System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting a news browsing history of a user; means for storing the collected news browsing history in a database; means for analyzing the stored news browsing history to generate a personalized news report; and means for providing the generated personalized news report to the user.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] News is an essential source of information in our lives, but the problem is that in our busy daily lives, we forget the content of the articles we read. Furthermore, even important news stories tend to be reported less frequently as new topics emerge, and they tend to fade from people's memories. Given this situation, there is a need for a method that allows users to look back on news and efficiently obtain follow-up and related information. Furthermore, in today's information-overloaded world, there is a need for a system that allows users to easily access reliable sources of information. [Means for solving the problem]

[0005] The present invention provides a system that includes a means for collecting a user's news viewing history and storing the collected news viewing history in a database, a means for analyzing the stored news viewing history to generate a personalized news report, and a means for providing the generated personalized news report to the user. This system allows the user to easily review and deepen their understanding of news viewed in the past. Furthermore, by including a means for searching for follow-up reports requested by the user and adding them to the news report, and a means for providing additional information related to the generated news report, the system achieves continuous news learning and provides reliable information.

[0006] "News browsing history" is a record of news articles that a user has viewed in the past.

[0007] A "database" is a system that systematically organizes and stores information, allowing it to be efficiently searched and accessed as needed.

[0008] A "personalized news report" is a news summary that is individually customized based on a user's interests and browsing history.

[0009] "Generative AI" is a system that uses artificial intelligence technology to generate new information and content.

[0010] A "user identifier" is a unique ID or cookie that identifies a specific user.

[0011] The "RAG model" refers to a model that provides information by combining search and generation.

[0012] A "news report" is a report summarizing news articles viewed by a user during a specific time range.

[0013] A "follow-up" is new information added after news has already been reported.

[0014] "Related information" is additional information or background knowledge related to a particular news article. [Brief explanation of the drawings]

[0015] [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

[0016] 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.

[0017] First, the terms used in the following description will be explained.

[0018] 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).

[0019] 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.

[0020] 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.

[0021] 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.

[0022] 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."

[0023] [First embodiment]

[0024] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0025] 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.

[0026] 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).

[0027] 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.

[0028] 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.

[0029] 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.

[0030] 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.

[0031] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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."

[0036] ---

[0037] The present invention relates to a system for collecting, storing, and analyzing a user's news browsing history, and generating and providing personalized news reports based on the results. The present invention is implemented by combining a specific software system and infrastructure.

[0038] 1. Collection and storage of news browsing information

[0039] server:

[0040] The server provides an API endpoint for receiving news browsing information from news sites and apps.

[0041] Receive metadata such as news article title, URL, author, date and time of access, and access duration.

[0042] Save the received data in the database for each user.

[0043] Device:

[0044] The device sets a user identifier and collects information about news articles the user has viewed on news sites and apps.

[0045] Send the collected data to the server's API endpoint.

[0046] User:

[0047] Users use news sites and apps to view news articles.

[0048] User operations are basically the same as when browsing news normally, so no special operations are required.

[0049] 2. Analyzing news browsing information and generating reports

[0050] server:

[0051] The server periodically analyzes the news browsing information stored in the database.

[0052] Leverage natural language processing techniques to identify trends in topics and articles that interest users.

[0053] Based on the analysis, generative AI is used to generate personalized news reports, including the user's recently viewed articles, related supplemental information, and follow-up reports.

[0054] Generation AI:

[0055] The generative AI takes into account the user's browsing history and interests and generates reports in natural language.

[0056] Example: "Updates on X: Updates and insights on last week's news story."

[0057] 3. Providing news reports

[0058] Device:

[0059] Once the generated news report is complete, a notification will be sent to your device.

[0060] The terminal notifies the user of the availability of the report.

[0061] User:

[0062] The user views the news report through the terminal.

[0063] Review reports and quickly review previous news and follow-ups to deepen your understanding.

[0064] 4. Requesting and Providing Follow-Up and Related Information

[0065] User:

[0066] While viewing a report, users can request further updates or related information on a particular topic.

[0067] server:

[0068] The server receives the request and uses RAG (Retrievable and Generative Model) to search for relevant up-to-date information on the Internet and other databases.

[0069] Update news reports with newly discovered information.

[0070] Generation AI:

[0071] The generator AI then updates the personalized news report again to include any follow-up or related information requested by the user.

[0072] Device:

[0073] Once the updated news report is ready, you will be notified again on your device.

[0074] User:

[0075] Users can view updated news reports to see the latest updates and additional information.

[0076] This system not only helps users efficiently track and deeply understand the news even in their busy daily lives, but also contributes to improving media literacy in society as a whole by providing reliable sources of information.

[0077] The processing flow will be explained below.

[0078] Explanation of the specific processing of the program

[0079] ---

[0080] Step 1:

[0081] User:

[0082] A user opens a news site or app and reads a news article.

[0083] Step 2:

[0084] Device:

[0085] The device uses the user identifier to record metadata such as the title, URL, and viewing time of news articles viewed by the user.

[0086] Step 3:

[0087] Device:

[0088] The device makes an API request to send the recorded news viewing information to the server in real time.

[0089] Step 4:

[0090] server:

[0091] The server processes the news browsing information received from the terminal and stores it in a database together with the corresponding user identifier.

[0092] Step 5:

[0093] server:

[0094] Periodically (eg, weekly), the server analyzes the news browsing history stored in the database.

[0095] Step 6:

[0096] server:

[0097] The server analyzes the browsing history using natural language processing techniques to identify the user's interests.

[0098] Step 7:

[0099] server:

[0100] The server uses generative AI to generate personalized news reports based on the analysis results.

[0101] Step 8:

[0102] server:

[0103] The generated news reports are stored in a database and are ready to be notified to the device.

[0104] Step 9:

[0105] Device:

[0106] The terminal notifies the user that a news report is available.

[0107] Step 10:

[0108] User:

[0109] The user receives a notification from the device and views the news report.

[0110] Step 11:

[0111] User:

[0112] Users review news reports and request follow-up or related information of interest.

[0113] Step 12:

[0114] Device:

[0115] The device makes an API request to send the request from the user to the server.

[0116] Step 13:

[0117] server:

[0118] Based on the received request, the server uses the RAG model to search the internet and other databases for relevant follow-up information.

[0119] Step 14:

[0120] server:

[0121] Based on the search results, the generating AI will update the personalized news report again, adding follow-up and related information.

[0122] Step 15:

[0123] server:

[0124] The updated news report is again stored in the database and notified to the terminal.

[0125] Step 16:

[0126] Device:

[0127] The terminal will again announce that an updated news report is available.

[0128] Step 17:

[0129] User:

[0130] Users can view updated news reports to see the latest updates and additions.

[0131] ---

[0132] Through this process, a system is provided that allows users to efficiently track and deeply understand news even in their busy daily lives.

[0133] Example 1

[0134] 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."

[0135] Conventional news article delivery systems lacked sufficient personalization functionality that took into account a user's news browsing history, making it difficult to accurately provide information that users were interested in. Furthermore, there was a lack of a way for users to easily request and obtain follow-up reports or related information on specific news stories. This made it difficult for users to quickly and efficiently obtain the latest and most relevant information.

[0136] 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.

[0137] In this invention, the server includes means for collecting news browsing information, means for storing the collected news browsing information in a database, means for analyzing the stored news browsing information using natural language processing technology and generating a personalized news report using a generative artificial intelligence model, and means for providing the generated personalized news report to a user. This makes it possible to provide a personalized news report that takes into account the user's news browsing history, and to quickly obtain and provide related information in response to a user's request.

[0138] "News viewing information" is viewing data when a user views a news article, and specifically includes metadata such as the news article title, URL, author, viewing date and time, and viewing time.

[0139] A "database" is an information system for efficiently storing, organizing, and retrieving large amounts of data.

[0140] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and generate human language, and includes classification, extraction, and text analysis.

[0141] A "generative artificial intelligence model" is an artificial intelligence technology that learns from large datasets and automatically generates text that is meaningful to users.

[0142] A "personalized news report" is a customized news summary generated based on a user's news browsing history and tailored to their individual interests and preferences.

[0143] A "server" is a computer system that provides services to other computers (clients) on a network.

[0144] A "request" is an action by a user requesting specific information or service, or the content of that request.

[0145] "Searching" is the act of locating information from a database or the Internet based on specified criteria.

[0146] "Related information" is additional data or insight that is relevant to the user's interests or the information they are currently viewing.

[0147] MODE FOR CARRYING OUT THE INVENTION

[0148] The present invention relates to a system for collecting, storing, and analyzing news browsing information, generating personalized news reports based on the collected information, and providing the reports to users. The present invention is implemented by combining a specific software system and infrastructure.

[0149] Collection and storage of news browsing information

[0150] server:

[0151] The server provides an API endpoint for receiving news viewing information from news sites and apps. It receives metadata such as the news article title, URL, author, viewing date and time, and viewing time. The received data is stored for each user ID in the "news_views" table in the MySQL database.

[0152] Device:

[0153] The device collects news articles that users have viewed through news sites and apps. The device generates a unique identifier (user ID) for each user and sends it to the server's API endpoint along with news viewing information.

[0154] User:

[0155] Users browse news articles on a daily basis using news websites and apps, with no special operations required.

[0156] News browsing analysis and report generation

[0157] server:

[0158] The server periodically queries the news viewing information stored in the "news_views" table to obtain the user's browsing history. It analyzes the obtained data and extracts article topics and keywords using a natural language processing library (e.g., SpaCy) that identifies topics and article trends that each user is interested in. Based on the analysis results, it generates prompts for a generative artificial intelligence model (e.g., OpenAI GPT-3.5) to create a personalized news report.

[0159] Generative AI models:

[0160] The generative AI model generates a news report in natural language based on the server's prompts and matches the user's interests.

[0161] Example: "Updates on X: Updates and insights on last week's news story."

[0162] Providing news reports

[0163] Device:

[0164] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability.

[0165] User:

[0166] Users receive notifications on their devices, view news reports, and check the report content to stay up to date on related news and updates.

[0167] Requesting and providing follow-up and related information

[0168] User:

[0169] While viewing a news report, users can request updates or additional information on a particular topic.

[0170] server:

[0171] The server receives user requests and performs searches based on the requests, using the RAG model (e.g., Haystack) to gather relevant, up-to-date information from the Internet and other databases.

[0172] Example prompt: "Create a personalized news report containing the latest news that may be of interest to the user based on their recent browsing history. Specifically, detail new information and related updates about X, as well as future projections."

[0173] Generative AI models:

[0174] The generative AI model then uses the additional information collected to generate an updated news report for the user.

[0175] Device:

[0176] Once the updated news report is ready, you will be notified again on your device.

[0177] User:

[0178] Users receive notifications, view updated news reports, and stay up to date with new developments and relevant information.

[0179] The system allows users to efficiently receive personalized news reports based on their news browsing history, and they can request additional information on specific topics to stay up to date.

[0180] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0181] Processing Steps

[0182] Step 1:

[0183] server:

[0184] The server provides an API endpoint for receiving news browsing information, including metadata such as the news article title, URL, author, browsing date and time, and browsing duration.

[0185] Input: News browsing information sent from your device.

[0186] Data processing: Analyze the received metadata and convert it into a format that can be saved in the database for each user ID.

[0187] Output: Organized metadata stored in a database.

[0188] Step 2:

[0189] Device:

[0190] The device collects news articles that users have viewed through news sites and apps. The device generates a unique identifier (user ID) for each user and sends it to the server's API endpoint along with news viewing information.

[0191] Input: News article information obtained from news sites and apps.

[0192] Data processing: A user ID is assigned and the news article metadata is converted into a format that can be sent to the server.

[0193] Output: News browsing information sent to the server's API endpoint.

[0194] Step 3:

[0195] User:

[0196] Users browse news articles on a daily basis using news websites and apps, with no special operations required.

[0197] Input: User's news article browsing behavior.

[0198] Data processing: None.

[0199] Output: News browsing data based on user browsing behavior.

[0200] Step 4:

[0201] server:

[0202] The server periodically queries the news browsing information stored in the database to obtain the user's browsing history, analyzes the obtained data, and identifies topics and article trends that each user is interested in. It then uses a natural language processing library (e.g., SpaCy) to extract article topics and keywords.

[0203] Input: News browsing information stored in a database.

[0204] Data processing: Using natural language processing technology, topics and keywords from news articles are extracted and user interests are analyzed.

[0205] Output: Analysis results that reflect each user's interests and concerns.

[0206] Step 5:

[0207] server:

[0208] Based on the analysis results, the server generates prompt sentences for a generative AI model (e.g., OpenAI GPT-3.5) to create a personalized news report.

[0209] Input: Results analyzed using natural language processing and user browsing history.

[0210] Data processing: Generate prompt sentences and feed them into a generative AI model to generate personalized news reports.

[0211] Output: The generated personalized news report.

[0212] Action: Example prompt: "Create a personalized news report containing the latest news that may be of interest to the user based on their recent browsing history. Specifically, include new information and related updates about ____, as well as future projections."

[0213] Step 6:

[0214] Device:

[0215] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability.

[0216] Input: A news report sent from the server.

[0217] Data processing: Convert to notification format.

[0218] Output: Notify the user and display a news report.

[0219] Step 7:

[0220] User:

[0221] Users receive notifications on their devices, view news reports, and check the report content to stay up to date on related news and updates.

[0222] Input: A news report posted to the device.

[0223] Data processing: None.

[0224] Output: Confirmed news report viewing behavior of the user.

[0225] Step 8:

[0226] User:

[0227] While viewing a news report, users can request updates or additional information on a particular topic.

[0228] Input: A user request for a news report.

[0229] Data processing: None.

[0230] Output: Request sent.

[0231] Step 9:

[0232] server:

[0233] The server receives user requests and performs searches based on the requests, using the RAG model (e.g., Haystack) to gather relevant, up-to-date information from the Internet and other databases.

[0234] Input: The user's request and the database to search.

[0235] Data processing: Search and retrieve relevant information using the RAG model.

[0236] Output: The most recent relevant information retrieved.

[0237] Step 10:

[0238] Generative AI models:

[0239] The generative AI model then uses the additional information collected to generate an updated news report for the user.

[0240] Input: retrieved related information, previous browsing history, user interests.

[0241] Data processing: Generate prompt sentences and feed them into a generative AI model to generate updated news reports.

[0242] Output: An updated personalized news report.

[0243] Step 11:

[0244] Device:

[0245] Once the updated news report is ready, you will be notified again on your device.

[0246] Input: Updated news report sent from the server.

[0247] Data processing: Convert to notification format.

[0248] Output: Notify user and display update news report.

[0249] Step 12:

[0250] User:

[0251] Users receive notifications, view updated news reports, and stay up to date with new developments and relevant information.

[0252] Input: Updated news report notified to terminal.

[0253] Data processing: None.

[0254] Output: User's viewing behavior for latest information.

[0255] (Application example 1)

[0256] 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."

[0257] In modern society, a huge amount of news information is provided via the Internet, but users face the challenge of efficiently gathering information relevant to them. In particular, the lack of personalized news reports tailored to individual users makes it difficult for users to access news information based on their interests. Furthermore, when users want more information on a particular topic, they have to search manually, which is inconvenient.

[0258] 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.

[0259] In this invention, the server includes means for collecting a user's news browsing history, means for storing the collected news browsing history in a database, means for analyzing the stored news browsing history and generating a prompt sentence using a generative AI model to generate a personalized news report, and means for providing the generated personalized news report to the user. This effectively provides a personalized news report to the user, allowing the user to easily access news information based on their interests. In addition, requested follow-up reports and updates to the news report are automatically searched for, reducing the burden on the user of information gathering.

[0260] "User's news browsing history" is a record of information about news articles that the user has viewed, and includes metadata such as the article title, URL, author, viewing date and time, and viewing time.

[0261] A "database" is a collection of data that systematically stores collected news browsing history and can be retrieved and analyzed later.

[0262] A "generative AI model" is a model that uses artificial intelligence technology to generate prompt sentences based on saved news viewing history and generate personalized news reports.

[0263] A "prompt sentence" is an input sentence used by a generative AI model, which is an instruction sentence to generate an appropriate news report based on the user's news browsing history.

[0264] A "personalized news report" is a personalized news summary generated based on a user's interests and concerns, including related supplemental information and follow-up reports.

[0265] A "request" is a request for information made by a user for more details about a particular news topic or related information.

[0266] A "follow-up" is additional news that includes new information or subsequent developments related to news already provided.

[0267] "Searching" is the process of finding specific information in a database or on the Internet.

[0268] "Update" refers to the reorganization of an existing news report to bring it up to date, for example by adding new information.

[0269] The "means for providing to the user" includes a method and infrastructure for notifying the generated personalized news report to the user's terminal and making it viewable.

[0270] This invention relates to a system that collects, stores, and analyzes a user's news browsing history, and generates and provides personalized news reports based on the results using a generative AI model.

[0271] 1. Collection and storage of news browsing information

[0272] server:

[0273] The server provides an API endpoint to receive news browsing information from news sites and apps, and receives metadata such as the title, URL, author, date and time of browsing, and the browsing time of the news article. The server then stores the received data in a database for each user.

[0274] The main software running on the server is a RESTful API server (e.g., Flask, Django) that provides API endpoints.

[0275] Device:

[0276] The device sets an identifier to identify the user and collects information about news articles the user has viewed on news sites and apps, and sends the collected data to the server's API endpoint.

[0277] A news application (e.g., Android or iOS app) is installed on the device and provides the user interface.

[0278] User:

[0279] Users browse news articles on news websites or apps, and the system automatically collects data based on normal news browsing behavior, so users do not need to take any special action.

[0280] 2. Analyzing news browsing information and generating reports

[0281] server:

[0282] The server periodically analyzes the news browsing information stored in the database, using natural language processing techniques (e.g., spaCy, NLTK) to identify user interests.

[0283] Based on the analysis results, a generative AI model (e.g., OpenAI GPT-3) is used to generate prompt sentences, which are then used to generate a personalized news report.

[0284] The generated personalized news report includes the user's recently viewed articles, related supplemental information, follow-up reports, and more.

[0285] Generation AI:

[0286] The generative AI takes into account the user's browsing history and interests and generates a report in natural language, using prompts such as:

[0287] Generate the following news report based on topics of interest to user 12345:

[0288] Recent news trends indicate an interest in environmental issues, so include content that informs users about the latest environmental news and research findings.

[0289] 3. Providing news reports

[0290] Device:

[0291] Once the generated news report is complete, a notification is sent to the device, using a mobile notification service (e.g., Firebase Cloud Messaging) to notify the user of the report's availability.

[0292] User:

[0293] Users can view news reports on their devices, allowing them to easily review news they have viewed in the past and follow-up reports.

[0294] 4. Requesting and Providing Follow-Up and Related Information

[0295] User:

[0296] If a user is viewing a report and wants to know more about a particular topic or related information, they can make a request.

[0297] server:

[0298] The server receives the request, uses RAG (Retrievable and Generative Model) to search the internet and other databases for relevant up-to-date information, and updates the news report based on the newly found information.

[0299] Generation AI:

[0300] The generator AI then updates the personalized news report again to include any follow-up or related information requested by the user.

[0301] Device:

[0302] Once the updated news report is ready, you will be notified again on your device.

[0303] User:

[0304] Users can view updated news reports to see the latest updates and additional information.

[0305] This system allows users to efficiently receive personalized news reports and easily access news information based on their interests. It also automatically searches for requested follow-up news and updates the reports, significantly reducing the burden on users of information gathering.

[0306] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0307] Step 1:

[0308] Collection of news browsing information

[0309] Input: News article information viewed by the user on a news site or app (title, URL, author, viewing date and time, viewing time)

[0310] What it does: When a user browses news, the device (e.g., a smartphone) collects the metadata and sends it to the server's API endpoint. The device has a news app installed, which automatically collects data every time the user reads news.

[0311] Step 2:

[0312] Saving news browsing information

[0313] Input: News article information sent to the server

[0314] Specific operation: The server stores the news article information received through the API endpoint in a database. The database management system (e.g., MySQL, PostgreSQL) organizes and stores the news data for each user.

[0315] Step 3:

[0316] Analysis of news browsing information

[0317] Input: News browsing history stored in the database

[0318] How it works: The server periodically analyzes the user's news browsing history using natural language processing techniques (e.g., spaCy, NLTK). The analysis identifies trends in topics and articles that interest the user. The server then identifies the user's areas of interest based on the content, frequency, and viewing time of articles.

[0319] Step 4:

[0320] Generate personalized news reports

[0321] Input: Analysis results (data indicating user interests)

[0322] Specific behavior: The server generates a prompt using a generative AI model (e.g., OpenAI GPT-3) and generates a personalized news report based on it. For example, the prompt could be, "Generate the following news report based on the topics that user 12345 is interested in: Recent news trends indicate that they are interested in environmental issues. Therefore, please include content that informs the user about the latest environmental news and research findings."

[0323] Step 5:

[0324] Providing news reports

[0325] Input: Generated personalized news report

[0326] Specific operation: The server sends the generated news report to the device and notifies the user via a mobile notification service (e.g., Firebase Cloud Messaging). The user can then view the report on their device.

[0327] Step 6:

[0328] Requesting and providing follow-up and related information

[0329] Input: User requests (follow-ups or related information on a particular topic)

[0330] How it works: When a user wants to know the latest updates or related information on a particular topic, they make a request from their device. The server receives the request and uses RAG (Search and Generative Model) to search the internet and other databases for relevant updates. It then uses the generative AI model again to generate an updated, personalized news report.

[0331] Step 7:

[0332] Providing updated news reports

[0333] Input: Updated personalized news report

[0334] Specific operation: The server sends the updated news report back to the device and notifies the user via the mobile notification service. The user can then view the updated news report and check the latest updates and additional information.

