed in a list view.

[2046] Step 6:

[2047] News read status management

[2048] Users tap on a news article to view the details, and when they are finished reading, they press the "read" button. The device sends the "read" mark data to the server, which stores it in a database.

[2049] Input: User presses the read button

[2050] Data processing: Saving read information to a database

[2051] Output: Updated user browsing history

[2052] Specifically, the read information is sent to the server in JSON format as a POST request.

[2053] Step 7:

[2054] Feedback collection

[2055] The user fills in the feedback form for the news article and presses the "Submit" button. The terminal sends the feedback data to the server, where it is stored in a database.

[2056] Input: User feedback data

[2057] Data processing: Saving feedback data to a database

[2058] Output: Updated feedback information

[2059] Specifically, the input contents of the feedback form are sent to the server in JSON format as a POST request.

[2060] Step 8:

[2061] View related news

[2062] The server categorizes summarized news articles by topic and stores them in a database. When a user views a particular article, other related articles are also retrieved and displayed.

[2063] Input: Summary news article data

[2064] Data processing: Acquisition of related news article information

[2065] Output: User interface with related news displayed

[2066] Specifically, news articles belonging to the same category are retrieved from the server as "related articles" via a query.

[2067] Step 9:

[2068] Emotion data analysis and storage

[2069] While a user is viewing a news article, their emotions are analyzed in real time using the device's built-in camera and microphone. The server receives the analyzed emotional data and stores it in a database.

[2070] Input: User facial and voice data

[2071] Data processing: Emotion data analysis using an emotion engine

[2072] Output: Emotion data stored in a database

[2073] Specifically, emotions are analyzed using facial recognition technology and voice emotion recognition technology and stored in a database.

[2074] Step 10:

[2075] Selecting the next news story

[2076] The server selects the next news article to serve based on the accumulated emotional data and the browsing history of news articles. For example, if the user shows a happy expression, it will prioritize positive news articles.

[2077] Input: Emotion data, news article browsing history

[2078] Data processing: Selecting the next news article to be served

[2079] Output: Selected news article data

[2080] Specifically, we apply a recommendation algorithm based on sentiment analysis.

[2081] (Application example 2)

[2082] 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."

[2083] Current news article distribution systems have difficulty providing appropriate and easy-to-understand summaries for children. Furthermore, they lack mechanisms for grasping the degree to which children understand a news article or the emotions it evokes. As a result, they are unable to fully stimulate children's interest or enhance their learning. Another issue is that they are unable to provide personalized news article content, resulting in a uniform distribution system.

[2084] 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.

[2085] In this invention, the server includes means for collecting news articles, means for summarizing the collected news articles according to age, means for providing the summarized news articles to the terminal, means for recognizing the user's emotions while the summarized news articles are displayed, and means for analyzing and saving the emotion data. This makes it possible to provide news articles in a format that is interesting and easy for children to understand, and to personalize the next news article to be provided based on the emotion data.

[2086] "News articles" are text data about current events and happenings distributed by news organizations and information providers.

[2087] "Means of collection" refers to the ability to obtain the latest news articles from news APIs and other information sources via the network.

[2088] "Age-appropriate summarization" refers to the function of making collected news articles easier to understand and concise for a specific age group.

[2089] "Means for providing" refers to the function of delivering summarized news articles to user terminals so that they can be viewed.

[2090] "Device" refers to a device such as a smartphone, tablet, or computer that a user uses to view news articles.

[2091] "Means for recognizing emotions" refers to a function that uses a camera or microphone to analyze facial expressions and tone of voice when a user is viewing a news article, in order to determine emotions.

[2092] "Emotional data" refers to digital data that indicates the emotional state of a user analyzed from facial expressions, tone of voice, etc.

[2093] "Means for analyzing and storing" refers to the function for processing the recognized emotional data and storing it in storage such as a database.

[2094] "Personalization" refers to optimizing and providing content and information to suit the interests and concerns of each individual user.

[2095] The present invention relates to a system that summarizes news articles in an easy-to-understand manner for children and combines it with an emotion engine that recognizes the user's emotions. To effectively implement this invention, the server, terminals, and users must work together to execute each step.

[2096] server

[2097] The server collects news articles, preprocesses them, generates summaries, and analyzes and stores sentiment data.

[2098] 1. Collect news articles periodically from a news API and store them in a temporary data store. Here, you can use an existing news API such as NewsAPI.

[2099] 2. Preprocess the collected news articles using a text analysis library such as TextBlob to remove noise data and normalize character encoding.

[2100] 3. Classify the preprocessed news articles using a topic classification model and generate age-appropriate summaries using a generative AI model (e.g., OpenAI GPT-3), with an example prompt such as "Summarize the news article for a 10-year-old: [insert news article here]."

[2101] 4. The summarized news articles are stored in a database.

[2102] 5. While the user is browsing a news article, the emotion engine analyzes the user's facial expressions and tone of voice through the camera and microphone to generate emotion data.

[2103] 6. Emotional data is analyzed and stored in a database along with browsing history and related news.

[2104] Terminal

[2105] The device displays news articles and collects and transmits emotion data through a user interface.

[2106] 1. When a user opens the application, it sends a request to the server to get the latest summary news articles.

[2107] 2. The retrieved news articles are displayed in a list format on the user interface.

[2108] 3. When a user selects a news article and views its details, the device's built-in camera and microphone are used to analyze the user's emotions in real time and send the data to the emotion engine.

[2109] 4. After the user reads a news article, a button is provided to mark it as "read" and the data is sent to the server.

[2110] User

[2111] Users (mainly children) operate the device to read news articles and cooperate in collecting emotion data.

[2112] 1. Select the news article that interests you from the list of news articles.

[2113] 2. Have the students read news articles and analyze emotional data via a camera or microphone.

[2114] 3. After reading the details of the news article, press the "read" button to mark the article and provide feedback.

[2115] As a concrete example, consider a 10-year-old child viewing a summary article titled "New Report on Climate Change." While the child is reading the article, the emotion engine analyzes the child's facial expressions and determines that the child is interested. It also displays a related news article about "The Importance of Recycling," which the child can continue reading. This emotion information and browsing history are stored in a database and used for future news distribution.

[2116] In this way, by operating this system in cooperation with the server, terminals, and users, news articles can be provided in a format that is easy for children to understand and that will interest them, thereby improving the effectiveness of their learning.

[2117] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2118] Step 1:

[2119] T...

Claims

1. a means of collecting news articles; A means of summarizing collected news articles in an age-appropriate manner; means for providing summarized news articles to a terminal; A system including:

2. means for displaying a news article selected by the user on the terminal; means for marking a displayed news article as read; a means for collecting user feedback; The system of claim 1 , comprising:

3. a means of categorizing news articles by related topics; a means of displaying related news articles together; a means for storing news articles in a knowledge base; The system of claim 1 , comprising:

4. A way to share news articles on social media, A way to optimise shared news articles for specific age groups; The system of claim 1 , comprising:

5. A means to analyze news article feedback data to improve the model; and a means for utilizing the improved model in subsequent summary generation; The system of claim 1 , comprising:

Images