[0335] The above is the overall processing flow of the system, and at each step it is possible to efficiently collect and analyze news information from users and provide personalized information.

[0336] 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.

[0337] The present invention relates to a system for generating and providing news reports optimized for users by recognizing, collecting, storing, and analyzing a user's news browsing history and emotions. Hereinafter, an embodiment of the present invention will be described in detail.

[0338] 1. Collection and storage of news browsing and sentiment data

[0339] server

[0340] The server provides an API endpoint to receive news browsing information and sentiment data from news sites and apps.

[0341] It receives metadata such as the news article title, URL, author, viewing date and time, and viewing time, as well as the user's emotion data recognized by the emotion engine.

[0342] Save the received data in the database for each user.

[0343] Terminal

[0344] The device sets a user identifier and collects information about news articles the user has viewed on news sites or apps, as well as emotion data recognized by the emotion engine.

[0345] Send the collected data to the server's API endpoint.

[0346] User

[0347] Users use news sites and apps to view news articles.

[0348] Users basically operate the app in the same way as they normally would when browsing the news, but the emotion engine uses facial and emotion recognition technology to record the user's expressions and reactions.

[0349] 2. Analyzing news browsing information and sentiment data and generating reports

[0350] server

[0351] The server periodically analyzes the news browsing history and sentiment data stored in the database.

[0352] It uses natural language processing and sentiment analysis technologies to identify topics and article trends that interest users, as well as their thoughts when reading the news.

[0353] Based on the analysis results, generative AI is used to generate personalized news reports.

[0354] Generation AI

[0355] Generative AI generates reports in natural language based on the user's browsing history and sentiment.

[0356] Reports include users' recently viewed articles, related supplemental information, and adjustments to the content and tone of news based on sentiment.

[0357] Example: "Latest news on XX: Updates on last week's news and recommendations based on sentiment analysis."

[0358] 3. Providing news reports

[0359] Terminal

[0360] Once the generated news report is complete, a notification will be sent to your device.

[0361] The terminal notifies the user of the availability of the report.

[0362] User

[0363] The user receives a notification from the device and views the news report.

[0364] Review reports and quickly review previous news and follow-ups to deepen your understanding.

[0365] 4. Requesting and Providing Follow-Up and Related Information

[0366] User

[0367] While viewing a report, users can request further updates or related information on a particular topic.

[0368] server

[0369] The server receives the request and uses the RAG model to search the internet and other databases for relevant follow-up information.

[0370] Update news reports with newly discovered information.

[0371] Generation AI

[0372] The generative AI then updates the personalized news report again, adding follow-up and related information.

[0373] Terminal

[0374] Once the updated news report is ready, you will be notified again on your device.

[0375] User

[0376] Users can view updated news reports to see the latest updates and additions.

[0377] Specific examples

[0378] 1. User A:

[0379] I was browsing an article about climate change on a news app.

[0380] The emotion engine analyzes user A's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article.

[0381] 2. Terminal:

[0382] News browsing information and emotion data are sent to the server.

[0383] 3. Server:

[0384] The received data is collected and stored in a database.

[0385] The data is analyzed periodically and a sentiment engine generates personalized news reports based on the collected data.

[0386] 4. Generation AI:

[0387] A news report was created that focused on "climate change," which User A was interested in.

[0388] Reflect sentiment data and include many articles with a positive tone.

[0389] 5. Terminal:

[0390] When the news report is complete, User A is notified.

[0391] 6. User A:

[0392] Open the report to see the latest news you've read and follow-up updates.

[0393] This embodiment provides a news experience that takes into account the user's emotions and leads to greater understanding and interest.

[0394] The processing flow will be explained below.

[0395] Explanation of the specific processing of the program

[0396] ---

[0397] Step 1:

[0398] User:

[0399] A user opens a news site or app and reads a news article.

[0400] Step 2:

[0401] Device:

[0402] The device uses the user identifier to record metadata such as the title, URL, and viewing time of news articles viewed by the user.

[0403] In addition, the device's built-in emotion engine analyzes emotional data from the user's facial expressions and voice in real time, recording the type of emotion (joy, surprise, sadness, etc.) and its intensity.

[0404] Step 3:

[0405] Device:

[0406] The device makes an API request to send the recorded news viewing information and emotion data to the server in real time.

[0407] Step 4:

[0408] server:

[0409] The server processes the news browsing information and emotion data received from the terminal and stores them in a database together with the corresponding user identifier.

[0410] Step 5:

[0411] server:

[0412] Periodically (eg, weekly), the server analyzes the news browsing history and sentiment data stored in the database.

[0413] Step 6:

[0414] server:

[0415] The server analyzes browsing history and emotional data using natural language processing and sentiment analysis techniques to identify user interests.

[0416] For example, if a user consistently expresses enjoyment or interest in a particular topic, that topic will be given more weight.

[0417] Step 7:

[0418] server:

[0419] The server uses generative AI to generate personalized news reports based on the analysis results.

[0420] Report content is optimized based on user interests and emotional responses.

[0421] Step 8:

[0422] server:

[0423] The generated news reports are stored in a database and are ready to be notified to the device.

[0424] Step 9:

[0425] Device:

[0426] The terminal notifies the user that a news report is available.

[0427] Step 10:

[0428] User:

[0429] The user receives a notification from the device and views the news report.

[0430] Step 11:

[0431] User:

[0432] Users can view news reports and see selected articles and follow-up updates based on sentiment data.

[0433] Request any follow-up or related information that may be of interest to you.

[0434] Step 12:

[0435] Device:

[0436] The device makes an API request to send the request from the user to the server.

[0437] Step 13:

[0438] server:

[0439] Based on the received request, the server uses the RAG model to search the internet and other databases for relevant follow-up information.

[0440] Step 14:

[0441] server:

[0442] Based on the search results, the generating AI will update the personalized news report again, adding follow-up and related information.

[0443] Step 15:

[0444] server:

[0445] The updated news report is again stored in the database and notified to the terminal.

[0446] Step 16:

[0447] Device:

[0448] The terminal will again announce that an updated news report is available.

[0449] Step 17:

[0450] User:

[0451] Users can view updated news reports to see the latest updates and additions.

[0452] ---

[0453] Through this process, users can receive more personalized news reports that take into account their emotions when reading the news. In particular, the collection and analysis of emotional data allows the content and tone of the news to be optimized for users, eliciting deeper understanding and interest.

[0454] Example 2

[0455] 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."

[0456] Current news delivery systems achieve some degree of personalization based on users' news browsing history, but do not adequately consider users' emotions when providing news. As a result, users are unable to efficiently obtain news that matches their interests and emotions, resulting in a poor news browsing experience. Furthermore, when users request follow-up or related information on a specific topic, responses are often delayed, further worsening the user experience.

[0457] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting a user's news browsing history and user emotion data, means for storing the collected news browsing history and user emotion data in a database, means for analyzing the stored news browsing history and emotion data to generate a personalized news report, and means for providing the generated personalized news report to the user. This makes it possible to provide personalized news based on the user's emotion data, improving the quality of the news browsing experience. Furthermore, follow-up reports and related information requested by the user can be quickly provided, thereby increasing user satisfaction.

[0458] "News viewing history" is information about news articles viewed by the user, and specifically includes metadata such as the article title, URL, author, viewing date and time, and viewing time.

[0459] "Emotion data" refers to information about the emotional state recognized from the user's facial expressions and behavior, and indicates data that expresses emotions such as joy, surprise, and sadness.

[0460] "Database" refers to a computer system for systematically storing and managing collected news viewing history and emotion data.

[0461] "Analysis" refers to the process of analyzing news browsing history and emotional data stored in a database using machine learning and natural language processing technologies to identify trends in users' interests and emotions.

[0462] A "personalized news report" is a news report customized based on a user's news viewing history and emotional data, and includes content optimized for the user's interests and emotions.

[0463] "Generative AI" refers to systems that use artificial intelligence technology to generate personalized news reports, specifically natural language processing models (e.g., GPT-4).

[0464] A "follow-up" is any new information provided in relation to an already reported news story.

[0465] "Related information" refers to supplementary information or topics related to the news article the user viewed.

[0466] "Server" refers to the central computer system that collects, stores, and analyzes users' news browsing history and sentiment data to generate and provide personalized news reports.

[0467] A "terminal" is a device that a user uses to view news articles, and includes a smartphone, tablet, PC, etc.

[0468] The present invention relates to a system for collecting, storing, and analyzing a user's news browsing history and emotion data, and generating and providing a news report optimized for the user. Hereinafter, an embodiment of the invention will be described in detail.

[0469] Collection and storage of news browsing and sentiment data

[0470] server

[0471] The server provides an API endpoint for receiving news browsing information and emotion data from news sites and apps. Specifically, it receives metadata such as the news article title, URL, author, viewing date and time, and viewing time, as well as user emotion data recognized by the emotion engine. The received data is stored in a database for each user. This database is built using, for example, MySQL or MongoDB.

[0472] Terminal

[0473] The device sets a user identifier and collects information about news articles the user has viewed on news sites and apps, as well as emotion data recognized by an emotion engine. The collected data is periodically sent to the server's API endpoint. Emotion data is collected using the device's camera and an emotion analysis engine (e.g., Face API).

[0474] User

[0475] Users browse news articles on news websites or apps. The user's operations are basically the same as when browsing the news normally, but the emotion engine uses facial and emotion recognition technology to record the user's expressions and reactions.

[0476] Analyzing news browsing and sentiment data and generating reports

[0477] server

[0478] The server periodically analyzes the news browsing history and sentiment data stored in the database. It utilizes natural language processing (NLTK and spaCy) and sentiment analysis (IBM Watson and Microsoft Azure sentiment analysis APIs) to identify topics and article trends that interest users, as well as their thoughts while browsing the news. Based on the analysis results, a generative AI model (e.g., GPT-4) is used to generate personalized news reports.

[0479] Generation AI

[0480] Based on a user's browsing history and sentiment, the generative AI generates a natural language report that includes the user's recently viewed articles, relevant supplemental information, and tailoring the content and tone of the news based on sentiment.

[0481] Prompt Sentence Examples

[0482] "Based on User A's recent news about climate change, create a summary of articles with a positive tone that reflects surprising sentiment data."

[0483] Providing news reports

[0484] Terminal

[0485] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability, ensuring that it remains accessible to the user.

[0486] User

[0487] Users receive notifications on their devices to view news reports, and can review the reports to gain more detailed information about previously viewed news stories and follow-ups.

[0488] Requesting and providing follow-up and related information

[0489] User

[0490] While viewing a report, users can request follow-up or related information on a specific topic. This request is made through the News app.

[0491] server

[0492] The server receives the request and uses the Retrieval-Augmented Generation (RAG) model to search the Internet and other databases for relevant follow-up news, updating the news report with the newly discovered information.

[0493] Generation AI

[0494] The generative AI then updates the personalized news report again, adding follow-up and related information.

[0495] Terminal

[0496] Once the updated news report is complete, the user will receive another notification on their device, and they can view the updated news report to see any updates or additional information.

[0497] Specific examples

[0498] 1. User A:

[0499] When a user reads an article about climate change in a news app, the emotion engine analyzes the user's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article.

[0500] 2. Terminal:

[0501] News browsing information and emotion data are sent to a server.

[0502] 3. Server:

[0503] The received data is collected and stored in a database. The data is periodically analyzed and a sentiment engine generates personalized news reports based on the collected data.

[0504] 4. Generation AI:

[0505] Create a news report that focuses on "climate change," which User A is interested in. Reflect the sentiment data and include many articles with a positive tone.

[0506] 5. Terminal:

[0507] When the news report is complete, User A is notified.

[0508] 6. User A:

[0509] Open the report to see the news you recently read and follow-ups. This embodiment provides a news experience that takes into account the user's emotions and provides a deeper understanding and interest.

[0510] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0511] Step 1: User Views News Article

[0512] A user browses a news article using a news website or app. Specifically, the user opens a news app and selects and reads an article about climate change.

[0513] Input: Select the article you want to view on the news app

[0514] Output: Transmission of viewed article information

[0515] Step 2: Collecting news browsing information and emotion data from the device

[0516] The device automatically collects metadata (title, URL, author, date and time of viewing, duration, etc.) of news articles viewed by the user, using the data collection function within the news app.

[0517] The device uses its built-in camera and sensors to collect the user's facial expressions and emotional data (happiness, surprise, sadness, etc.) through an emotion engine (e.g., Face API).

[0518] Input: Browsing information in news apps, emotion data from cameras and sensors

[0519] Output: A batch of collected news metadata and sentiment data

[0520] Step 3: Send data from the device to the server

[0521] The device sends the collected news browsing information and sentiment data to a server through a pre-configured API endpoint, using an internet connection for this communication.

[0522] Input: Batches of news metadata and sentiment data collected on the device

[0523] Output: Confirmation of data transmission to the server

[0524] Step 4: Server saves data

[0525] The server organizes the received news browsing information and emotion data for each user and stores them in a database (e.g., MySQL or MongoDB) using a database management system.

[0526] Input: A batch of news metadata and sentiment data sent from the device

[0527] Output: Save confirmation message

[0528] Step 5: Periodic analysis of data by the server

[0529] The server periodically analyzes the news browsing information and sentiment data in the database, using natural language processing technologies (e.g., NLTK, spaCy) and sentiment analysis technologies (e.g., IBM Watson, Microsoft Azure sentiment analysis API) to identify trends in user interests and sentiment.

[0530] Input: News metadata and sentiment data in a database

[0531] Output: Analysis result report

[0532] Step 6: Generative AI creates personalized news reports

[0533] Based on the analysis results, the server sends prompts to a generative AI (e.g., GPT-4) to generate a personalized news report, which is written in natural language based on the user's browsing history and emotions.

[0534] Input: Analysis result report and prompt statement

[0535] Output: Personalized news report

[0536] Prompt Sentence Examples

[0537] "Based on User A's recent news about climate change, create a summary of articles with a positive tone that reflects surprising sentiment data."

[0538] Step 7: News report notification on your device

[0539] The server sends the generated news report to the terminal, which displays a notification to the user that a new news report is available.

[0540] Input: Personalized News Report

[0541] Output: Notification message to terminal

[0542] Step 8: User Views News Report

[0543] Users receive notifications from their devices and open and view news reports, which include recently viewed stories, follow-ups, and tone adjustments based on emotion.

[0544] Input: Notifications from your device

[0545] Output: Content of the news report

[0546] Step 9: User requests follow-up or related information

[0547] While viewing a news report, users interact with an interface to request updates or related information on a particular topic.

[0548] Input: Request button in news report

[0549] Output: Request information

[0550] Step 10: Server searches for relevant information and updates the report

[0551] The server receives user requests, uses the RAG (Retrieval-Augmented Generation) model to search the Internet and other databases for related or follow-up information, and updates the news report based on the newly found information.

[0552] Input: Request information from the user

[0553] Output: Updated news report

[0554] Step 11: Regenerating the news report with generative AI

[0555] The generative AI generates a personalized news report again based on the new information.

[0556] Input: List of newly found information

[0557] Output: Updated personalized news report

[0558] Step 12: Update report notification to device

[0559] The server retransmits the updated news report to the terminal, which displays a notification to the user that the report has been updated.

[0560] Input: Updated news report

[0561] Output: Notification message to terminal

[0562] Step 13: User Views Update News Report

[0563] Users can view updated news reports to see the latest updates and additions.

[0564] Input: Notifications from your device

[0565] Output: Updated news report content

[0566] Through these steps, the system can personalize users' news browsing experience and provide optimal news reports that reflect emotional data.

[0567] (Application example 2)

[0568] 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."

[0569] Conventional news delivery systems personalized news based on users' browsing history, but it was difficult to consider users' emotional responses. This made it difficult to provide optimal news reports that would deepen users' interest and understanding. Furthermore, there was a lack of mechanisms for generating reports that accurately reflected users' emotions toward specific news articles. To address these issues, it is necessary to collect and analyze not only news browsing history but also user emotional data during browsing to enrich the user experience. Furthermore, there is a need for more intuitive, real-time data collection and reporting using wearable devices such as smart glasses.

[0570] 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.

[0571] In this invention, the server includes means for collecting a user's news viewing history and emotional data, means for storing the collected news viewing history and emotional data in a database, means for analyzing the stored news viewing history and emotional data to generate a personalized news report, means for providing the generated personalized news report to the user, means for collecting the user's news viewing and emotional data in real time through the smart glasses, means for analyzing the user's emotional data, and means for providing the user with a function for viewing the news report using voice commands or gestures. This enables the generation and provision of more personalized news reports based on the user's emotional responses. Furthermore, the user's daily data can be collected naturally through the smart glasses, allowing the user to intuitively view the news report, thereby improving the user experience.

[0572] "News browsing history" is a collection of metadata such as the title, URL, author, browsing date and time, and browsing time of news articles that a user has viewed.

[0573] "Emotion data" is information on emotions that are recorded and recognized by the emotion engine while the user is browsing the news.

[0574] A "database" is a collection that stores collected news viewing history and emotional data and manages it so that it can be searched and analyzed as needed.

[0575] A "personalized news report" is a news summary optimized for a user based on the user's news viewing history and emotional data.

[0576] "Smart glasses" are wearable devices equipped with functions such as a camera and a display, and can recognize the user's gaze direction and facial expressions and display information.

[0577] "Means for collecting data in real time" refers to a method for capturing the user's facial expressions and reactions while viewing the news in real time through the camera in the smart glasses and instantly transmitting them to a server.

[0578] "Ability to browse using voice commands and gestures" refers to an interface that allows a user to navigate and browse news reports using voice or body movements.

[0579] The "means for analyzing user emotional data" is a technology that uses collected emotional data to evaluate the emotions felt by users while browsing news and changes in those emotions.

[0580] This invention is a system that collects, stores, and analyzes users' news browsing history and emotional data to generate and provide personalized news reports.

[0581] 1. Collection and storage of news browsing and sentiment data

[0582] server

[0583] The server provides an API endpoint for receiving news browsing information and emotion data from news sites and applications. It receives metadata such as the title, URL, author, viewing date and time of news articles, as well as user emotion data detected by the emotion engine, and stores this data in a database for each user.

[0584] Terminal

[0585] The device collects information about news articles viewed by users on news sites and apps through the smart glasses, as well as emotional data recognized by the emotion engine, and sends this data to the server's API endpoint.

[0586] User

[0587] Users use smart glasses to browse news articles. While browsing, the emotion engine uses facial and emotion recognition technology to record the user's facial expressions and reactions. The data is then sent to a server in real time.

[0588] 2. Analyzing news browsing information and sentiment data and generating reports

[0589] server

[0590] The server periodically analyzes the news browsing history and sentiment data stored in the database to identify trends in topics and articles that users are interested in. This is done using natural language processing and sentiment analysis technologies, and uses generative AI to generate personalized news reports.

[0591] Generation AI

[0592] Based on the user's browsing history and emotional data, the generative AI creates a news report that can be viewed in-car or via gesture control, including recently viewed articles, relevant supplemental information, and an adjustment to the tone of the news article based on the emotional data.

[0593] 3. Providing news reports

[0594] Terminal

[0595] Once the generated news report is complete, it notifies the user through the smart glasses, who can then use voice commands or gestures to view or interact with the report.

[0596] User

[0597] Users receive notifications, access personalized news reports, and easily review previously viewed news stories and follow-up updates. If they want more updates or related information on a particular topic, they can request that information.

[0598] 4. Requesting and Providing Follow-Up and Related Information

[0599] User

[0600] While viewing a report, users can request follow-up or related information on a particular topic.

[0601] server

[0602] The server receives the request and uses the Retrieval-Augmented Generation (RAG) model to search for relevant follow-up information from the Internet and other databases, updating the news report based on the information found.

[0603] Generation AI

[0604] The generative AI is then personalized again to update the news report, adding any necessary follow-up or related information.

[0605] Terminal

[0606] Once the updated news report is ready, another notification will be sent to the smart glasses.

[0607] User

[0608] Users can view updated news reports to see the latest updates and additions.

[0609] Example 1

[0610] A user browses an article about climate change in a news app. The emotion engine analyzes the user's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article. The news browsing information and emotional data are sent to the server via the smart glasses.

[0611] Prompt Sentence Examples

[0612] List of news recently viewed by user 123 (including metadata)

[0613] In addition, based on the emotional data (Happy, Sad, Angry, etc.) when viewing each news article,

[0614] Generate personalized news reports.

[0615] The report includes a summary of your recently viewed articles along with information on topics of interest.

[0616] Provide a list of recommended articles.

[0617] If possible, frame your news in a positive tone.

[0618] This format allows for a deeper, more engaging news experience that takes into account the user's emotions.

[0619] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0620] Step 1:

[0621] The device collects the user's news viewing information and emotion data through the smart glasses. The input is real-time video footage of the news being viewed and the metadata of the news article. The video footage captured by the smart glasses' camera is input to a face recognition and emotion recognition engine to analyze the user's emotion (e.g., surprise, interest, etc.). The output is emotion data corresponding to the metadata of the news article.

[0622] Step 2:

[0623] The device sends the collected news browsing information and emotion data to the server. The input is the news article metadata and emotion data collected in step 1. This is sent as an HTTP request to the server's API endpoint. The output is a notification of successful data transmission to the server.

[0624] Step 3:

[0625] The server stores the received news browsing information and emotion data in a database. The input is the metadata and emotion data of the news article received as an HTTP request. These are stored in the database as records corresponding to each user. The output is a confirmation notification of the record stored in the database.

[0626] Step 4:

[0627] The server periodically analyzes the stored news browsing history and sentiment data. The input is the metadata and sentiment data of news articles stored in the database. A data analysis engine is used to identify user interests and sentiment trends and create the basis for generating personalized news reports. The output is the analysis results data.

[0628] Step 5:

[0629] The server uses a generative AI to generate a personalized news report. The input is the analysis result data obtained in step 4. The generative AI uses natural language processing technology to create a news report based on the user's interests and emotions. The output is a completed personalized news report.

[0630] Step 6:

[0631] The server notifies the terminal of the generated news report. The input is the completed personalized news report, which is sent to the terminal and notified to the user. The output is a notification display from the terminal.

[0632] Step 7:

[0633] The device allows users to browse news reports using voice commands or gestures. The input is a personalized news report sent from the server and the user's voice or gesture instructions. The device displays the news report through the interface of the smart glasses, and displays and changes the content based on the user's actions. The output is the content viewed or manipulated by the user.

[0634] Step 8:

[0635] While viewing a report, a user requests follow-up or related information on a specific topic. The input is a request made by the user through voice commands or gestures. The device receives this request and sends it to the server. The output is a notification of the request to the server.

[0636] Step 9:

[0637] The server receives requests and searches for relevant follow-up information. The input is a request from a user. The RAG model (Retrieval-Augmented Generation model) is used to search for relevant information from the Internet and other databases. The output is the relevant information found.

[0638] Step 10:

[0639] The server updates the news report using a generative AI. The input is the relevant information found and the existing news report. The generative AI adds the relevant information and creates a personalized news report again. The output is the updated news report.

[0640] Step 11:

[0641] The server notifies the terminal of the updated news report. The input is the updated news report. It is sent to the terminal and notified to the user. The output is an update notification display from the terminal.

[0642] Step 12:

[0643] The user browses the updated news report to check the latest updates and additional information. The input is the user's operation upon receiving the update notification. The user checks the news report through the smart glasses and obtains the necessary information. The output is the latest information obtained by the user.

[0644] 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.

[0645] 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.

[0646] 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.

[0647] [Second embodiment]

[0648] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0649] 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.

[0650] 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).

[0651] 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.

[0652] 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.

[0653] 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).

[0654] 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.

[0655] 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.

[0656] 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.

[0657] 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.

[0658] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0659] 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."

[0660] ---

[0661] The present invention relates to a system for collecting, storing, and analyzing a user's news browsing history, and generating and providing personalized news reports based on the results. The present invention is implemented by combining a specific software system and infrastructure.

[0662] 1. Collection and storage of news browsing information

[0663] server:

[0664] The server provides an API endpoint for receiving news browsing information from news sites and apps.

[0665] Receive metadata such as news article title, URL, author, date and time of access, and access duration.

[0666] Save the received data in the database for each user.

[0667] Device:

[0668] The device sets a user identifier and collects information about news articles the user has viewed on news sites and apps.

[0669] Send the collected data to the server's API endpoint.

[0670] User:

[0671] Users use news sites and apps to view news articles.

[0672] User operations are basically the same as when browsing news normally, so no special operations are required.

[0673] 2. Analyzing news browsing information and generating reports

[0674] server:

[0675] The server periodically analyzes the news browsing information stored in the database.

[0676] Leverage natural language processing techniques to identify trends in topics and articles that interest users.

[0677] Based on the analysis, generative AI is used to generate personalized news reports, including the user's recently viewed articles, related supplemental information, and follow-up reports.

[0678] Generation AI:

[0679] The generative AI takes into account the user's browsing history and interests and generates reports in natural language.

[0680] Example: "Updates on X: Updates and insights on last week's news story."

[0681] 3. Providing news reports

[0682] Device:

[0683] Once the generated news report is complete, a notification will be sent to your device.

[0684] The terminal notifies the user of the availability of the report.

[0685] User:

[0686] The user views the news report through the terminal.

[0687] Review reports and quickly review previous news and follow-ups to deepen your understanding.

[0688] 4. Requesting and Providing Follow-Up and Related Information

[0689] User:

[0690] While viewing a report, users can request further updates or related information on a particular topic.

[0691] server:

[0692] The server receives the request and uses RAG (Retrievable and Generative Model) to search for relevant up-to-date information on the Internet and other databases.

[0693] Update news reports with newly discovered information.

[0694] Generation AI:

[0695] The generator AI then updates the personalized news report again to include any follow-up or related information requested by the user.

[0696] Device:

[0697] Once the updated news report is ready, you will be notified again on your device.

[0698] User:

[0699] Users can view updated news reports to see the latest updates and additional information.

[0700] This system not only helps users efficiently track and deeply understand the news even in their busy daily lives, but also contributes to improving media literacy in society as a whole by providing reliable sources of information.

[0701] The processing flow will be explained below.

[0702] Explanation of the specific processing of the program

[0703] ---

[0704] Step 1:

[0705] User:

[0706] A user opens a news site or app and reads a news article.

[0707] Step 2:

[0708] Device:

[0709] The device uses the user identifier to record metadata such as the title, URL, and viewing time of news articles viewed by the user.

[0710] Step 3:

[0711] Device:

[0712] The device makes an API request to send the recorded news viewing information to the server in real time.

[0713] Step 4:

[0714] server:

[0715] The server processes the news browsing information received from the terminal and stores it in a database together with the corresponding user identifier.

[0716] Step 5:

[0717] server:

[0718] Periodically (eg, weekly), the server analyzes the news browsing history stored in the database.

[0719] Step 6:

[0720] server:

[0721] The server analyzes the browsing history using natural language processing techniques to identify the user's interests.

[0722] Step 7:

[0723] server:

[0724] The server uses generative AI to generate personalized news reports based on the analysis results.

[0725] Step 8:

[0726] server:

[0727] The generated news reports are stored in a database and are ready to be notified to the device.

[0728] Step 9:

[0729] Device:

[0730] The terminal notifies the user that a news report is available.

[0731] Step 10:

[0732] User:

[0733] The user receives a notification from the device and views the news report.

[0734] Step 11:

[0735] User:

[0736] Users review news reports and request follow-up or related information of interest.

[0737] Step 12:

[0738] Device:

[0739] The device makes an API request to send the request from the user to the server.

[0740] Step 13:

[0741] server:

[0742] Based on the received request, the server uses the RAG model to search the internet and other databases for relevant follow-up information.

[0743] Step 14:

[0744] server:

[0745] Based on the search results, the generating AI will update the personalized news report again, adding follow-up and related information.

[0746] Step 15:

[0747] server:

[0748] The updated news report is again stored in the database and notified to the terminal.

[0749] Step 16:

[0750] Device:

[0751] The terminal will again announce that an updated news report is available.

[0752] Step 17:

[0753] User:

[0754] Users can view updated news reports to see the latest updates and additions.

[0755] ---

[0756] Through this process, a system is provided that allows users to efficiently track and deeply understand news even in their busy daily lives.

[0757] Example 1

[0758] 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."

[0759] Conventional news article delivery systems lacked sufficient personalization functionality that took into account a user's news browsing history, making it difficult to accurately provide information that users were interested in. Furthermore, there was a lack of a way for users to easily request and obtain follow-up reports or related information on specific news stories. This made it difficult for users to quickly and efficiently obtain the latest and most relevant information.

[0760] 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.

[0761] In this invention, the server includes means for collecting news browsing information, means for storing the collected news browsing information in a database, means for analyzing the stored news browsing information using natural language processing technology and generating a personalized news report using a generative artificial intelligence model, and means for providing the generated personalized news report to a user. This makes it possible to provide a personalized news report that takes into account the user's news browsing history, and to quickly obtain and provide related information in response to a user's request.

[0762] "News viewing information" is viewing data when a user views a news article, and specifically includes metadata such as the news article title, URL, author, viewing date and time, and viewing time.

[0763] A "database" is an information system for efficiently storing, organizing, and retrieving large amounts of data.

[0764] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and generate human language, and includes classification, extraction, and text analysis.

[0765] A "generative artificial intelligence model" is an artificial intelligence technology that learns from large datasets and automatically generates text that is meaningful to users.

[0766] A "personalized news report" is a customized news summary generated based on a user's news browsing history and tailored to their individual interests and preferences.

[0767] A "server" is a computer system that provides services to other computers (clients) on a network.

[0768] A "request" is an action by a user requesting specific information or service, or the content of that request.

[0769] "Searching" is the act of locating information from a database or the Internet based on specified criteria.

[0770] "Related information" is additional data or insight that is relevant to the user's interests or the information they are currently viewing.

[0771] MODE FOR CARRYING OUT THE INVENTION

[0772] The present invention relates to a system for collecting, storing, and analyzing news browsing information, generating personalized news reports based on the collected information, and providing the reports to users. The present invention is implemented by combining a specific software system and infrastructure.

[0773] Collection and storage of news browsing information

[0774] server:

[0775] The server provides an API endpoint for receiving news viewing information from news sites and apps. It receives metadata such as the news article title, URL, author, viewing date and time, and viewing time. The received data is stored for each user ID in the "news_views" table in the MySQL database.

[0776] Device:

[0777] The device collects news articles that users have viewed through news sites and apps. The device generates a unique identifier (user ID) for each user and sends it to the server's API endpoint along with news viewing information.

[0778] User:

[0779] Users browse news articles on a daily basis using news websites and apps, with no special operations required.

[0780] News browsing analysis and report generation

[0781] server:

[0782] The server periodically queries the news viewing information stored in the "news_views" table to obtain the user's browsing history. It analyzes the obtained data and extracts article topics and keywords using a natural language processing library (e.g., SpaCy) that identifies topics and article trends that each user is interested in. Based on the analysis results, it generates prompts for a generative artificial intelligence model (e.g., OpenAI GPT-3.5) to create a personalized news report.

[0783] Generative AI models:

[0784] The generative AI model generates a news report in natural language based on the server's prompts and matches the user's interests.

[0785] Example: "Updates on X: Updates and insights on last week's news story."

[0786] Providing news reports

[0787] Device:

[0788] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability.

[0789] User:

[0790] Users receive notifications on their devices, view news reports, and check the report content to stay up to date on related news and updates.

[0791] Requesting and providing follow-up and related information

[0792] User:

[0793] While viewing a news report, users can request updates or additional information on a particular topic.

[0794] server:

[0795] The server receives user requests and performs searches based on the requests, using the RAG model (e.g., Haystack) to gather relevant, up-to-date information from the Internet and other databases.

[0796] Example prompt: "Create a personalized news report containing the latest news that may be of interest to the user based on their recent browsing history. Specifically, detail new information and related updates about X, as well as future projections."

[0797] Generative AI models:

[0798] The generative AI model then uses the additional information collected to generate an updated news report for the user.

[0799] Device:

[0800] Once the updated news report is ready, you will be notified again on your device.

[0801] User:

[0802] Users receive notifications, view updated news reports, and stay up to date with new developments and relevant information.

[0803] The system allows users to efficiently receive personalized news reports based on their news browsing history, and they can request additional information on specific topics to stay up to date.

[0804] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0805] Processing Steps

[0806] Step 1:

[0807] server:

[0808] The server provides an API endpoint for receiving news browsing information, including metadata such as the news article title, URL, author, browsing date and time, and browsing duration.

[0809] Input: News browsing information sent from your device.

[0810] Data processing: Analyze the received metadata and convert it into a format that can be saved in the database for each user ID.

[0811] Output: Organized metadata stored in a database.

[0812] Step 2:

[0813] Device:

[0814] The device collects news articles that users have viewed through news sites and apps. The device generates a unique identifier (user ID) for each user and sends it to the server's API endpoint along with news viewing information.

[0815] Input: News article information obtained from news sites and apps.

[0816] Data processing: A user ID is assigned and the news article metadata is converted into a format that can be sent to the server.

[0817] Output: News browsing information sent to the server's API endpoint.

[0818] Step 3:

[0819] User:

[0820] Users browse news articles on a daily basis using news websites and apps, with no special operations required.

[0821] Input: User's news article browsing behavior.

[0822] Data processing: None.

[0823] Output: News browsing data based on user browsing behavior.

[0824] Step 4:

[0825] server:

[0826] The server periodically queries the news browsing information stored in the database to obtain the user's browsing history, analyzes the obtained data, and identifies topics and article trends that each user is interested in. It then uses a natural language processing library (e.g., SpaCy) to extract article topics and keywords.

[0827] Input: News browsing information stored in a database.

[0828] Data processing: Using natural language processing technology, topics and keywords from news articles are extracted and user interests are analyzed.

[0829] Output: Analysis results that reflect each user's interests and concerns.

[0830] Step 5:

[0831] server:

[0832] Based on the analysis results, the server generates prompt sentences for a generative AI model (e.g., OpenAI GPT-3.5) to create a personalized news report.

[0833] Input: Results analyzed using natural language processing and user browsing history.

[0834] Data processing: Generate prompt sentences and feed them into a generative AI model to generate personalized news reports.

[0835] Output: The generated personalized news report.

[0836] Action: Example prompt: "Create a personalized news report containing the latest news that may be of interest to the user based on their recent browsing history. Specifically, include new information and related updates about ____, as well as future projections."

[0837] Step 6:

[0838] Device:

[0839] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability.

[0840] Input: A news report sent from the server.

[0841] Data processing: Convert to notification format.

[0842] Output: Notify the user and display a news report.

[0843] Step 7:

[0844] User:

[0845] Users receive notifications on their devices, view news reports, and check the report content to stay up to date on related news and updates.

[0846] Input: A news report posted to the device.

[0847] Data processing: None.

[0848] Output: Confirmed news report viewing behavior of the user.

[0849] Step 8:

[0850] User:

[0851] While viewing a news report, users can request updates or additional information on a particular topic.

[0852] Input: A user request for a news report.

[0853] Data processing: None.

[0854] Output: Request sent.

[0855] Step 9:

[0856] server:

[0857] The server receives user requests and performs searches based on the requests, using the RAG model (e.g., Haystack) to gather relevant, up-to-date information from the Internet and other databases.

[0858] Input: The user's request and the database to search.

[0859] Data processing: Search and retrieve relevant information using the RAG model.

[0860] Output: The most recent relevant information retrieved.

[0861] Step 10:

[0862] Generative AI models:

[0863] The generative AI model then uses the additional information collected to generate an updated news report for the user.

[0864] Input: retrieved related information, previous browsing history, user interests.

[0865] Data processing: Generate prompt sentences and feed them into a generative AI model to generate updated news reports.

[0866] Output: An updated personalized news report.

[0867] Step 11:

[0868] Device:

[0869] Once the updated news report is ready, you will be notified again on your device.

[0870] Input: Updated news report sent from the server.

[0871] Data processing: Convert to notification format.

[0872] Output: Notify user and display update news report.

[0873] Step 12:

[0874] User:

[0875] Users receive notifications, view updated news reports, and stay up to date with new developments and relevant information.

[0876] Input: Updated news report notified to terminal.

[0877] Data processing: None.

[0878] Output: User's viewing behavior for latest information.

[0879] (Application example 1)

[0880] 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."

[0881] In modern society, a huge amount of news information is provided via the Internet, but users face the challenge of efficiently gathering information relevant to them. In particular, the lack of personalized news reports tailored to individual users makes it difficult for users to access news information based on their interests. Furthermore, when users want more information on a particular topic, they have to search manually, which is inconvenient.

[0882] 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.

[0883] In this invention, the server includes means for collecting a user's news browsing history, means for storing the collected news browsing history in a database, means for analyzing the stored news browsing history and generating a prompt sentence using a generative AI model to generate a personalized news report, and means for providing the generated personalized news report to the user. This effectively provides a personalized news report to the user, allowing the user to easily access news information based on their interests. In addition, requested follow-up reports and updates to the news report are automatically searched for, reducing the burden on the user of information gathering.

[0884] "User's news browsing history" is a record of information about news articles that the user has viewed, and includes metadata such as the article title, URL, author, viewing date and time, and viewing time.

[0885] A "database" is a collection of data that systematically stores collected news browsing history and can be retrieved and analyzed later.

[0886] A "generative AI model" is a model that uses artificial intelligence technology to generate prompt sentences based on saved news viewing history and generate personalized news reports.

[0887] A "prompt sentence" is an input sentence used by a generative AI model, which is an instruction sentence to generate an appropriate news report based on the user's news browsing history.

[0888] A "personalized news report" is a personalized news summary generated based on a user's interests and concerns, including related supplemental information and follow-up reports.

[0889] A "request" is a request for information made by a user for more details about a particular news topic or related information.

[0890] A "follow-up" is additional news that includes new information or subsequent developments related to news already provided.

[0891] "Searching" is the process of finding specific information in a database or on the Internet.

[0892] "Update" refers to the reorganization of an existing news report to bring it up to date, for example by adding new information.

[0893] The "means for providing to the user" includes a method and infrastructure for notifying the generated personalized news report to the user's terminal and making it viewable.

[0894] This invention relates to a system that collects, stores, and analyzes a user's news browsing history, and generates and provides personalized news reports based on the results using a generative AI model.

[0895] 1. Collection and storage of news browsing information

[0896] server:

[0897] The server provides an API endpoint to receive news browsing information from news sites and apps, and receives metadata such as the title, URL, author, date and time of browsing, and the browsing time of the news article. The server then stores the received data in a database for each user.

[0898] The main software running on the server is a RESTful API server (e.g., Flask, Django) that provides API endpoints.

[0899] Device:

[0900] The device sets an identifier to identify the user and collects information about news articles the user has viewed on news sites and apps, and sends the collected data to the server's API endpoint.

[0901] A news application (e.g., Android or iOS app) is installed on the device and provides the user interface.

[0902] User:

[0903] Users browse news articles on news websites or apps, and the system automatically collects data based on normal news browsing behavior, so users do not need to take any special action.

[0904] 2. Analyzing news browsing information and generating reports

[0905] server:

[0906] The server periodically analyzes the news browsing information stored in the database, using natural language processing techniques (e.g., spaCy, NLTK) to identify user interests.

[0907] Based on the analysis results, a generative AI model (e.g., OpenAI GPT-3) is used to generate prompt sentences, which are then used to generate a personalized news report.

[0908] The generated personalized news report includes the user's recently viewed articles, related supplemental information, follow-up reports, and more.

[0909] Generation AI:

[0910] The generative AI takes into account the user's browsing history and interests and generates a report in natural language, using prompts such as:

[0911] Generate the following news report based on topics of interest to user 12345:

[0912] Recent news trends indicate an interest in environmental issues, so include content that informs users about the latest environmental news and research findings.

[0913] 3. Providing news reports

[0914] Device:

[0915] Once the generated news report is complete, a notification is sent to the device, using a mobile notification service (e.g., Firebase Cloud Messaging) to notify the user of the report's availability.

[0916] User:

[0917] Users can view news reports on their devices, allowing them to easily review news they have viewed in the past and follow-up reports.

[0918] 4. Requesting and Providing Follow-Up and Related Information

[0919] User:

[0920] If a user is viewing a report and wants to know more about a particular topic or related information, they can make a request.

[0921] server:

[0922] The server receives the request, uses RAG (Retrievable and Generative Model) to search the internet and other databases for relevant up-to-date information, and updates the news report based on the newly found information.

[0923] Generation AI:

[0924] The generator AI then updates the personalized news report again to include any follow-up or related information requested by the user.

[0925] Device:

[0926] Once the updated news report is ready, you will be notified again on your device.

[0927] User:

[0928] Users can view updated news reports to see the latest updates and additional information.

[0929] This system allows users to efficiently receive personalized news reports and easily access news information based on their interests. It also automatically searches for requested follow-up news and updates the reports, significantly reducing the burden on users of information gathering.

[0930] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0931] Step 1:

[0932] Collection of news browsing information

[0933] Input: News article information viewed by the user on a news site or app (title, URL, author, viewing date and time, viewing time)

[0934] What it does: When a user browses news, the device (e.g., a smartphone) collects the metadata and sends it to the server's API endpoint. The device has a news app installed, which automatically collects data every time the user reads news.

[0935] Step 2:

[0936] Saving news browsing information

[0937] Input: News article information sent to the server

[0938] Specific operation: The server stores the news article information received through the API endpoint in a database. The database management system (e.g., MySQL, PostgreSQL) organizes and stores the news data for each user.

[0939] Step 3:

[0940] Analysis of news browsing information

[0941] Input: News browsing history stored in the database

[0942] How it works: The server periodically analyzes the user's news browsing history using natural language processing techniques (e.g., spaCy, NLTK). The analysis identifies trends in topics and articles that interest the user. The server then identifies the user's areas of interest based on the content, frequency, and viewing time of articles.

[0943] Step 4:

[0944] Generate personalized news reports

[0945] Input: Analysis results (data indicating user interests)

[0946] Specific behavior: The server generates a prompt using a generative AI model (e.g., OpenAI GPT-3) and generates a personalized news report based on it. For example, the prompt could be, "Generate the following news report based on the topics that user 12345 is interested in: Recent news trends indicate that they are interested in environmental issues. Therefore, please include content that informs the user about the latest environmental news and research findings."

[0947] Step 5:

[0948] Providing news reports

[0949] Input: Generated personalized news report

[0950] Specific operation: The server sends the generated news report to the device and notifies the user via a mobile notification service (e.g., Firebase Cloud Messaging). The user can then view the report on their device.

[0951] Step 6:

[0952] Requesting and providing follow-up and related information

[0953] Input: User requests (follow-ups or related information on a particular topic)

[0954] How it works: When a user wants to know the latest updates or related information on a particular topic, they make a request from their device. The server receives the request and uses RAG (Search and Generative Model) to search the internet and other databases for relevant updates. It then uses the generative AI model again to generate an updated, personalized news report.

[0955] Step 7:

[0956] Providing updated news reports

[0957] Input: Updated personalized news report

[0958] Specific operation: The server sends the updated news report back to the device and notifies the user via the mobile notification service. The user can then view the updated news report and check the latest updates and additional information.

[0959] The above is the overall processing flow of the system, and at each step it is possible to efficiently collect and analyze news information from users and provide personalized information.

[0960] 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.

[0961] The present invention relates to a system for generating and providing news reports optimized for users by recognizing, collecting, storing, and analyzing a user's news browsing history and emotions. Hereinafter, an embodiment of the present invention will be described in detail.

[0962] 1. Collection and storage of news browsing and sentiment data

[0963] server

[0964] The server provides an API endpoint to receive news browsing information and sentiment data from news sites and apps.

[0965] It receives metadata such as the news article title, URL, author, viewing date and time, and viewing time, as well as the user's emotion data recognized by the emotion engine.

[0966] Save the received data in the database for each user.

[0967] Terminal

[0968] The device sets a user identifier and collects information about news articles the user has viewed on news sites or apps, as well as emotion data recognized by the emotion engine.

[0969] Send the collected data to the server's API endpoint.

[0970] User

[0971] Users use news sites and apps to view news articles.

[0972] Users basically operate the app in the same way as they normally would when browsing the news, but the emotion engine uses facial and emotion recognition technology to record the user's expressions and reactions.

[0973] 2. Analyzing news browsing information and sentiment data and generating reports

[0974] server

[0975] The server periodically analyzes the news browsing history and sentiment data stored in the database.

[0976] It uses natural language processing and sentiment analysis technologies to identify topics and article trends that interest users, as well as their thoughts when reading the news.

[0977] Based on the analysis results, generative AI is used to generate personalized news reports.

[0978] Generation AI

[0979] Generative AI generates reports in natural language based on the user's browsing history and sentiment.

[0980] Reports include users' recently viewed articles, related supplemental information, and adjustments to the content and tone of news based on sentiment.

[0981] Example: "Latest news on XX: Updates on last week's news and recommendations based on sentiment analysis."

[0982] 3. Providing news reports

[0983] Terminal

[0984] Once the generated news report is complete, a notification will be sent to your device.

[0985] The terminal notifies the user of the availability of the report.

[0986] User

[0987] The user receives a notification from the device and views the news report.

[0988] Review reports and quickly review previous news and follow-ups to deepen your understanding.

[0989] 4. Requesting and Providing Follow-Up and Related Information

[0990] User

[0991] While viewing a report, users can request further updates or related information on a particular topic.

[0992] server

[0993] The server receives the request and uses the RAG model to search the internet and other databases for relevant follow-up information.

[0994] Update news reports with newly discovered information.

[0995] Generation AI

[0996] The generative AI then updates the personalized news report again, adding follow-up and related information.

[0997] Terminal

[0998] Once the updated news report is ready, you will be notified again on your device.

[0999] User

[1000] Users can view updated news reports to see the latest updates and additions.

[1001] Specific examples

[1002] 1. User A:

[1003] I was browsing an article about climate change on a news app.

[1004] The emotion engine analyzes user A's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article.

[1005] 2. Terminal:

[1006] News browsing information and emotion data are sent to the server.

[1007] 3. Server:

[1008] The received data is collected and stored in a database.

[1009] The data is analyzed periodically and a sentiment engine generates personalized news reports based on the collected data.

[1010] 4. Generation AI:

[1011] A news report was created that focused on "climate change," which User A was interested in.

[1012] Reflect sentiment data and include many articles with a positive tone.

[1013] 5. Terminal:

[1014] When the news report is complete, User A is notified.

[1015] 6. User A:

[1016] Open the report to see the latest news you've read and follow-up updates.

[1017] This embodiment provides a news experience that takes into account the user's emotions and leads to greater understanding and interest.

[1018] The processing flow will be explained below.

[1019] Explanation of the specific processing of the program

[1020] ---

[1021] Step 1:

[1022] User:

[1023] A user opens a news site or app and reads a news article.

[1024] Step 2:

[1025] Device:

[1026] The device uses the user identifier to record metadata such as the title, URL, and viewing time of news articles viewed by the user.

[1027] In addition, the device's built-in emotion engine analyzes emotional data from the user's facial expressions and voice in real time, recording the type of emotion (joy, surprise, sadness, etc.) and its intensity.

[1028] Step 3:

[1029] Device:

[1030] The device makes an API request to send the recorded news viewing information and emotion data to the server in real time.

[1031] Step 4:

[1032] server:

[1033] The server processes the news browsing information and emotion data received from the terminal and stores them in a database together with the corresponding user identifier.

[1034] Step 5:

[1035] server:

[1036] Periodically (eg, weekly), the server analyzes the news browsing history and sentiment data stored in the database.

[1037] Step 6:

[1038] server:

[1039] The server analyzes browsing history and emotional data using natural language processing and sentiment analysis techniques to identify user interests.

[1040] For example, if a user consistently expresses enjoyment or interest in a particular topic, that topic will be given more weight.

[1041] Step 7:

[1042] server:

[1043] The server uses generative AI to generate personalized news reports based on the analysis results.

[1044] Report content is optimized based on user interests and emotional responses.

[1045] Step 8:

[1046] server:

[1047] The generated news reports are stored in a database and are ready to be notified to the device.

[1048] Step 9:

[1049] Device:

[1050] The terminal notifies the user that a news report is available.

[1051] Step 10:

[1052] User:

[1053] The user receives a notification from the device and views the news report.

[1054] Step 11:

[1055] User:

[1056] Users can view news reports and see selected articles and follow-up updates based on sentiment data.

[1057] Request any follow-up or related information that may be of interest to you.

[1058] Step 12:

[1059] Device:

[1060] The device makes an API request to send the request from the user to the server.

[1061] Step 13:

[1062] server:

[1063] Based on the received request, the server uses the RAG model to search the internet and other databases for relevant follow-up information.

[1064] Step 14:

[1065] server:

[1066] Based on the search results, the generating AI will update the personalized news report again, adding follow-up and related information.

[1067] Step 15:

[1068] server:

[1069] The updated news report is again stored in the database and notified to the terminal.

[1070] Step 16:

[1071] Device:

[1072] The terminal will again announce that an updated news report is available.

[1073] Step 17:

[1074] User:

[1075] Users can view updated news reports to see the latest updates and additions.

[1076] ---

[1077] Through this process, users can receive more personalized news reports that take into account their emotions when reading the news. In particular, the collection and analysis of emotional data allows the content and tone of the news to be optimized for users, eliciting deeper understanding and interest.

[1078] Example 2

[1079] 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."

[1080] Current news delivery systems achieve some degree of personalization based on users' news browsing history, but do not adequately consider users' emotions when providing news. As a result, users are unable to efficiently obtain news that matches their interests and emotions, resulting in a poor news browsing experience. Furthermore, when users request follow-up or related information on a specific topic, responses are often delayed, further worsening the user experience.

[1081] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting a user's news browsing history and user emotion data, means for storing the collected news browsing history and user emotion data in a database, means for analyzing the stored news browsing history and emotion data to generate a personalized news report, and means for providing the generated personalized news report to the user. This makes it possible to provide personalized news based on the user's emotion data, improving the quality of the news browsing experience. Furthermore, follow-up reports and related information requested by the user can be quickly provided, thereby increasing user satisfaction.

[1082] "News viewing history" is information about news articles viewed by the user, and specifically includes metadata such as the article title, URL, author, viewing date and time, and viewing time.

[1083] "Emotion data" refers to information about the emotional state recognized from the user's facial expressions and behavior, and indicates data that expresses emotions such as joy, surprise, and sadness.

[1084] "Database" refers to a computer system for systematically storing and managing collected news viewing history and emotion data.

[1085] "Analysis" refers to the process of analyzing news browsing history and emotional data stored in a database using machine learning and natural language processing technologies to identify trends in users' interests and emotions.

[1086] A "personalized news report" is a news report customized based on a user's news viewing history and emotional data, and includes content optimized for the user's interests and emotions.

[1087] "Generative AI" refers to systems that use artificial intelligence technology to generate personalized news reports, specifically natural language processing models (e.g., GPT-4).

[1088] A "follow-up" is any new information provided in relation to an already reported news story.

[1089] "Related information" refers to supplementary information or topics related to the news article the user viewed.

[1090] "Server" refers to the central computer system that collects, stores, and analyzes users' news browsing history and sentiment data to generate and provide personalized news reports.

[1091] A "terminal" is a device that a user uses to view news articles, and includes a smartphone, tablet, PC, etc.

[1092] The present invention relates to a system for collecting, storing, and analyzing a user's news browsing history and emotion data, and generating and providing a news report optimized for the user. Hereinafter, an embodiment of the invention will be described in detail.

[1093] Collection and storage of news browsing and sentiment data

[1094] server

[1095] The server provides an API endpoint for receiving news browsing information and emotion data from news sites and apps. Specifically, it receives metadata such as the news article title, URL, author, viewing date and time, and viewing time, as well as user emotion data recognized by the emotion engine. The received data is stored in a database for each user. This database is built using, for example, MySQL or MongoDB.

[1096] Terminal

[1097] The device sets a user identifier and collects information about news articles the user has viewed on news sites and apps, as well as emotion data recognized by an emotion engine. The collected data is periodically sent to the server's API endpoint. Emotion data is collected using the device's camera and an emotion analysis engine (e.g., Face API).

[1098] User

[1099] Users browse news articles on news websites or apps. The user's operations are basically the same as when browsing the news normally, but the emotion engine uses facial and emotion recognition technology to record the user's expressions and reactions.

[1100] Analyzing news browsing and sentiment data and generating reports

[1101] server

[1102] The server periodically analyzes the news browsing history and sentiment data stored in the database. It utilizes natural language processing (NLTK and spaCy) and sentiment analysis (IBM Watson and Microsoft Azure sentiment analysis APIs) to identify topics and article trends that interest users, as well as their thoughts while browsing the news. Based on the analysis results, a generative AI model (e.g., GPT-4) is used to generate personalized news reports.

[1103] Generation AI

[1104] Based on a user's browsing history and sentiment, the generative AI generates a natural language report that includes the user's recently viewed articles, relevant supplemental information, and tailoring the content and tone of the news based on sentiment.

[1105] Prompt Sentence Examples

[1106] "Based on User A's recent news about climate change, create a summary of articles with a positive tone that reflects surprising sentiment data."

[1107] Providing news reports

[1108] Terminal

[1109] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability, ensuring that it remains accessible to the user.

[1110] User

[1111] Users receive notifications on their devices to view news reports, and can review the reports to gain more detailed information about previously viewed news stories and follow-ups.

[1112] Requesting and providing follow-up and related information

[1113] User

[1114] While viewing a report, users can request follow-up or related information on a specific topic. This request is made through the News app.

[1115] server

[1116] The server receives the request and uses the Retrieval-Augmented Generation (RAG) model to search the Internet and other databases for relevant follow-up news, updating the news report with the newly discovered information.

[1117] Generation AI

[1118] The generative AI then updates the personalized news report again, adding follow-up and related information.

[1119] Terminal

[1120] Once the updated news report is complete, the user will receive another notification on their device, and they can view the updated news report to see any updates or additional information.

[1121] Specific examples

[1122] 1. User A:

[1123] When a user reads an article about climate change in a news app, the emotion engine analyzes the user's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article.

[1124] 2. Terminal:

[1125] News browsing information and emotion data are sent to a server.

[1126] 3. Server:

[1127] The received data is collected and stored in a database. The data is periodically analyzed and a sentiment engine generates personalized news reports based on the collected data.

[1128] 4. Generation AI:

[1129] Create a news report that focuses on "climate change," which User A is interested in. Reflect the sentiment data and include many articles with a positive tone.

[1130] 5. Terminal:

[1131] When the news report is complete, User A is notified.

[1132] 6. User A:

[1133] Open the report to see the news you recently read and follow-ups. This embodiment provides a news experience that takes into account the user's emotions and provides a deeper understanding and interest.

[1134] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1135] Step 1: User Views News Article

[1136] A user browses a news article using a news website or app. Specifically, the user opens a news app and selects and reads an article about climate change.

[1137] Input: Select the article you want to view on the news app

[1138] Output: Transmission of viewed article information

[1139] Step 2: Collecting news browsing information and emotion data from the device

[1140] The device automatically collects metadata (title, URL, author, date and time of viewing, duration, etc.) of news articles viewed by the user, using the data collection function within the news app.

[1141] The device uses its built-in camera and sensors to collect the user's facial expressions and emotional data (happiness, surprise, sadness, etc.) through an emotion engine (e.g., Face API).

[1142] Input: Browsing information in news apps, emotion data from cameras and sensors

[1143] Output: A batch of collected news metadata and sentiment data

[1144] Step 3: Send data from the device to the server

[1145] The device sends the collected news browsing information and sentiment data to a server through a pre-configured API endpoint, using an internet connection for this communication.

[1146] Input: Batches of news metadata and sentiment data collected on the device

[1147] Output: Confirmation of data transmission to the server

[1148] Step 4: Server saves data

[1149] The server organizes the received news browsing information and emotion data for each user and stores them in a database (e.g., MySQL or MongoDB) using a database management system.

[1150] Input: A batch of news metadata and sentiment data sent from the device

[1151] Output: Save confirmation message

[1152] Step 5: Periodic analysis of data by the server

[1153] The server periodically analyzes the news browsing information and sentiment data in the database, using natural language processing technologies (e.g., NLTK, spaCy) and sentiment analysis technologies (e.g., IBM Watson, Microsoft Azure sentiment analysis API) to identify trends in user interests and sentiment.

[1154] Input: News metadata and sentiment data in a database

[1155] Output: Analysis result report

[1156] Step 6: Generative AI creates personalized news reports

[1157] Based on the analysis results, the server sends prompts to a generative AI (e.g., GPT-4) to generate a personalized news report, which is written in natural language based on the user's browsing history and emotions.

[1158] Input: Analysis result report and prompt statement

[1159] Output: Personalized news report

[1160] Prompt Sentence Examples

[1161] "Based on User A's recent news about climate change, create a summary of articles with a positive tone that reflects surprising sentiment data."

[1162] Step 7: News report notification on your device

[1163] The server sends the generated news report to the terminal, which displays a notification to the user that a new news report is available.

[1164] Input: Personalized News Report

[1165] Output: Notification message to terminal

[1166] Step 8: User Views News Report

[1167] Users receive notifications from their devices and open and view news reports, which include recently viewed stories, follow-ups, and tone adjustments based on emotion.

[1168] Input: Notifications from your device

[1169] Output: Content of the news report

[1170] Step 9: User requests follow-up or related information

[1171] While viewing a news report, users interact with an interface to request updates or related information on a particular topic.

[1172] Input: Request button in news report

[1173] Output: Request information

[1174] Step 10: Server searches for relevant information and updates the report

[1175] The server receives user requests, uses the RAG (Retrieval-Augmented Generation) model to search the Internet and other databases for related or follow-up information, and updates the news report based on the newly found information.

[1176] Input: Request information from the user

[1177] Output: Updated news report

[1178] Step 11: Regenerating the news report with generative AI

[1179] The generative AI generates a personalized news report again based on the new information.

[1180] Input: List of newly found information

[1181] Output: Updated personalized news report

[1182] Step 12: Update report notification to device

[1183] The server retransmits the updated news report to the terminal, which displays a notification to the user that the report has been updated.

[1184] Input: Updated news report

[1185] Output: Notification message to terminal

[1186] Step 13: User Views Update News Report

[1187] Users can view updated news reports to see the latest updates and additions.

[1188] Input: Notifications from your device

[1189] Output: Updated news report content

[1190] Through these steps, the system can personalize users' news browsing experience and provide optimal news reports that reflect emotional data.

[1191] (Application example 2)

[1192] 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."

[1193] Conventional news delivery systems personalized news based on users' browsing history, but it was difficult to consider users' emotional responses. This made it difficult to provide optimal news reports that would deepen users' interest and understanding. Furthermore, there was a lack of mechanisms for generating reports that accurately reflected users' emotions toward specific news articles. To address these issues, it is necessary to collect and analyze not only news browsing history but also user emotional data during browsing to enrich the user experience. Furthermore, there is a need for more intuitive, real-time data collection and reporting using wearable devices such as smart glasses.

[1194] 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.

[1195] In this invention, the server includes means for collecting a user's news viewing history and emotional data, means for storing the collected news viewing history and emotional data in a database, means for analyzing the stored news viewing history and emotional data to generate a personalized news report, means for providing the generated personalized news report to the user, means for collecting the user's news viewing and emotional data in real time through the smart glasses, means for analyzing the user's emotional data, and means for providing the user with a function for viewing the news report using voice commands or gestures. This enables the generation and provision of more personalized news reports based on the user's emotional responses. Furthermore, the user's daily data can be collected naturally through the smart glasses, allowing the user to intuitively view the news report, thereby improving the user experience.

[1196] "News browsing history" is a collection of metadata such as the title, URL, author, browsing date and time, and browsing time of news articles that a user has viewed.

[1197] "Emotion data" is information on emotions that are recorded and recognized by the emotion engine while the user is browsing the news.

[1198] A "database" is a collection that stores collected news viewing history and emotional data and manages it so that it can be searched and analyzed as needed.

[1199] A "personalized news report" is a news summary optimized for a user based on the user's news viewing history and emotional data.

[1200] "Smart glasses" are wearable devices equipped with functions such as a camera and a display, and can recognize the user's gaze direction and facial expressions and display information.

[1201] "Means for collecting data in real time" refers to a method for capturing the user's facial expressions and reactions while viewing the news in real time through the camera in the smart glasses and instantly transmitting them to a server.

[1202] "Ability to browse using voice commands and gestures" refers to an interface that allows a user to navigate and browse news reports using voice or body movements.

[1203] The "means for analyzing user emotional data" is a technology that uses collected emotional data to evaluate the emotions felt by users while browsing news and changes in those emotions.

[1204] This invention is a system that collects, stores, and analyzes users' news browsing history and emotional data to generate and provide personalized news reports.

[1205] 1. Collection and storage of news browsing and sentiment data

[1206] server

[1207] The server provides an API endpoint for receiving news browsing information and emotion data from news sites and applications. It receives metadata such as the title, URL, author, viewing date and time of news articles, as well as user emotion data detected by the emotion engine, and stores this data in a database for each user.

[1208] Terminal

[1209] The device collects information about news articles viewed by users on news sites and apps through the smart glasses, as well as emotional data recognized by the emotion engine, and sends this data to the server's API endpoint.

[1210] User

[1211] Users use smart glasses to browse news articles. While browsing, the emotion engine uses facial and emotion recognition technology to record the user's facial expressions and reactions. The data is then sent to a server in real time.

[1212] 2. Analyzing news browsing information and sentiment data and generating reports

[1213] server

[1214] The server periodically analyzes the news browsing history and sentiment data stored in the database to identify trends in topics and articles that users are interested in. This is done using natural language processing and sentiment analysis technologies, and uses generative AI to generate personalized news reports.

[1215] Generation AI

[1216] Based on the user's browsing history and emotional data, the generative AI creates a news report that can be viewed in-car or via gesture control, including recently viewed articles, relevant supplemental information, and an adjustment to the tone of the news article based on the emotional data.

[1217] 3. Providing news reports

[1218] Terminal

[1219] Once the generated news report is complete, it notifies the user through the smart glasses, who can then use voice commands or gestures to view or interact with the report.

[1220] User

[1221] Users receive notifications, access personalized news reports, and easily review previously viewed news stories and follow-up updates. If they want more updates or related information on a particular topic, they can request that information.

[1222] 4. Requesting and Providing Follow-Up and Related Information

[1223] User

[1224] While viewing a report, users can request follow-up or related information on a particular topic.

[1225] server

[1226] The server receives the request and uses the Retrieval-Augmented Generation (RAG) model to search for relevant follow-up information from the Internet and other databases, updating the news report based on the information found.

[1227] Generation AI

[1228] The generative AI is then personalized again to update the news report, adding any necessary follow-up or related information.

[1229] Terminal

[1230] Once the updated news report is ready, another notification will be sent to the smart glasses.

[1231] User

[1232] Users can view updated news reports to see the latest updates and additions.

[1233] Example 1

[1234] A user browses an article about climate change in a news app. The emotion engine analyzes the user's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article. The news browsing information and emotional data are sent to the server via the smart glasses.

[1235] Prompt Sentence Examples

[1236] List of news recently viewed by user 123 (including metadata)

[1237] In addition, based on the emotional data (Happy, Sad, Angry, etc.) when viewing each news article,

[1238] Generate personalized news reports.

[1239] The report includes a summary of your recently viewed articles along with information on topics of interest.

[1240] Provide a list of recommended articles.

[1241] If possible, frame your news in a positive tone.

[1242] This format allows for a deeper, more engaging news experience that takes into account the user's emotions.

[1243] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1244] Step 1:

[1245] The device collects the user's news viewing information and emotion data through the smart glasses. The input is real-time video footage of the news being viewed and the metadata of the news article. The video footage captured by the smart glasses' camera is input to a face recognition and emotion recognition engine to analyze the user's emotion (e.g., surprise, interest, etc.). The output is emotion data corresponding to the metadata of the news article.

[1246] Step 2:

[1247] The device sends the collected news browsing information and emotion data to the server. The input is the news article metadata and emotion data collected in step 1. This is sent as an HTTP request to the server's API endpoint. The output is a notification of successful data transmission to the server.

[1248] Step 3:

[1249] The server stores the received news browsing information and emotion data in a database. The input is the metadata and emotion data of the news article received as an HTTP request. These are stored in the database as records corresponding to each user. The output is a confirmation notification of the record stored in the database.

[1250] Step 4:

[1251] The server periodically analyzes the stored news browsing history and sentiment data. The input is the metadata and sentiment data of news articles stored in the database. A data analysis engine is used to identify user interests and sentiment trends and create the basis for generating personalized news reports. The output is the analysis results data.

[1252] Step 5:

[1253] The server uses a generative AI to generate a personalized news report. The input is the analysis result data obtained in step 4. The generative AI uses natural language processing technology to create a news report based on the user's interests and emotions. The output is a completed personalized news report.

[1254] Step 6:

[1255] The server notifies the terminal of the generated news report. The input is the completed personalized news report, which is sent to the terminal and notified to the user. The output is a notification display from the terminal.

[1256] Step 7:

[1257] The device allows users to browse news reports using voice commands or gestures. The input is a personalized news report sent from the server and the user's voice or gesture instructions. The device displays the news report through the interface of the smart glasses, and displays and changes the content based on the user's actions. The output is the content viewed or manipulated by the user.

[1258] Step 8:

[1259] While viewing a report, a user requests follow-up or related information on a specific topic. The input is a request made by the user through voice commands or gestures. The device receives this request and sends it to the server. The output is a notification of the request to the server.

[1260] Step 9:

[1261] The server receives requests and searches for relevant follow-up information. The input is a request from a user. The RAG model (Retrieval-Augmented Generation model) is used to search for relevant information from the Internet and other databases. The output is the relevant information found.

[1262] Step 10:

[1263] The server updates the news report using a generative AI. The input is the relevant information found and the existing news report. The generative AI adds the relevant information and creates a personalized news report again. The output is the updated news report.

[1264] Step 11:

[1265] The server notifies the terminal of the updated news report. The input is the updated news report. It is sent to the terminal and notified to the user. The output is an update notification display from the terminal.

[1266] Step 12:

[1267] The user browses the updated news report to check the latest updates and additional information. The input is the user's operation upon receiving the update notification. The user checks the news report through the smart glasses and obtains the necessary information. The output is the latest information obtained by the user.

[1268] 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.

[1269] 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.

[1270] 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.

[1271] [Third embodiment]

[1272] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1273] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1274] 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).

[1275] 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.

[1276] 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.

[1277] 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).

[1278] 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.

[1279] 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.

[1280] 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.

[1281] 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.

[1282] 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.

[1283] 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."

[1284] ---

[1285] The present invention relates to a system for collecting, storing, and analyzing a user's news browsing history, and generating and providing personalized news reports based on the results. The present invention is implemented by combining a specific software system and infrastructure.

[1286] 1. Collection and storage of news browsing information

[1287] server:

[1288] The server provides an API endpoint for receiving news browsing information from news sites and apps.

[1289] Receive metadata such as news article title, URL, author, date and time of access, and access duration.

[1290] Save the received data in the database for each user.

[1291] Device:

[1292] The device sets a user identifier and collects information about news articles the user has viewed on news sites and apps.

[1293] Send the collected data to the server's API endpoint.

[1294] User:

[1295] Users use news sites and apps to view news articles.

[1296] User operations are basically the same as when browsing news normally, so no special operations are required.

[1297] 2. Analyzing news browsing information and generating reports

[1298] server:

[1299] The server periodically analyzes the news browsing information stored in the database.

[1300] Leverage natural language processing techniques to identify trends in topics and articles that interest users.

[1301] Based on the analysis, generative AI is used to generate personalized news reports, including the user's recently viewed articles, related supplemental information, and follow-up reports.

[1302] Generation AI:

[1303] The generative AI takes into account the user's browsing history and interests and generates reports in natural language.

[1304] Example: "Updates on X: Updates and insights on last week's news story."

[1305] 3. Providing news reports

[1306] Device:

[1307] Once the generated news report is complete, a notification will be sent to your device.

[1308] The terminal notifies the user of the availability of the report.

[1309] User:

[1310] The user views the news report through the terminal.

[1311] Review reports and quickly review previous news and follow-ups to deepen your understanding.

[1312] 4. Requesting and Providing Follow-Up and Related Information

[1313] User:

[1314] While viewing a report, users can request further updates or related information on a particular topic.

[1315] server:

[1316] The server receives the request and uses RAG (Retrievable and Generative Model) to search for relevant up-to-date information on the Internet and other databases.

[1317] Update news reports with newly discovered information.

[1318] Generation AI:

[1319] The generator AI then updates the personalized news report again to include any follow-up or related information requested by the user.

[1320] Device:

[1321] Once the updated news report is ready, you will be notified again on your device.

[1322] User:

[1323] Users can view updated news reports to see the latest updates and additional information.

[1324] This system not only helps users efficiently track and deeply understand the news even in their busy daily lives, but also contributes to improving media literacy in society as a whole by providing reliable sources of information.

[1325] The processing flow will be explained below.

[1326] Explanation of the specific processing of the program

[1327] ---

[1328] Step 1:

[1329] User:

[1330] A user opens a news site or app and reads a news article.

[1331] Step 2:

[1332] Device:

[1333] The device uses the user identifier to record metadata such as the title, URL, and viewing time of news articles viewed by the user.

[1334] Step 3:

[1335] Device:

[1336] The device makes an API request to send the recorded news viewing information to the server in real time.

[1337] Step 4:

[1338] server:

[1339] The server processes the news browsing information received from the terminal and stores it in a database together with the corresponding user identifier.

[1340] Step 5:

[1341] server:

[1342] Periodically (eg, weekly), the server analyzes the news browsing history stored in the database.

[1343] Step 6:

[1344] server:

[1345] The server analyzes the browsing history using natural language processing techniques to identify the user's interests.

[1346] Step 7:

[1347] server:

[1348] The server uses generative AI to generate personalized news reports based on the analysis results.

[1349] Step 8:

[1350] server:

[1351] The generated news reports are stored in a database and are ready to be notified to the device.

[1352] Step 9:

[1353] Device:

[1354] The terminal notifies the user that a news report is available.

[1355] Step 10:

[1356] User:

[1357] The user receives a notification from the device and views the news report.

[1358] Step 11:

[1359] User:

[1360] Users review news reports and request follow-up or related information of interest.

[1361] Step 12:

[1362] Device:

[1363] The device makes an API request to send the request from the user to the server.

[1364] Step 13:

[1365] server:

[1366] Based on the received request, the server uses the RAG model to search the internet and other databases for relevant follow-up information.

[1367] Step 14:

[1368] server:

[1369] Based on the search results, the generating AI will update the personalized news report again, adding follow-up and related information.

[1370] Step 15:

[1371] server:

[1372] The updated news report is again stored in the database and notified to the terminal.

[1373] Step 16:

[1374] Device:

[1375] The terminal will again announce that an updated news report is available.

[1376] Step 17:

[1377] User:

[1378] Users can view updated news reports to see the latest updates and additions.

[1379] ---

[1380] Through this process, a system is provided that allows users to efficiently track and deeply understand news even in their busy daily lives.

[1381] Example 1

[1382] 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."

[1383] Conventional news article delivery systems lacked sufficient personalization functionality that took into account a user's news browsing history, making it difficult to accurately provide information that users were interested in. Furthermore, there was a lack of a way for users to easily request and obtain follow-up reports or related information on specific news stories. This made it difficult for users to quickly and efficiently obtain the latest and most relevant information.

[1384] 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.

[1385] In this invention, the server includes means for collecting news browsing information, means for storing the collected news browsing information in a database, means for analyzing the stored news browsing information using natural language processing technology and generating a personalized news report using a generative artificial intelligence model, and means for providing the generated personalized news report to a user. This makes it possible to provide a personalized news report that takes into account the user's news browsing history, and to quickly obtain and provide related information in response to a user's request.

[1386] "News viewing information" is viewing data when a user views a news article, and specifically includes metadata such as the news article title, URL, author, viewing date and time, and viewing time.

[1387] A "database" is an information system for efficiently storing, organizing, and retrieving large amounts of data.

[1388] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and generate human language, and includes classification, extraction, and text analysis.

[1389] A "generative artificial intelligence model" is an artificial intelligence technology that learns from large datasets and automatically generates text that is meaningful to users.

[1390] A "personalized news report" is a customized news summary generated based on a user's news browsing history and tailored to their individual interests and preferences.

[1391] A "server" is a computer system that provides services to other computers (clients) on a network.

[1392] A "request" is an action by a user requesting specific information or service, or the content of that request.

[1393] "Searching" is the act of locating information from a database or the Internet based on specified criteria.

[1394] "Related information" is additional data or insight that is relevant to the user's interests or the information they are currently viewing.

[1395] MODE FOR CARRYING OUT THE INVENTION

[1396] The present invention relates to a system for collecting, storing, and analyzing news browsing information, generating personalized news reports based on the collected information, and providing the reports to users. The present invention is implemented by combining a specific software system and infrastructure.

[1397] Collection and storage of news browsing information

[1398] server:

[1399] The server provides an API endpoint for receiving news viewing information from news sites and apps. It receives metadata such as the news article title, URL, author, viewing date and time, and viewing time. The received data is stored for each user ID in the "news_views" table in the MySQL database.

[1400] Device:

[1401] The device collects news articles that users have viewed through news sites and apps. The device generates a unique identifier (user ID) for each user and sends it to the server's API endpoint along with news viewing information.

[1402] User:

[1403] Users browse news articles on a daily basis using news websites and apps, with no special operations required.

[1404] News browsing analysis and report generation

[1405] server:

[1406] The server periodically queries the news viewing information stored in the "news_views" table to obtain the user's browsing history. It analyzes the obtained data and extracts article topics and keywords using a natural language processing library (e.g., SpaCy) that identifies topics and article trends that each user is interested in. Based on the analysis results, it generates prompts for a generative artificial intelligence model (e.g., OpenAI GPT-3.5) to create a personalized news report.

[1407] Generative AI models:

[1408] The generative AI model generates a news report in natural language based on the server's prompts and matches the user's interests.

[1409] Example: "Updates on X: Updates and insights on last week's news story."

[1410] Providing news reports

[1411] Device:

[1412] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability.

[1413] User:

[1414] Users receive notifications on their devices, view news reports, and check the report content to stay up to date on related news and updates.

[1415] Requesting and providing follow-up and related information

[1416] User:

[1417] While viewing a news report, users can request updates or additional information on a particular topic.

[1418] server:

[1419] The server receives user requests and performs searches based on the requests, using the RAG model (e.g., Haystack) to gather relevant, up-to-date information from the Internet and other databases.

[1420] Example prompt: "Create a personalized news report containing the latest news that may be of interest to the user based on their recent browsing history. Specifically, detail new information and related updates about X, as well as future projections."

[1421] Generative AI models:

[1422] The generative AI model then uses the additional information collected to generate an updated news report for the user.

[1423] Device:

[1424] Once the updated news report is ready, you will be notified again on your device.

[1425] User:

[1426] Users receive notifications, view updated news reports, and stay up to date with new developments and relevant information.

[1427] The system allows users to efficiently receive personalized news reports based on their news browsing history, and they can request additional information on specific topics to stay up to date.

[1428] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1429] Processing Steps

[1430] Step 1:

[1431] server:

[1432] The server provides an API endpoint for receiving news browsing information, including metadata such as the news article title, URL, author, browsing date and time, and browsing duration.

[1433] Input: News browsing information sent from your device.

[1434] Data processing: Analyze the received metadata and convert it into a format that can be saved in the database for each user ID.

[1435] Output: Organized metadata stored in a database.

[1436] Step 2:

[1437] Device:

[1438] The device collects news articles that users have viewed through news sites and apps. The device generates a unique identifier (user ID) for each user and sends it to the server's API endpoint along with news viewing information.

[1439] Input: News article information obtained from news sites and apps.

[1440] Data processing: A user ID is assigned and the news article metadata is converted into a format that can be sent to the server.

[1441] Output: News browsing information sent to the server's API endpoint.

[1442] Step 3:

[1443] User:

[1444] Users browse news articles on a daily basis using news websites and apps, with no special operations required.

[1445] Input: User's news article browsing behavior.

[1446] Data processing: None.

[1447] Output: News browsing data based on user browsing behavior.

[1448] Step 4:

[1449] server:

[1450] The server periodically queries the news browsing information stored in the database to obtain the user's browsing history, analyzes the obtained data, and identifies topics and article trends that each user is interested in. It then uses a natural language processing library (e.g., SpaCy) to extract article topics and keywords.

[1451] Input: News browsing information stored in a database.

[1452] Data processing: Using natural language processing technology, topics and keywords from news articles are extracted and user interests are analyzed.

[1453] Output: Analysis results that reflect each user's interests and concerns.

[1454] Step 5:

[1455] server:

[1456] Based on the analysis results, the server generates prompt sentences for a generative AI model (e.g., OpenAI GPT-3.5) to create a personalized news report.

[1457] Input: Results analyzed using natural language processing and user browsing history.

[1458] Data processing: Generate prompt sentences and feed them into a generative AI model to generate personalized news reports.

[1459] Output: The generated personalized news report.

[1460] Action: Example prompt: "Create a personalized news report containing the latest news that may be of interest to the user based on their recent browsing history. Specifically, include new information and related updates about ____, as well as future projections."

[1461] Step 6:

[1462] Device:

[1463] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability.

[1464] Input: A news report sent from the server.

[1465] Data processing: Convert to notification format.

[1466] Output: Notify the user and display a news report.

[1467] Step 7:

[1468] User:

[1469] Users receive notifications on their devices, view news reports, and check the report content to stay up to date on related news and updates.

[1470] Input: A news report posted to the device.

[1471] Data processing: None.

[1472] Output: Confirmed news report viewing behavior of the user.

[1473] Step 8:

[1474] User:

[1475] While viewing a news report, users can request updates or additional information on a particular topic.

[1476] Input: A user request for a news report.

[1477] Data processing: None.

[1478] Output: Request sent.

[1479] Step 9:

[1480] server:

[1481] The server receives user requests and performs searches based on the requests, using the RAG model (e.g., Haystack) to gather relevant, up-to-date information from the Internet and other databases.

[1482] Input: The user's request and the database to search.

[1483] Data processing: Search and retrieve relevant information using the RAG model.

[1484] Output: The most recent relevant information retrieved.

[1485] Step 10:

[1486] Generative AI models:

[1487] The generative AI model then uses the additional information collected to generate an updated news report for the user.

[1488] Input: retrieved related information, previous browsing history, user interests.

[1489] Data processing: Generate prompt sentences and feed them into a generative AI model to generate updated news reports.

[1490] Output: An updated personalized news report.

[1491] Step 11:

[1492] Device:

[1493] Once the updated news report is ready, you will be notified again on your device.

[1494] Input: Updated news report sent from the server.

[1495] Data processing: Convert to notification format.

[1496] Output: Notify user and display update news report.

[1497] Step 12:

[1498] User:

[1499] Users receive notifications, view updated news reports, and stay up to date with new developments and relevant information.

[1500] Input: Updated news report notified to terminal.

[1501] Data processing: None.

[1502] Output: User's viewing behavior for latest information.

[1503] (Application example 1)

[1504] 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."

[1505] In modern society, a huge amount of news information is provided via the Internet, but users face the challenge of efficiently gathering information relevant to them. In particular, the lack of personalized news reports tailored to individual users makes it difficult for users to access news information based on their interests. Furthermore, when users want more information on a particular topic, they have to search manually, which is inconvenient.

[1506] 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.

[1507] In this invention, the server includes means for collecting a user's news browsing history, means for storing the collected news browsing history in a database, means for analyzing the stored news browsing history and generating a prompt sentence using a generative AI model to generate a personalized news report, and means for providing the generated personalized news report to the user. This effectively provides a personalized news report to the user, allowing the user to easily access news information based on their interests. In addition, requested follow-up reports and updates to the news report are automatically searched for, reducing the burden on the user of information gathering.

[1508] "User's news browsing history" is a record of information about news articles that the user has viewed, and includes metadata such as the article title, URL, author, viewing date and time, and viewing time.

[1509] A "database" is a collection of data that systematically stores collected news browsing history and can be retrieved and analyzed later.

[1510] A "generative AI model" is a model that uses artificial intelligence technology to generate prompt sentences based on saved news viewing history and generate personalized news reports.

[1511] A "prompt sentence" is an input sentence used by a generative AI model, which is an instruction sentence to generate an appropriate news report based on the user's news browsing history.

[1512] A "personalized news report" is a personalized news summary generated based on a user's interests and concerns, including related supplemental information and follow-up reports.

[1513] A "request" is a request for information made by a user for more details about a particular news topic or related information.

[1514] A "follow-up" is additional news that includes new information or subsequent developments related to news already provided.

[1515] "Searching" is the process of finding specific information in a database or on the Internet.

[1516] "Update" refers to the reorganization of an existing news report to bring it up to date, for example by adding new information.

[1517] The "means for providing to the user" includes a method and infrastructure for notifying the generated personalized news report to the user's terminal and making it viewable.

[1518] This invention relates to a system that collects, stores, and analyzes a user's news browsing history, and generates and provides personalized news reports based on the results using a generative AI model.

[1519] 1. Collection and storage of news browsing information

[1520] server:

[1521] The server provides an API endpoint to receive news browsing information from news sites and apps, and receives metadata such as the title, URL, author, date and time of browsing, and the browsing time of the news article. The server then stores the received data in a database for each user.

[1522] The main software running on the server is a RESTful API server (e.g., Flask, Django) that provides API endpoints.

[1523] Device:

[1524] The device sets an identifier to identify the user and collects information about news articles the user has viewed on news sites and apps, and sends the collected data to the server's API endpoint.

[1525] A news application (e.g., Android or iOS app) is installed on the device and provides the user interface.

[1526] User:

[1527] Users browse news articles on news websites or apps, and the system automatically collects data based on normal news browsing behavior, so users do not need to take any special action.

[1528] 2. Analyzing news browsing information and generating reports

[1529] server:

[1530] The server periodically analyzes the news browsing information stored in the database, using natural language processing techniques (e.g., spaCy, NLTK) to identify user interests.

[1531] Based on the analysis results, a generative AI model (e.g., OpenAI GPT-3) is used to generate prompt sentences, which are then used to generate a personalized news report.

[1532] The generated personalized news report includes the user's recently viewed articles, related supplemental information, follow-up reports, and more.

[1533] Generation AI:

[1534] The generative AI takes into account the user's browsing history and interests and generates a report in natural language, using prompts such as:

[1535] Generate the following news report based on topics of interest to user 12345:

[1536] Recent news trends indicate an interest in environmental issues, so include content that informs users about the latest environmental news and research findings.

[1537] 3. Providing news reports

[1538] Device:

[1539] Once the generated news report is complete, a notification is sent to the device, using a mobile notification service (e.g., Firebase Cloud Messaging) to notify the user of the report's availability.

[1540] User:

[1541] Users can view news reports on their devices, allowing them to easily review news they have viewed in the past and follow-up reports.

[1542] 4. Requesting and Providing Follow-Up and Related Information

[1543] User:

[1544] If a user is viewing a report and wants to know more about a particular topic or related information, they can make a request.

[1545] server:

[1546] The server receives the request, uses RAG (Retrievable and Generative Model) to search the internet and other databases for relevant up-to-date information, and updates the news report based on the newly found information.

[1547] Generation AI:

[1548] The generator AI then updates the personalized news report again to include any follow-up or related information requested by the user.

[1549] Device:

[1550] Once the updated news report is ready, you will be notified again on your device.

[1551] User:

[1552] Users can view updated news reports to see the latest updates and additional information.

[1553] This system allows users to efficiently receive personalized news reports and easily access news information based on their interests. It also automatically searches for requested follow-up news and updates the reports, significantly reducing the burden on users of information gathering.

[1554] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1555] Step 1:

[1556] Collection of news browsing information

[1557] Input: News article information viewed by the user on a news site or app (title, URL, author, viewing date and time, viewing time)

[1558] What it does: When a user browses news, the device (e.g., a smartphone) collects the metadata and sends it to the server's API endpoint. The device has a news app installed, which automatically collects data every time the user reads news.

[1559] Step 2:

[1560] Saving news browsing information

[1561] Input: News article information sent to the server

[1562] Specific operation: The server stores the news article information received through the API endpoint in a database. The database management system (e.g., MySQL, PostgreSQL) organizes and stores the news data for each user.

[1563] Step 3:

[1564] Analysis of news browsing information

[1565] Input: News browsing history stored in the database

[1566] How it works: The server periodically analyzes the user's news browsing history using natural language processing techniques (e.g., spaCy, NLTK). The analysis identifies trends in topics and articles that interest the user. The server then identifies the user's areas of interest based on the content, frequency, and viewing time of articles.

[1567] Step 4:

[1568] Generate personalized news reports

[1569] Input: Analysis results (data indicating user interests)

[1570] Specific behavior: The server generates a prompt using a generative AI model (e.g., OpenAI GPT-3) and generates a personalized news report based on it. For example, the prompt could be, "Generate the following news report based on the topics that user 12345 is interested in: Recent news trends indicate that they are interested in environmental issues. Therefore, please include content that informs the user about the latest environmental news and research findings."

[1571] Step 5:

[1572] Providing news reports

[1573] Input: Generated personalized news report

[1574] Specific operation: The server sends the generated news report to the device and notifies the user via a mobile notification service (e.g., Firebase Cloud Messaging). The user can then view the report on their device.

[1575] Step 6:

[1576] Requesting and providing follow-up and related information

[1577] Input: User requests (follow-ups or related information on a particular topic)

[1578] How it works: When a user wants to know the latest updates or related information on a particular topic, they make a request from their device. The server receives the request and uses RAG (Search and Generative Model) to search the internet and other databases for relevant updates. It then uses the generative AI model again to generate an updated, personalized news report.

[1579] Step 7:

[1580] Providing updated news reports

[1581] Input: Updated personalized news report

[1582] Specific operation: The server sends the updated news report back to the device and notifies the user via the mobile notification service. The user can then view the updated news report and check the latest updates and additional information.

[1583] The above is the overall processing flow of the system, and at each step it is possible to efficiently collect and analyze news information from users and provide personalized information.

[1584] 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.

[1585] The present invention relates to a system for generating and providing news reports optimized for users by recognizing, collecting, storing, and analyzing a user's news browsing history and emotions. Hereinafter, an embodiment of the present invention will be described in detail.

[1586] 1. Collection and storage of news browsing and sentiment data

[1587] server

[1588] The server provides an API endpoint to receive news browsing information and sentiment data from news sites and apps.

[1589] It receives metadata such as the news article title, URL, author, viewing date and time, and viewing time, as well as the user's emotion data recognized by the emotion engine.

[1590] Save the received data in the database for each user.

[1591] Terminal

[1592] The device sets a user identifier and collects information about news articles the user has viewed on news sites or apps, as well as emotion data recognized by the emotion engine.

[1593] Send the collected data to the server's API endpoint.

[1594] User

[1595] Users use news sites and apps to view news articles.

[1596] Users basically operate the app in the same way as they normally would when browsing the news, but the emotion engine uses facial and emotion recognition technology to record the user's expressions and reactions.

[1597] 2. Analyzing news browsing information and sentiment data and generating reports

[1598] server

[1599] The server periodically analyzes the news browsing history and sentiment data stored in the database.

[1600] It uses natural language processing and sentiment analysis technologies to identify topics and article trends that interest users, as well as their thoughts when reading the news.

[1601] Based on the analysis results, generative AI is used to generate personalized news reports.

[1602] Generation AI

[1603] Generative AI generates reports in natural language based on the user's browsing history and sentiment.

[1604] Reports include users' recently viewed articles, related supplemental information, and adjustments to the content and tone of news based on sentiment.

[1605] Example: "Latest news on XX: Updates on last week's news and recommendations based on sentiment analysis."

[1606] 3. Providing news reports

[1607] Terminal

[1608] Once the generated news report is complete, a notification will be sent to your device.

[1609] The terminal notifies the user of the availability of the report.

[1610] User

[1611] The user receives a notification from the device and views the news report.

[1612] Review reports and quickly review previous news and follow-ups to deepen your understanding.

[1613] 4. Requesting and Providing Follow-Up and Related Information

[1614] User

[1615] While viewing a report, users can request further updates or related information on a particular topic.

[1616] server

[1617] The server receives the request and uses the RAG model to search the internet and other databases for relevant follow-up information.

[1618] Update news reports with newly discovered information.

[1619] Generation AI

[1620] The generative AI then updates the personalized news report again, adding follow-up and related information.

[1621] Terminal

[1622] Once the updated news report is ready, you will be notified again on your device.

[1623] User

[1624] Users can view updated news reports to see the latest updates and additions.

[1625] Specific examples

[1626] 1. User A:

[1627] I was browsing an article about climate change on a news app.

[1628] The emotion engine analyzes user A's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article.

[1629] 2. Terminal:

[1630] News browsing information and emotion data are sent to the server.

[1631] 3. Server:

[1632] The received data is collected and stored in a database.

[1633] The data is analyzed periodically and a sentiment engine generates personalized news reports based on the collected data.

[1634] 4. Generation AI:

[1635] A news report was created that focused on "climate change," which User A was interested in.

[1636] Reflect sentiment data and include many articles with a positive tone.

[1637] 5. Terminal:

[1638] When the news report is complete, User A is notified.

[1639] 6. User A:

[1640] Open the report to see the latest news you've read and follow-up updates.

[1641] This embodiment provides a news experience that takes into account the user's emotions and leads to greater understanding and interest.

[1642] The processing flow will be explained below.

[1643] Explanation of the specific processing of the program

[1644] ---

[1645] Step 1:

[1646] User:

[1647] A user opens a news site or app and reads a news article.

[1648] Step 2:

[1649] Device:

[1650] The device uses the user identifier to record metadata such as the title, URL, and viewing time of news articles viewed by the user.

[1651] In addition, the device's built-in emotion engine analyzes emotional data from the user's facial expressions and voice in real time, recording the type of emotion (joy, surprise, sadness, etc.) and its intensity.

[1652] Step 3:

[1653] Device:

[1654] The device makes an API request to send the recorded news viewing information and emotion data to the server in real time.

[1655] Step 4:

[1656] server:

[1657] The server processes the news browsing information and emotion data received from the terminal and stores them in a database together with the corresponding user identifier.

[1658] Step 5:

[1659] server:

[1660] Periodically (eg, weekly), the server analyzes the news browsing history and sentiment data stored in the database.

[1661] Step 6:

[1662] server:

[1663] The server analyzes browsing history and emotional data using natural language processing and sentiment analysis techniques to identify user interests.

[1664] For example, if a user consistently expresses enjoyment or interest in a particular topic, that topic will be given more weight.

[1665] Step 7:

[1666] server:

[1667] The server uses generative AI to generate personalized news reports based on the analysis results.

[1668] Report content is optimized based on user interests and emotional responses.

[1669] Step 8:

[1670] server:

[1671] The generated news reports are stored in a database and are ready to be notified to the device.

[1672] Step 9:

[1673] Device:

[1674] The terminal notifies the user that a news report is available.

[1675] Step 10:

[1676] User:

[1677] The user receives a notification from the device and views the news report.

[1678] Step 11:

[1679] User:

[1680] Users can view news reports and see selected articles and follow-up updates based on sentiment data.

[1681] Request any follow-up or related information that may be of interest to you.

[1682] Step 12:

[1683] Device:

[1684] The device makes an API request to send the request from the user to the server.

[1685] Step 13:

[1686] server:

[1687] Based on the received request, the server uses the RAG model to search the internet and other databases for relevant follow-up information.

[1688] Step 14:

[1689] server:

[1690] Based on the search results, the generating AI will update the personalized news report again, adding follow-up and related information.

[1691] Step 15:

[1692] server:

[1693] The updated news report is again stored in the database and notified to the terminal.

[1694] Step 16:

[1695] Device:

[1696] The terminal will again announce that an updated news report is available.

[1697] Step 17:

[1698] User:

[1699] Users can view updated news reports to see the latest updates and additions.

[1700] ---

[1701] Through this process, users can receive more personalized news reports that take into account their emotions when reading the news. In particular, the collection and analysis of emotional data allows the content and tone of the news to be optimized for users, eliciting deeper understanding and interest.

[1702] Example 2

[1703] 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."

[1704] Current news delivery systems achieve some degree of personalization based on users' news browsing history, but do not adequately consider users' emotions when providing news. As a result, users are unable to efficiently obtain news that matches their interests and emotions, resulting in a poor news browsing experience. Furthermore, when users request follow-up or related information on a specific topic, responses are often delayed, further worsening the user experience.

[1705] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting a user's news browsing history and user emotion data, means for storing the collected news browsing history and user emotion data in a database, means for analyzing the stored news browsing history and emotion data to generate a personalized news report, and means for providing the generated personalized news report to the user. This makes it possible to provide personalized news based on the user's emotion data, improving the quality of the news browsing experience. Furthermore, follow-up reports and related information requested by the user can be quickly provided, thereby increasing user satisfaction.

[1706] "News viewing history" is information about news articles viewed by the user, and specifically includes metadata such as the article title, URL, author, viewing date and time, and viewing time.

[1707] "Emotion data" refers to information about the emotional state recognized from the user's facial expressions and behavior, and indicates data that expresses emotions such as joy, surprise, and sadness.

[1708] "Database" refers to a computer system for systematically storing and managing collected news viewing history and emotion data.

[1709] "Analysis" refers to the process of analyzing news browsing history and emotional data stored in a database using machine learning and natural language processing technologies to identify trends in users' interests and emotions.

[1710] A "personalized news report" is a news report customized based on a user's news viewing history and emotional data, and includes content optimized for the user's interests and emotions.

[1711] "Generative AI" refers to systems that use artificial intelligence technology to generate personalized news reports, specifically natural language processing models (e.g., GPT-4).

[1712] A "follow-up" is any new information provided in relation to an already reported news story.

[1713] "Related information" refers to supplementary information or topics related to the news article the user viewed.

[1714] "Server" refers to the central computer system that collects, stores, and analyzes users' news browsing history and sentiment data to generate and provide personalized news reports.

[1715] A "terminal" is a device that a user uses to view news articles, and includes a smartphone, tablet, PC, etc.

[1716] The present invention relates to a system for collecting, storing, and analyzing a user's news browsing history and emotion data, and generating and providing a news report optimized for the user. Hereinafter, an embodiment of the invention will be described in detail.

[1717] Collection and storage of news browsing and sentiment data

[1718] server

[1719] The server provides an API endpoint for receiving news browsing information and emotion data from news sites and apps. Specifically, it receives metadata such as the news article title, URL, author, viewing date and time, and viewing time, as well as user emotion data recognized by the emotion engine. The received data is stored in a database for each user. This database is built using, for example, MySQL or MongoDB.

[1720] Terminal

[1721] The device sets a user identifier and collects information about news articles the user has viewed on news sites and apps, as well as emotion data recognized by an emotion engine. The collected data is periodically sent to the server's API endpoint. Emotion data is collected using the device's camera and an emotion analysis engine (e.g., Face API).

[1722] User

[1723] Users browse news articles on news websites or apps. The user's operations are basically the same as when browsing the news normally, but the emotion engine uses facial and emotion recognition technology to record the user's expressions and reactions.

[1724] Analyzing news browsing and sentiment data and generating reports

[1725] server

[1726] The server periodically analyzes the news browsing history and sentiment data stored in the database. It utilizes natural language processing (NLTK and spaCy) and sentiment analysis (IBM Watson and Microsoft Azure sentiment analysis APIs) to identify topics and article trends that interest users, as well as their thoughts while browsing the news. Based on the analysis results, a generative AI model (e.g., GPT-4) is used to generate personalized news reports.

[1727] Generation AI

[1728] Based on a user's browsing history and sentiment, the generative AI generates a natural language report that includes the user's recently viewed articles, relevant supplemental information, and tailoring the content and tone of the news based on sentiment.

[1729] Prompt Sentence Examples

[1730] "Based on User A's recent news about climate change, create a summary of articles with a positive tone that reflects surprising sentiment data."

[1731] Providing news reports

[1732] Terminal

[1733] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability, ensuring that it remains accessible to the user.

[1734] User

[1735] Users receive notifications on their devices to view news reports, and can review the reports to gain more detailed information about previously viewed news stories and follow-ups.

[1736] Requesting and providing follow-up and related information

[1737] User

[1738] While viewing a report, users can request follow-up or related information on a specific topic. This request is made through the News app.

[1739] server

[1740] The server receives the request and uses the Retrieval-Augmented Generation (RAG) model to search the Internet and other databases for relevant follow-up news, updating the news report with the newly discovered information.

[1741] Generation AI

[1742] The generative AI then updates the personalized news report again, adding follow-up and related information.

[1743] Terminal

[1744] Once the updated news report is complete, the user will receive another notification on their device, and they can view the updated news report to see any updates or additional information.

[1745] Specific examples

[1746] 1. User A:

[1747] When a user reads an article about climate change in a news app, the emotion engine analyzes the user's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article.

[1748] 2. Terminal:

[1749] News browsing information and emotion data are sent to a server.

[1750] 3. Server:

[1751] The received data is collected and stored in a database. The data is periodically analyzed and a sentiment engine generates personalized news reports based on the collected data.

[1752] 4. Generation AI:

[1753] Create a news report that focuses on "climate change," which User A is interested in. Reflect the sentiment data and include many articles with a positive tone.

[1754] 5. Terminal:

[1755] When the news report is complete, User A is notified.

[1756] 6. User A:

[1757] Open the report to see the news you recently read and follow-ups. This embodiment provides a news experience that takes into account the user's emotions and provides a deeper understanding and interest.

[1758] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1759] Step 1: User Views News Article

[1760] A user browses a news article using a news website or app. Specifically, the user opens a news app and selects and reads an article about climate change.

[1761] Input: Select the article you want to view on the news app

[1762] Output: Transmission of viewed article information

[1763] Step 2: Collecting news browsing information and emotion data from the device

[1764] The device automatically collects metadata (title, URL, author, date and time of viewing, duration, etc.) of news articles viewed by the user, using the data collection function within the news app.

[1765] The device uses its built-in camera and sensors to collect the user's facial expressions and emotional data (happiness, surprise, sadness, etc.) through an emotion engine (e.g., Face API).

[1766] Input: Browsing information in news apps, emotion data from cameras and sensors

[1767] Output: A batch of collected news metadata and sentiment data

[1768] Step 3: Send data from the device to the server

[1769] The device sends the collected news browsing information and sentiment data to a server through a pre-configured API endpoint, using an internet connection for this communication.

[1770] Input: Batches of news metadata and sentiment data collected on the device

[1771] Output: Confirmation of data transmission to the server

[1772] Step 4: Server saves data

[1773] The server organizes the received news browsing information and emotion data for each user and stores them in a database (e.g., MySQL or MongoDB) using a database management system.

[1774] Input: A batch of news metadata and sentiment data sent from the device

[1775] Output: Save confirmation message

[1776] Step 5: Periodic analysis of data by the server

[1777] The server periodically analyzes the news browsing information and sentiment data in the database, using natural language processing technologies (e.g., NLTK, spaCy) and sentiment analysis technologies (e.g., IBM Watson, Microsoft Azure sentiment analysis API) to identify trends in user interests and sentiment.

[1778] Input: News metadata and sentiment data in a database

[1779] Output: Analysis result report

[1780] Step 6: Generative AI creates personalized news reports

[1781] Based on the analysis results, the server sends prompts to a generative AI (e.g., GPT-4) to generate a personalized news report, which is written in natural language based on the user's browsing history and emotions.

[1782] Input: Analysis result report and prompt statement

[1783] Output: Personalized news report

[1784] Prompt Sentence Examples

[1785] "Based on User A's recent news about climate change, create a summary of articles with a positive tone that reflects surprising sentiment data."

[1786] Step 7: News report notification on your device

[1787] The server sends the generated news report to the terminal, which displays a notification to the user that a new news report is available.

[1788] Input: Personalized News Report

[1789] Output: Notification message to terminal

[1790] Step 8: User Views News Report

[1791] Users receive notifications from their devices and open and view news reports, which include recently viewed stories, follow-ups, and tone adjustments based on emotion.

[1792] Input: Notifications from your device

[1793] Output: Content of the news report

[1794] Step 9: User requests follow-up or related information

[1795] While viewing a news report, users interact with an interface to request updates or related information on a particular topic.

[1796] Input: Request button in news report

[1797] Output: Request information

[1798] Step 10: Server searches for relevant information and updates the report

[1799] The server receives user requests, uses the RAG (Retrieval-Augmented Generation) model to search the Internet and other databases for related or follow-up information, and updates the news report based on the newly found information.

[1800] Input: Request information from the user

[1801] Output: Updated news report

[1802] Step 11: Regenerating the news report with generative AI

[1803] The generative AI generates a personalized news report again based on the new information.

[1804] Input: List of newly found information

[1805] Output: Updated personalized news report

[1806] Step 12: Update report notification to device

[1807] The server retransmits the updated news report to the terminal, which displays a notification to the user that the report has been updated.

[1808] Input: Updated news report

[1809] Output: Notification message to terminal

[1810] Step 13: User Views Update News Report

[1811] Users can view updated news reports to see the latest updates and additions.

[1812] Input: Notifications from your device

[1813] Output: Updated news report content

[1814] Through these steps, the system can personalize users' news browsing experience and provide optimal news reports that reflect emotional data.

[1815] (Application example 2)

[1816] 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."

[1817] Conventional news delivery systems personalized news based on users' browsing history, but it was difficult to consider users' emotional responses. This made it difficult to provide optimal news reports that would deepen users' interest and understanding. Furthermore, there was a lack of mechanisms for generating reports that accurately reflected users' emotions toward specific news articles. To address these issues, it is necessary to collect and analyze not only news browsing history but also user emotional data during browsing to enrich the user experience. Furthermore, there is a need for more intuitive, real-time data collection and reporting using wearable devices such as smart glasses.

[1818] 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.

[1819] In this invention, the server includes means for collecting a user's news viewing history and emotional data, means for storing the collected news viewing history and emotional data in a database, means for analyzing the stored news viewing history and emotional data to generate a personalized news report, means for providing the generated personalized news report to the user, means for collecting the user's news viewing and emotional data in real time through the smart glasses, means for analyzing the user's emotional data, and means for providing the user with a function for viewing the news report using voice commands or gestures. This enables the generation and provision of more personalized news reports based on the user's emotional responses. Furthermore, the user's daily data can be collected naturally through the smart glasses, allowing the user to intuitively view the news report, thereby improving the user experience.

[1820] "News browsing history" is a collection of metadata such as the title, URL, author, browsing date and time, and browsing time of news articles that a user has viewed.

[1821] "Emotion data" is information on emotions that are recorded and recognized by the emotion engine while the user is browsing the news.

[1822] A "database" is a collection that stores collected news viewing history and emotional data and manages it so that it can be searched and analyzed as needed.

[1823] A "personalized news report" is a news summary optimized for a user based on the user's news viewing history and emotional data.

[1824] "Smart glasses" are wearable devices equipped with functions such as a camera and a display, and can recognize the user's gaze direction and facial expressions and display information.

[1825] "Means for collecting data in real time" refers to a method for capturing the user's facial expressions and reactions while viewing the news in real time through the camera in the smart glasses and instantly transmitting them to a server.

[1826] "Ability to browse using voice commands and gestures" refers to an interface that allows a user to navigate and browse news reports using voice or body movements.

[1827] The "means for analyzing user emotional data" is a technology that uses collected emotional data to evaluate the emotions felt by users while browsing news and changes in those emotions.

[1828] This invention is a system that collects, stores, and analyzes users' news browsing history and emotional data to generate and provide personalized news reports.

[1829] 1. Collection and storage of news browsing and sentiment data

[1830] server

[1831] The server provides an API endpoint for receiving news browsing information and emotion data from news sites and applications. It receives metadata such as the title, URL, author, viewing date and time of news articles, as well as user emotion data detected by the emotion engine, and stores this data in a database for each user.

[1832] Terminal

[1833] The device collects information about news articles viewed by users on news sites and apps through the smart glasses, as well as emotional data recognized by the emotion engine, and sends this data to the server's API endpoint.

[1834] User

[1835] Users use smart glasses to browse news articles. While browsing, the emotion engine uses facial and emotion recognition technology to record the user's facial expressions and reactions. The data is then sent to a server in real time.

[1836] 2. Analyzing news browsing information and sentiment data and generating reports

[1837] server

[1838] The server periodically analyzes the news browsing history and sentiment data stored in the database to identify trends in topics and articles that users are interested in. This is done using natural language processing and sentiment analysis technologies, and uses generative AI to generate personalized news reports.

[1839] Generation AI

[1840] Based on the user's browsing history and emotional data, the generative AI creates a news report that can be viewed in-car or via gesture control, including recently viewed articles, relevant supplemental information, and an adjustment to the tone of the news article based on the emotional data.

[1841] 3. Providing news reports

[1842] Terminal

[1843] Once the generated news report is complete, it notifies the user through the smart glasses, who can then use voice commands or gestures to view or interact with the report.

[1844] User

[1845] Users receive notifications, access personalized news reports, and easily review previously viewed news stories and follow-up updates. If they want more updates or related information on a particular topic, they can request that information.

[1846] 4. Requesting and Providing Follow-Up and Related Information

[1847] User

[1848] While viewing a report, users can request follow-up or related information on a particular topic.

[1849] server

[1850] The server receives the request and uses the Retrieval-Augmented Generation (RAG) model to search for relevant follow-up information from the Internet and other databases, updating the news report based on the information found.

[1851] Generation AI

[1852] The generative AI is then personalized again to update the news report, adding any necessary follow-up or related information.

[1853] Terminal

[1854] Once the updated news report is ready, another notification will be sent to the smart glasses.

[1855] User

[1856] Users can view updated news reports to see the latest updates and additions.

[1857] Example 1

[1858] A user browses an article about climate change in a news app. The emotion engine analyzes the user's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article. The news browsing information and emotional data are sent to the server via the smart glasses.

[1859] Prompt Sentence Examples

[1860] List of news recently viewed by user 123 (including metadata)

[1861] In addition, based on the emotional data (Happy, Sad, Angry, etc.) when viewing each news article,

[1862] Generate personalized news reports.

[1863] The report includes a summary of your recently viewed articles along with information on topics of interest.

[1864] Provide a list of recommended articles.

[1865] If possible, frame your news in a positive tone.

[1866] This format allows for a deeper, more engaging news experience that takes into account the user's emotions.

[1867] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1868] Step 1:

[1869] The device collects the user's news viewing information and emotion data through the smart glasses. The input is real-time video footage of the news being viewed and the metadata of the news article. The video footage captured by the smart glasses' camera is input to a face recognition and emotion recognition engine to analyze the user's emotion (e.g., surprise, interest, etc.). The output is emotion data corresponding to the metadata of the news article.

[1870] Step 2:

[1871] The device sends the collected news browsing information and emotion data to the server. The input is the news article metadata and emotion data collected in step 1. This is sent as an HTTP request to the server's API endpoint. The output is a notification of successful data transmission to the server.

[1872] Step 3:

[1873] The server stores the received news browsing information and emotion data in a database. The input is the metadata and emotion data of the news article received as an HTTP request. These are stored in the database as records corresponding to each user. The output is a confirmation notification of the record stored in the database.

[1874] Step 4:

[1875] The server periodically analyzes the stored news browsing history and sentiment data. The input is the metadata and sentiment data of news articles stored in the database. A data analysis engine is used to identify user interests and sentiment trends and create the basis for generating personalized news reports. The output is the analysis results data.

[1876] Step 5:

[1877] The server uses a generative AI to generate a personalized news report. The input is the analysis result data obtained in step 4. The generative AI uses natural language processing technology to create a news report based on the user's interests and emotions. The output is a completed personalized news report.

[1878] Step 6:

[1879] The server notifies the terminal of the generated news report. The input is the completed personalized news report, which is sent to the terminal and notified to the user. The output is a notification display from the terminal.

[1880] Step 7:

[1881] The device allows users to browse news reports using voice commands or gestures. The input is a personalized news report sent from the server and the user's voice or gesture instructions. The device displays the news report through the interface of the smart glasses, and displays and changes the content based on the user's actions. The output is the content viewed or manipulated by the user.

[1882] Step 8:

[1883] While viewing a report, a user requests follow-up or related information on a specific topic. The input is a request made by the user through voice commands or gestures. The device receives this request and sends it to the server. The output is a notification of the request to the server.

[1884] Step 9:

[1885] The server receives requests and searches for relevant follow-up information. The input is a request from a user. The RAG model (Retrieval-Augmented Generation model) is used to search for relevant information from the Internet and other databases. The output is the relevant information found.

[1886] Step 10:

[1887] The server updates the news report using a generative AI. The input is the relevant information found and the existing news report. The generative AI adds the relevant information and creates a personalized news report again. The output is the updated news report.

[1888] Step 11:

[1889] The server notifies the terminal of the updated news report. The input is the updated news report. It is sent to the terminal and notified to the user. The output is an update notification display from the terminal.

[1890] Step 12:

[1891] The user browses the updated news report to check the latest updates and additional information. The input is the user's operation upon receiving the update notification. The user checks the news report through the smart glasses and obtains the necessary information. The output is the latest information obtained by the user.

[1892] 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.

[1893] 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.

[1894] 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.

[1895] [Fourth embodiment]

[1896] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1897] 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.

[1898] 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).

[1899] 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.

[1900] 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.

[1901] 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).

[1902] 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.

[1903] 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.

[1904] 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.

[1905] 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.

[1906] 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.

[1907] 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.

[1908] 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."

[1909] ---

[1910] The present invention relates to a system for collecting, storing, and analyzing a user's news browsing history, and generating and providing personalized news reports based on the results. The present invention is implemented by combining a specific software system and infrastructure.

[1911] 1. Collection and storage of news browsing information

[1912] server:

[1913] The server provides an API endpoint for receiving news browsing information from news sites and apps.

[1914] Receive metadata such as news article title, URL, author, date and time of access, and access duration.

[1915] Save the received data in the database for each user.

[1916] Device:

[1917] The device sets a user identifier and collects information about news articles the user has viewed on news sites and apps.

[1918] Send the collected data to the server's API endpoint.

[1919] User:

[1920] Users use news sites and apps to view news articles.

[1921] User operations are basically the same as when browsing news normally, so no special operations are required.

[1922] 2. Analyzing news browsing information and generating reports

[1923] server:

[1924] The server periodically analyzes the news browsing information stored in the database.

[1925] Leverage natural language processing techniques to identify trends in topics and articles that interest users.

[1926] Based on the analysis, generative AI is used to generate personalized news reports, including the user's recently viewed articles, related supplemental information, and follow-up reports.

[1927] Generation AI:

[1928] The generative AI takes into account the user's browsing history and interests and generates reports in natural language.

[1929] Example: "Updates on X: Updates and insights on last week's news story."

[1930] 3. Providing news reports

[1931] Device:

[1932] Once the generated news report is complete, a notification will be sent to your device.

[1933] The terminal notifies the user of the availability of the report.

[1934] User:

[1935] The user views the news report through the terminal.

[1936] Review reports and quickly review previous news and follow-ups to deepen your understanding.

[1937] 4. Requesting and Providing Follow-Up and Related Information

[1938] User:

[1939] While viewing a report, users can request further updates or related information on a particular topic.

[1940] server:

[1941] The server receives the request and uses RAG (Retrievable and Generative Model) to search for relevant up-to-date information on the Internet and other databases.

[1942] Update news reports with newly discovered information.

[1943] Generation AI:

[1944] The generator AI then updates the personalized news report again to include any follow-up or related information requested by the user.

[1945] Device:

[1946] Once the updated news report is ready, you will be notified again on your device.

[1947] User:

[1948] Users can view updated news reports to see the latest updates and additional information.

[1949] This system not only helps users efficiently track and deeply understand the news even in their busy daily lives, but also contributes to improving media literacy in society as a whole by providing reliable sources of information.

[1950] The processing flow will be explained below.

[1951] Explanation of the specific processing of the program

[1952] ---

[1953] Step 1:

[1954] User:

[1955] A user opens a news site or app and reads a news article.

[1956] Step 2:

[1957] Device:

[1958] The device uses the user identifier to record metadata such as the title, URL, and viewing time of news articles viewed by the user.

[1959] Step 3:

[1960] Device:

[1961] The device makes an API request to send the recorded news viewing information to the server in real time.

[1962] Step 4:

[1963] server:

[1964] The server processes the news browsing information received from the terminal and stores it in a database together with the corresponding user identifier.

[1965] Step 5:

[1966] server:

[1967] Periodically (eg, weekly), the server analyzes the news browsing history stored in the database.

[1968] Step 6:

[1969] server:

[1970] The server analyzes the browsing history using natural language processing techniques to identify the user's interests.

[1971] Step 7:

[1972] server:

[1973] The server uses generative AI to generate personalized news reports based on the analysis results.

[1974] Step 8:

[1975] server:

[1976] The generated news reports are stored in a database and are ready to be notified to the device.

[1977] Step 9:

[1978] Device:

[1979] The terminal notifies the user that a news report is available.

[1980] Step 10:

[1981] User:

[1982] The user receives a notification from the device and views the news report.

[1983] Step 11:

[1984] User:

[1985] Users review news reports and request follow-up or related information of interest.

[1986] Step 12:

[1987] Device:

[1988] The device makes an API request to send the request from the user to the server.

[1989] Step 13:

[1990] server:

[1991] Based on the received request, the server uses the RAG model to search the internet and other databases for relevant follow-up information.

[1992] Step 14:

[1993] server:

[1994] Based on the search results, the generating AI will update the personalized news report again, adding follow-up and related information.

[1995] Step 15:

[1996] server:

[1997] The updated news report is again stored in the database and notified to the terminal.

[1998] Step 16:

[1999] Device:

[2000] The terminal will again announce that an updated news report is available.

[2001] Step 17:

[2002] User:

[2003] Users can view updated news reports to see the latest updates and additions.

[2004] ---

[2005] Through this process, a system is provided that allows users to efficiently track and deeply understand news even in their busy daily lives.

[2006] Example 1

[2007] 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."

[2008] Conventional news article delivery systems lacked sufficient personalization functionality that took into account a user's news browsing history, making it difficult to accurately provide information that users were interested in. Furthermore, there was a lack of a way for users to easily request and obtain follow-up reports or related information on specific news stories. This made it difficult for users to quickly and efficiently obtain the latest and most relevant information.

[2009] 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.

[2010] In this invention, the server includes means for collecting news browsing information, means for storing the collected news browsing information in a database, means for analyzing the stored news browsing information using natural language processing technology and generating a personalized news report using a generative artificial intelligence model, and means for providing the generated personalized news report to a user. This makes it possible to provide a personalized news report that takes into account the user's news browsing history, and to quickly obtain and provide related information in response to a user's request.

[2011] "News viewing information" is viewing data when a user views a news article, and specifically includes metadata such as the news article title, URL, author, viewing date and time, and viewing time.

[2012] A "database" is an information system for efficiently storing, organizing, and retrieving large amounts of data.

[2013] "Natural language processing technology" refers to technology that enables computers to understand, interpret, and generate human language, and includes classification, extraction, and text analysis.

[2014] A "generative artificial intelligence model" is an artificial intelligence technology that learns from large datasets and automatically generates text that is meaningful to users.

[2015] A "personalized news report" is a customized news summary generated based on a user's news browsing history and tailored to their individual interests and preferences.

[2016] A "server" is a computer system that provides services to other computers (clients) on a network.

[2017] A "request" is an action by a user requesting specific information or service, or the content of that request.

[2018] "Searching" is the act of locating information from a database or the Internet based on specified criteria.

[2019] "Related information" is additional data or insight that is relevant to the user's interests or the information they are currently viewing.

[2020] MODE FOR CARRYING OUT THE INVENTION

[2021] The present invention relates to a system for collecting, storing, and analyzing news browsing information, generating personalized news reports based on the collected information, and providing the reports to users. The present invention is implemented by combining a specific software system and infrastructure.

[2022] Collection and storage of news browsing information

[2023] server:

[2024] The server provides an API endpoint for receiving news viewing information from news sites and apps. It receives metadata such as the news article title, URL, author, viewing date and time, and viewing time. The received data is stored for each user ID in the "news_views" table in the MySQL database.

[2025] Device:

[2026] The device collects news articles that users have viewed through news sites and apps. The device generates a unique identifier (user ID) for each user and sends it to the server's API endpoint along with news viewing information.

[2027] User:

[2028] Users browse news articles on a daily basis using news websites and apps, with no special operations required.

[2029] News browsing analysis and report generation

[2030] server:

[2031] The server periodically queries the news viewing information stored in the "news_views" table to obtain the user's browsing history. It analyzes the obtained data and extracts article topics and keywords using a natural language processing library (e.g., SpaCy) that identifies topics and article trends that each user is interested in. Based on the analysis results, it generates prompts for a generative artificial intelligence model (e.g., OpenAI GPT-3.5) to create a personalized news report.

[2032] Generative AI models:

[2033] The generative AI model generates a news report in natural language based on the server's prompts and matches the user's interests.

[2034] Example: "Updates on X: Updates and insights on last week's news story."

[2035] Providing news reports

[2036] Device:

[2037] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability.

[2038] User:

[2039] Users receive notifications on their devices, view news reports, and check the report content to stay up to date on related news and updates.

[2040] Requesting and providing follow-up and related information

[2041] User:

[2042] While viewing a news report, users can request updates or additional information on a particular topic.

[2043] server:

[2044] The server receives user requests and performs searches based on the requests, using the RAG model (e.g., Haystack) to gather relevant, up-to-date information from the Internet and other databases.

[2045] Example prompt: "Create a personalized news report containing the latest news that may be of interest to the user based on their recent browsing history. Specifically, detail new information and related updates about X, as well as future projections."

[2046] Generative AI models:

[2047] The generative AI model then uses the additional information collected to generate an updated news report for the user.

[2048] Device:

[2049] Once the updated news report is ready, you will be notified again on your device.

[2050] User:

[2051] Users receive notifications, view updated news reports, and stay up to date with new developments and relevant information.

[2052] The system allows users to efficiently receive personalized news reports based on their news browsing history, and they can request additional information on specific topics to stay up to date.

[2053] The flow of the identification process in the first embodiment will be described with reference to FIG.

[2054] Processing Steps

[2055] Step 1:

[2056] server:

[2057] The server provides an API endpoint for receiving news browsing information, including metadata such as the news article title, URL, author, browsing date and time, and browsing duration.

[2058] Input: News browsing information sent from your device.

[2059] Data processing: Analyze the received metadata and convert it into a format that can be saved in the database for each user ID.

[2060] Output: Organized metadata stored in a database.

[2061] Step 2:

[2062] Device:

[2063] The device collects news articles that users have viewed through news sites and apps. The device generates a unique identifier (user ID) for each user and sends it to the server's API endpoint along with news viewing information.

[2064] Input: News article information obtained from news sites and apps.

[2065] Data processing: A user ID is assigned and the news article metadata is converted into a format that can be sent to the server.

[2066] Output: News browsing information sent to the server's API endpoint.

[2067] Step 3:

[2068] User:

[2069] Users browse news articles on a daily basis using news websites and apps, with no special operations required.

[2070] Input: User's news article browsing behavior.

[2071] Data processing: None.

[2072] Output: News browsing data based on user browsing behavior.

[2073] Step 4:

[2074] server:

[2075] The server periodically queries the news browsing information stored in the database to obtain the user's browsing history, analyzes the obtained data, and identifies topics and article trends that each user is interested in. It then uses a natural language processing library (e.g., SpaCy) to extract article topics and keywords.

[2076] Input: News browsing information stored in a database.

[2077] Data processing: Using natural language processing technology, topics and keywords from news articles are extracted and user interests are analyzed.

[2078] Output: Analysis results that reflect each user's interests and concerns.

[2079] Step 5:

[2080] server:

[2081] Based on the analysis results, the server generates prompt sentences for a generative AI model (e.g., OpenAI GPT-3.5) to create a personalized news report.

[2082] Input: Results analyzed using natural language processing and user browsing history.

[2083] Data processing: Generate prompt sentences and feed them into a generative AI model to generate personalized news reports.

[2084] Output: The generated personalized news report.

[2085] Action: Example prompt: "Create a personalized news report containing the latest news that may be of interest to the user based on their recent browsing history. Specifically, include new information and related updates about ____, as well as future projections."

[2086] Step 6:

[2087] Device:

[2088] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability.

[2089] Input: A news report sent from the server.

[2090] Data processing: Convert to notification format.

[2091] Output: Notify the user and display a news report.

[2092] Step 7:

[2093] User:

[2094] Users receive notifications on their devices, view news reports, and check the report content to stay up to date on related news and updates.

[2095] Input: A news report posted to the device.

[2096] Data processing: None.

[2097] Output: Confirmed news report viewing behavior of the user.

[2098] Step 8:

[2099] User:

[2100] While viewing a news report, users can request updates or additional information on a particular topic.

[2101] Input: A user request for a news report.

[2102] Data processing: None.

[2103] Output: Request sent.

[2104] Step 9:

[2105] server:

[2106] The server receives user requests and performs searches based on the requests, using the RAG model (e.g., Haystack) to gather relevant, up-to-date information from the Internet and other databases.

[2107] Input: The user's request and the database to search.

[2108] Data processing: Search and retrieve relevant information using the RAG model.

[2109] Output: The most recent relevant information retrieved.

[2110] Step 10:

[2111] Generative AI models:

[2112] The generative AI model then uses the additional information collected to generate an updated news report for the user.

[2113] Input: retrieved related information, previous browsing history, user interests.

[2114] Data processing: Generate prompt sentences and feed them into a generative AI model to generate updated news reports.

[2115] Output: An updated personalized news report.

[2116] Step 11:

[2117] Device:

[2118] Once the updated news report is ready, you will be notified again on your device.

[2119] Input: Updated news report sent from the server.

[2120] Data processing: Convert to notification format.

[2121] Output: Notify user and display update news report.

[2122] Step 12:

[2123] User:

[2124] Users receive notifications, view updated news reports, and stay up to date with new developments and relevant information.

[2125] Input: Updated news report notified to terminal.

[2126] Data processing: None.

[2127] Output: User's viewing behavior for latest information.

[2128] (Application example 1)

[2129] 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."

[2130] In modern society, a huge amount of news information is provided via the Internet, but users face the challenge of efficiently gathering information relevant to them. In particular, the lack of personalized news reports tailored to individual users makes it difficult for users to access news information based on their interests. Furthermore, when users want more information on a particular topic, they have to search manually, which is inconvenient.

[2131] 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.

[2132] In this invention, the server includes means for collecting a user's news browsing history, means for storing the collected news browsing history in a database, means for analyzing the stored news browsing history and generating a prompt sentence using a generative AI model to generate a personalized news report, and means for providing the generated personalized news report to the user. This effectively provides a personalized news report to the user, allowing the user to easily access news information based on their interests. In addition, requested follow-up reports and updates to the news report are automatically searched for, reducing the burden on the user of information gathering.

[2133] "User's news browsing history" is a record of information about news articles that the user has viewed, and includes metadata such as the article title, URL, author, viewing date and time, and viewing time.

[2134] A "database" is a collection of data that systematically stores collected news browsing history and can be retrieved and analyzed later.

[2135] A "generative AI model" is a model that uses artificial intelligence technology to generate prompt sentences based on saved news viewing history and generate personalized news reports.

[2136] A "prompt sentence" is an input sentence used by a generative AI model, which is an instruction sentence to generate an appropriate news report based on the user's news browsing history.

[2137] A "personalized news report" is a personalized news summary generated based on a user's interests and concerns, including related supplemental information and follow-up reports.

[2138] A "request" is a request for information made by a user for more details about a particular news topic or related information.

[2139] A "follow-up" is additional news that includes new information or subsequent developments related to news already provided.

[2140] "Searching" is the process of finding specific information in a database or on the Internet.

[2141] "Update" refers to the reorganization of an existing news report to bring it up to date, for example by adding new information.

[2142] The "means for providing to the user" includes a method and infrastructure for notifying the generated personalized news report to the user's terminal and making it viewable.

[2143] This invention relates to a system that collects, stores, and analyzes a user's news browsing history, and generates and provides personalized news reports based on the results using a generative AI model.

[2144] 1. Collection and storage of news browsing information

[2145] server:

[2146] The server provides an API endpoint to receive news browsing information from news sites and apps, and receives metadata such as the title, URL, author, date and time of browsing, and the browsing time of the news article. The server then stores the received data in a database for each user.

[2147] The main software running on the server is a RESTful API server (e.g., Flask, Django) that provides API endpoints.

[2148] Device:

[2149] The device sets an identifier to identify the user and collects information about news articles the user has viewed on news sites and apps, and sends the collected data to the server's API endpoint.

[2150] A news application (e.g., Android or iOS app) is installed on the device and provides the user interface.

[2151] User:

[2152] Users browse news articles on news websites or apps, and the system automatically collects data based on normal news browsing behavior, so users do not need to take any special action.

[2153] 2. Analyzing news browsing information and generating reports

[2154] server:

[2155] The server periodically analyzes the news browsing information stored in the database, using natural language processing techniques (e.g., spaCy, NLTK) to identify user interests.

[2156] Based on the analysis results, a generative AI model (e.g., OpenAI GPT-3) is used to generate prompt sentences, which are then used to generate a personalized news report.

[2157] The generated personalized news report includes the user's recently viewed articles, related supplemental information, follow-up reports, and more.

[2158] Generation AI:

[2159] The generative AI takes into account the user's browsing history and interests and generates a report in natural language, using prompts such as:

[2160] Generate the following news report based on topics of interest to user 12345:

[2161] Recent news trends indicate an interest in environmental issues, so include content that informs users about the latest environmental news and research findings.

[2162] 3. Providing news reports

[2163] Device:

[2164] Once the generated news report is complete, a notification is sent to the device, using a mobile notification service (e.g., Firebase Cloud Messaging) to notify the user of the report's availability.

[2165] User:

[2166] Users can view news reports on their devices, allowing them to easily review news they have viewed in the past and follow-up reports.

[2167] 4. Requesting and Providing Follow-Up and Related Information

[2168] User:

[2169] If a user is viewing a report and wants to know more about a particular topic or related information, they can make a request.

[2170] server:

[2171] The server receives the request, uses RAG (Retrievable and Generative Model) to search the internet and other databases for relevant up-to-date information, and updates the news report based on the newly found information.

[2172] Generation AI:

[2173] The generator AI then updates the personalized news report again to include any follow-up or related information requested by the user.

[2174] Device:

[2175] Once the updated news report is ready, you will be notified again on your device.

[2176] User:

[2177] Users can view updated news reports to see the latest updates and additional information.

[2178] This system allows users to efficiently receive personalized news reports and easily access news information based on their interests. It also automatically searches for requested follow-up news and updates the reports, significantly reducing the burden on users of information gathering.

[2179] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[2180] Step 1:

[2181] Collection of news browsing information

[2182] Input: News article information viewed by the user on a news site or app (title, URL, author, viewing date and time, viewing time)

[2183] What it does: When a user browses news, the device (e.g., a smartphone) collects the metadata and sends it to the server's API endpoint. The device has a news app installed, which automatically collects data every time the user reads news.

[2184] Step 2:

[2185] Saving news browsing information

[2186] Input: News article information sent to the server

[2187] Specific operation: The server stores the news article information received through the API endpoint in a database. The database management system (e.g., MySQL, PostgreSQL) organizes and stores the news data for each user.

[2188] Step 3:

[2189] Analysis of news browsing information

[2190] Input: News browsing history stored in the database

[2191] How it works: The server periodically analyzes the user's news browsing history using natural language processing techniques (e.g., spaCy, NLTK). The analysis identifies trends in topics and articles that interest the user. The server then identifies the user's areas of interest based on the content, frequency, and viewing time of articles.

[2192] Step 4:

[2193] Generate personalized news reports

[2194] Input: Analysis results (data indicating user interests)

[2195] Specific behavior: The server generates a prompt using a generative AI model (e.g., OpenAI GPT-3) and generates a personalized news report based on it. For example, the prompt could be, "Generate the following news report based on the topics that user 12345 is interested in: Recent news trends indicate that they are interested in environmental issues. Therefore, please include content that informs the user about the latest environmental news and research findings."

[2196] Step 5:

[2197] Providing news reports

[2198] Input: Generated personalized news report

[2199] Specific operation: The server sends the generated news report to the device and notifies the user via a mobile notification service (e.g., Firebase Cloud Messaging). The user can then view the report on their device.

[2200] Step 6:

[2201] Requesting and providing follow-up and related information

[2202] Input: User requests (follow-ups or related information on a particular topic)

[2203] How it works: When a user wants to know the latest updates or related information on a particular topic, they make a request from their device. The server receives the request and uses RAG (Search and Generative Model) to search the internet and other databases for relevant updates. It then uses the generative AI model again to generate an updated, personalized news report.

[2204] Step 7:

[2205] Providing updated news reports

[2206] Input: Updated personalized news report

[2207] Specific operation: The server sends the updated news report back to the device and notifies the user via the mobile notification service. The user can then view the updated news report and check the latest updates and additional information.

[2208] The above is the overall processing flow of the system, and at each step it is possible to efficiently collect and analyze news information from users and provide personalized information.

[2209] 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.

[2210] The present invention relates to a system for generating and providing news reports optimized for users by recognizing, collecting, storing, and analyzing a user's news browsing history and emotions. Hereinafter, an embodiment of the present invention will be described in detail.

[2211] 1. Collection and storage of news browsing and sentiment data

[2212] server

[2213] The server provides an API endpoint to receive news browsing information and sentiment data from news sites and apps.

[2214] It receives metadata such as the news article title, URL, author, viewing date and time, and viewing time, as well as the user's emotion data recognized by the emotion engine.

[2215] Save the received data in the database for each user.

[2216] Terminal

[2217] The device sets a user identifier and collects information about news articles the user has viewed on news sites or apps, as well as emotion data recognized by the emotion engine.

[2218] Send the collected data to the server's API endpoint.

[2219] User

[2220] Users use news sites and apps to view news articles.

[2221] Users basically operate the app in the same way as they normally would when browsing the news, but the emotion engine uses facial and emotion recognition technology to record the user's expressions and reactions.

[2222] 2. Analyzing news browsing information and sentiment data and generating reports

[2223] server

[2224] The server periodically analyzes the news browsing history and sentiment data stored in the database.

[2225] It uses natural language processing and sentiment analysis technologies to identify topics and article trends that interest users, as well as their thoughts when reading the news.

[2226] Based on the analysis results, generative AI is used to generate personalized news reports.

[2227] Generation AI

[2228] Generative AI generates reports in natural language based on the user's browsing history and sentiment.

[2229] Reports include users' recently viewed articles, related supplemental information, and adjustments to the content and tone of news based on sentiment.

[2230] Example: "Latest news on XX: Updates on last week's news and recommendations based on sentiment analysis."

[2231] 3. Providing news reports

[2232] Terminal

[2233] Once the generated news report is complete, a notification will be sent to your device.

[2234] The terminal notifies the user of the availability of the report.

[2235] User

[2236] The user receives a notification from the device and views the news report.

[2237] Review reports and quickly review previous news and follow-ups to deepen your understanding.

[2238] 4. Requesting and Providing Follow-Up and Related Information

[2239] User

[2240] While viewing a report, users can request further updates or related information on a particular topic.

[2241] server

[2242] The server receives the request and uses the RAG model to search the internet and other databases for relevant follow-up information.

[2243] Update news reports with newly discovered information.

[2244] Generation AI

[2245] The generative AI then updates the personalized news report again, adding follow-up and related information.

[2246] Terminal

[2247] Once the updated news report is ready, you will be notified again on your device.

[2248] User

[2249] Users can view updated news reports to see the latest updates and additions.

[2250] Specific examples

[2251] 1. User A:

[2252] I was browsing an article about climate change on a news app.

[2253] The emotion engine analyzes user A's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article.

[2254] 2. Terminal:

[2255] News browsing information and emotion data are sent to the server.

[2256] 3. Server:

[2257] The received data is collected and stored in a database.

[2258] The data is analyzed periodically and a sentiment engine generates personalized news reports based on the collected data.

[2259] 4. Generation AI:

[2260] A news report was created that focused on "climate change," which User A was interested in.

[2261] Reflect sentiment data and include many articles with a positive tone.

[2262] 5. Terminal:

[2263] When the news report is complete, User A is notified.

[2264] 6. User A:

[2265] Open the report to see the latest news you've read and follow-up updates.

[2266] This embodiment provides a news experience that takes into account the user's emotions and leads to greater understanding and interest.

[2267] The processing flow will be explained below.

[2268] Explanation of the specific processing of the program

[2269] ---

[2270] Step 1:

[2271] User:

[2272] A user opens a news site or app and reads a news article.

[2273] Step 2:

[2274] Device:

[2275] The device uses the user identifier to record metadata such as the title, URL, and viewing time of news articles viewed by the user.

[2276] In addition, the device's built-in emotion engine analyzes emotional data from the user's facial expressions and voice in real time, recording the type of emotion (joy, surprise, sadness, etc.) and its intensity.

[2277] Step 3:

[2278] Device:

[2279] The device makes an API request to send the recorded news viewing information and emotion data to the server in real time.

[2280] Step 4:

[2281] server:

[2282] The server processes the news browsing information and emotion data received from the terminal and stores them in a database together with the corresponding user identifier.

[2283] Step 5:

[2284] server:

[2285] Periodically (eg, weekly), the server analyzes the news browsing history and sentiment data stored in the database.

[2286] Step 6:

[2287] server:

[2288] The server analyzes browsing history and emotional data using natural language processing and sentiment analysis techniques to identify user interests.

[2289] For example, if a user consistently expresses enjoyment or interest in a particular topic, that topic will be given more weight.

[2290] Step 7:

[2291] server:

[2292] The server uses generative AI to generate personalized news reports based on the analysis results.

[2293] Report content is optimized based on user interests and emotional responses.

[2294] Step 8:

[2295] server:

[2296] The generated news reports are stored in a database and are ready to be notified to the device.

[2297] Step 9:

[2298] Device:

[2299] The terminal notifies the user that a news report is available.

[2300] Step 10:

[2301] User:

[2302] The user receives a notification from the device and views the news report.

[2303] Step 11:

[2304] User:

[2305] Users can view news reports and see selected articles and follow-up updates based on sentiment data.

[2306] Request any follow-up or related information that may be of interest to you.

[2307] Step 12:

[2308] Device:

[2309] The device makes an API request to send the request from the user to the server.

[2310] Step 13:

[2311] server:

[2312] Based on the received request, the server uses the RAG model to search the internet and other databases for relevant follow-up information.

[2313] Step 14:

[2314] server:

[2315] Based on the search results, the generating AI will update the personalized news report again, adding follow-up and related information.

[2316] Step 15:

[2317] server:

[2318] The updated news report is again stored in the database and notified to the terminal.

[2319] Step 16:

[2320] Device:

[2321] The terminal will again announce that an updated news report is available.

[2322] Step 17:

[2323] User:

[2324] Users can view updated news reports to see the latest updates and additions.

[2325] ---

[2326] Through this process, users can receive more personalized news reports that take into account their emotions when reading the news. In particular, the collection and analysis of emotional data allows the content and tone of the news to be optimized for users, eliciting deeper understanding and interest.

[2327] Example 2

[2328] 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."

[2329] Current news delivery systems achieve some degree of personalization based on users' news browsing history, but do not adequately consider users' emotions when providing news. As a result, users are unable to efficiently obtain news that matches their interests and emotions, resulting in a poor news browsing experience. Furthermore, when users request follow-up or related information on a specific topic, responses are often delayed, further worsening the user experience.

[2330] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting a user's news browsing history and user emotion data, means for storing the collected news browsing history and user emotion data in a database, means for analyzing the stored news browsing history and emotion data to generate a personalized news report, and means for providing the generated personalized news report to the user. This makes it possible to provide personalized news based on the user's emotion data, improving the quality of the news browsing experience. Furthermore, follow-up reports and related information requested by the user can be quickly provided, thereby increasing user satisfaction.

[2331] "News viewing history" is information about news articles viewed by the user, and specifically includes metadata such as the article title, URL, author, viewing date and time, and viewing time.

[2332] "Emotion data" refers to information about the emotional state recognized from the user's facial expressions and behavior, and indicates data that expresses emotions such as joy, surprise, and sadness.

[2333] "Database" refers to a computer system for systematically storing and managing collected news viewing history and emotion data.

[2334] "Analysis" refers to the process of analyzing news browsing history and emotional data stored in a database using machine learning and natural language processing technologies to identify trends in users' interests and emotions.

[2335] A "personalized news report" is a news report customized based on a user's news viewing history and emotional data, and includes content optimized for the user's interests and emotions.

[2336] "Generative AI" refers to systems that use artificial intelligence technology to generate personalized news reports, specifically natural language processing models (e.g., GPT-4).

[2337] A "follow-up" is any new information provided in relation to an already reported news story.

[2338] "Related information" refers to supplementary information or topics related to the news article the user viewed.

[2339] "Server" refers to the central computer system that collects, stores, and analyzes users' news browsing history and sentiment data to generate and provide personalized news reports.

[2340] A "terminal" is a device that a user uses to view news articles, and includes a smartphone, tablet, PC, etc.

[2341] The present invention relates to a system for collecting, storing, and analyzing a user's news browsing history and emotion data, and generating and providing a news report optimized for the user. Hereinafter, an embodiment of the invention will be described in detail.

[2342] Collection and storage of news browsing and sentiment data

[2343] server

[2344] The server provides an API endpoint for receiving news browsing information and emotion data from news sites and apps. Specifically, it receives metadata such as the news article title, URL, author, viewing date and time, and viewing time, as well as user emotion data recognized by the emotion engine. The received data is stored in a database for each user. This database is built using, for example, MySQL or MongoDB.

[2345] Terminal

[2346] The device sets a user identifier and collects information about news articles the user has viewed on news sites and apps, as well as emotion data recognized by an emotion engine. The collected data is periodically sent to the server's API endpoint. Emotion data is collected using the device's camera and an emotion analysis engine (e.g., Face API).

[2347] User

[2348] Users browse news articles on news websites or apps. The user's operations are basically the same as when browsing the news normally, but the emotion engine uses facial and emotion recognition technology to record the user's expressions and reactions.

[2349] Analyzing news browsing and sentiment data and generating reports

[2350] server

[2351] The server periodically analyzes the news browsing history and sentiment data stored in the database. It utilizes natural language processing (NLTK and spaCy) and sentiment analysis (IBM Watson and Microsoft Azure sentiment analysis APIs) to identify topics and article trends that interest users, as well as their thoughts while browsing the news. Based on the analysis results, a generative AI model (e.g., GPT-4) is used to generate personalized news reports.

[2352] Generation AI

[2353] Based on a user's browsing history and sentiment, the generative AI generates a natural language report that includes the user's recently viewed articles, relevant supplemental information, and tailoring the content and tone of the news based on sentiment.

[2354] Prompt Sentence Examples

[2355] "Based on User A's recent news about climate change, create a summary of articles with a positive tone that reflects surprising sentiment data."

[2356] Providing news reports

[2357] Terminal

[2358] Once the generated news report is complete, a notification is sent to the terminal, which notifies the user of the report's availability, ensuring that it remains accessible to the user.

[2359] User

[2360] Users receive notifications on their devices to view news reports, and can review the reports to gain more detailed information about previously viewed news stories and follow-ups.

[2361] Requesting and providing follow-up and related information

[2362] User

[2363] While viewing a report, users can request follow-up or related information on a specific topic. This request is made through the News app.

[2364] server

[2365] The server receives the request and uses the Retrieval-Augmented Generation (RAG) model to search the Internet and other databases for relevant follow-up news, updating the news report with the newly discovered information.

[2366] Generation AI

[2367] The generative AI then updates the personalized news report again, adding follow-up and related information.

[2368] Terminal

[2369] Once the updated news report is complete, the user will receive another notification on their device, and they can view the updated news report to see any updates or additional information.

[2370] Specific examples

[2371] 1. User A:

[2372] When a user reads an article about climate change in a news app, the emotion engine analyzes the user's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article.

[2373] 2. Terminal:

[2374] News browsing information and emotion data are sent to a server.

[2375] 3. Server:

[2376] The received data is collected and stored in a database. The data is periodically analyzed and a sentiment engine generates personalized news reports based on the collected data.

[2377] 4. Generation AI:

[2378] Create a news report that focuses on "climate change," which User A is interested in. Reflect the sentiment data and include many articles with a positive tone.

[2379] 5. Terminal:

[2380] When the news report is complete, User A is notified.

[2381] 6. User A:

[2382] Open the report to see the news you recently read and follow-ups. This embodiment provides a news experience that takes into account the user's emotions and provides a deeper understanding and interest.

[2383] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2384] Step 1: User Views News Article

[2385] A user browses a news article using a news website or app. Specifically, the user opens a news app and selects and reads an article about climate change.

[2386] Input: Select the article you want to view on the news app

[2387] Output: Transmission of viewed article information

[2388] Step 2: Collecting news browsing information and emotion data from the device

[2389] The device automatically collects metadata (title, URL, author, date and time of viewing, duration, etc.) of news articles viewed by the user, using the data collection function within the news app.

[2390] The device uses its built-in camera and sensors to collect the user's facial expressions and emotional data (happiness, surprise, sadness, etc.) through an emotion engine (e.g., Face API).

[2391] Input: Browsing information in news apps, emotion data from cameras and sensors

[2392] Output: A batch of collected news metadata and sentiment data

[2393] Step 3: Send data from the device to the server

[2394] The device sends the collected news browsing information and sentiment data to a server through a pre-configured API endpoint, using an internet connection for this communication.

[2395] Input: Batches of news metadata and sentiment data collected on the device

[2396] Output: Confirmation of data transmission to the server

[2397] Step 4: Server saves data

[2398] The server organizes the received news browsing information and emotion data for each user and stores them in a database (e.g., MySQL or MongoDB) using a database management system.

[2399] Input: A batch of news metadata and sentiment data sent from the device

[2400] Output: Save confirmation message

[2401] Step 5: Periodic analysis of data by the server

[2402] The server periodically analyzes the news browsing information and sentiment data in the database, using natural language processing technologies (e.g., NLTK, spaCy) and sentiment analysis technologies (e.g., IBM Watson, Microsoft Azure sentiment analysis API) to identify trends in user interests and sentiment.

[2403] Input: News metadata and sentiment data in a database

[2404] Output: Analysis result report

[2405] Step 6: Generative AI creates personalized news reports

[2406] Based on the analysis results, the server sends prompts to a generative AI (e.g., GPT-4) to generate a personalized news report, which is written in natural language based on the user's browsing history and emotions.

[2407] Input: Analysis result report and prompt statement

[2408] Output: Personalized news report

[2409] Prompt Sentence Examples

[2410] "Based on User A's recent news about climate change, create a summary of articles with a positive tone that reflects surprising sentiment data."

[2411] Step 7: News report notification on your device

[2412] The server sends the generated news report to the terminal, which displays a notification to the user that a new news report is available.

[2413] Input: Personalized News Report

[2414] Output: Notification message to terminal

[2415] Step 8: User Views News Report

[2416] Users receive notifications from their devices and open and view news reports, which include recently viewed stories, follow-ups, and tone adjustments based on emotion.

[2417] Input: Notifications from your device

[2418] Output: Content of the news report

[2419] Step 9: User requests follow-up or related information

[2420] While viewing a news report, users interact with an interface to request updates or related information on a particular topic.

[2421] Input: Request button in news report

[2422] Output: Request information

[2423] Step 10: Server searches for relevant information and updates the report

[2424] The server receives user requests, uses the RAG (Retrieval-Augmented Generation) model to search the Internet and other databases for related or follow-up information, and updates the news report based on the newly found information.

[2425] Input: Request information from the user

[2426] Output: Updated news report

[2427] Step 11: Regenerating the news report with generative AI

[2428] The generative AI generates a personalized news report again based on the new information.

[2429] Input: List of newly found information

[2430] Output: Updated personalized news report

[2431] Step 12: Update report notification to device

[2432] The server retransmits the updated news report to the terminal, which displays a notification to the user that the report has been updated.

[2433] Input: Updated news report

[2434] Output: Notification message to terminal

[2435] Step 13: User Views Update News Report

[2436] Users can view updated news reports to see the latest updates and additions.

[2437] Input: Notifications from your device

[2438] Output: Updated news report content

[2439] Through these steps, the system can personalize users' news browsing experience and provide optimal news reports that reflect emotional data.

[2440] (Application example 2)

[2441] 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."

[2442] Conventional news delivery systems personalized news based on users' browsing history, but it was difficult to consider users' emotional responses. This made it difficult to provide optimal news reports that would deepen users' interest and understanding. Furthermore, there was a lack of mechanisms for generating reports that accurately reflected users' emotions toward specific news articles. To address these issues, it is necessary to collect and analyze not only news browsing history but also user emotional data during browsing to enrich the user experience. Furthermore, there is a need for more intuitive, real-time data collection and reporting using wearable devices such as smart glasses.

[2443] 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.

[2444] In this invention, the server includes means for collecting a user's news viewing history and emotional data, means for storing the collected news viewing history and emotional data in a database, means for analyzing the stored news viewing history and emotional data to generate a personalized news report, means for providing the generated personalized news report to the user, means for collecting the user's news viewing and emotional data in real time through the smart glasses, means for analyzing the user's emotional data, and means for providing the user with a function for viewing the news report using voice commands or gestures. This enables the generation and provision of more personalized news reports based on the user's emotional responses. Furthermore, the user's daily data can be collected naturally through the smart glasses, allowing the user to intuitively view the news report, thereby improving the user experience.

[2445] "News browsing history" is a collection of metadata such as the title, URL, author, browsing date and time, and browsing time of news articles that a user has viewed.

[2446] "Emotion data" is information on emotions that are recorded and recognized by the emotion engine while the user is browsing the news.

[2447] A "database" is a collection that stores collected news viewing history and emotional data and manages it so that it can be searched and analyzed as needed.

[2448] A "personalized news report" is a news summary optimized for a user based on the user's news viewing history and emotional data.

[2449] "Smart glasses" are wearable devices equipped with functions such as a camera and a display, and can recognize the user's gaze direction and facial expressions and display information.

[2450] "Means for collecting data in real time" refers to a method for capturing the user's facial expressions and reactions while viewing the news in real time through the camera in the smart glasses and instantly transmitting them to a server.

[2451] "Ability to browse using voice commands and gestures" refers to an interface that allows a user to navigate and browse news reports using voice or body movements.

[2452] The "means for analyzing user emotional data" is a technology that uses collected emotional data to evaluate the emotions felt by users while browsing news and changes in those emotions.

[2453] This invention is a system that collects, stores, and analyzes users' news browsing history and emotional data to generate and provide personalized news reports.

[2454] 1. Collection and storage of news browsing and sentiment data

[2455] server

[2456] The server provides an API endpoint for receiving news browsing information and emotion data from news sites and applications. It receives metadata such as the title, URL, author, viewing date and time of news articles, as well as user emotion data detected by the emotion engine, and stores this data in a database for each user.

[2457] Terminal

[2458] The device collects information about news articles viewed by users on news sites and apps through the smart glasses, as well as emotional data recognized by the emotion engine, and sends this data to the server's API endpoint.

[2459] User

[2460] Users use smart glasses to browse news articles. While browsing, the emotion engine uses facial and emotion recognition technology to record the user's facial expressions and reactions. The data is then sent to a server in real time.

[2461] 2. Analyzing news browsing information and sentiment data and generating reports

[2462] server

[2463] The server periodically analyzes the news browsing history and sentiment data stored in the database to identify trends in topics and articles that users are interested in. This is done using natural language processing and sentiment analysis technologies, and uses generative AI to generate personalized news reports.

[2464] Generation AI

[2465] Based on the user's browsing history and emotional data, the generative AI creates a news report that can be viewed in-car or via gesture control, including recently viewed articles, relevant supplemental information, and an adjustment to the tone of the news article based on the emotional data.

[2466] 3. Providing news reports

[2467] Terminal

[2468] Once the generated news report is complete, it notifies the user through the smart glasses, who can then use voice commands or gestures to view or interact with the report.

[2469] User

[2470] Users receive notifications, access personalized news reports, and easily review previously viewed news stories and follow-up updates. If they want more updates or related information on a particular topic, they can request that information.

[2471] 4. Requesting and Providing Follow-Up and Related Information

[2472] User

[2473] While viewing a report, users can request follow-up or related information on a particular topic.

[2474] server

[2475] The server receives the request and uses the Retrieval-Augmented Generation (RAG) model to search for relevant follow-up information from the Internet and other databases, updating the news report based on the information found.

[2476] Generation AI

[2477] The generative AI is then personalized again to update the news report, adding any necessary follow-up or related information.

[2478] Terminal

[2479] Once the updated news report is ready, another notification will be sent to the smart glasses.

[2480] User

[2481] Users can view updated news reports to see the latest updates and additions.

[2482] Example 1

[2483] A user browses an article about climate change in a news app. The emotion engine analyzes the user's facial expressions and collects emotional data (e.g., surprise, interest) when reading the article. The news browsing information and emotional data are sent to the server via the smart glasses.

[2484] Prompt Sentence Examples

[2485] List of news recently viewed by user 123 (including metadata)

[2486] In addition, based on the emotional data (Happy, Sad, Angry, etc.) when viewing each news article,

[2487] Generate personalized news reports.

[2488] The report includes a summary of your recently viewed articles along with information on topics of interest.

[2489] Provide a list of recommended articles.

[2490] If possible, frame your news in a positive tone.

[2491] This format allows for a deeper, more engaging news experience that takes into account the user's emotions.

[2492] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2493] Step 1:

[2494] The device collects the user's news viewing information and emotion data through the smart glasses. The input is real-time video footage of the news being viewed and the metadata of the news article. The video footage captured by the smart glasses' camera is input to a face recognition and emotion recognition engine to analyze the user's emotion (e.g., surprise, interest, etc.). The output is emotion data corresponding to the metadata of the news article.

[2495] Step 2:

[2496] The device sends the collected news browsing information and emotion data to the server. The input is the news article metadata and emotion data collected in step 1. This is sent as an HTTP request to the server's API endpoint. The output is a notification of successful data transmission to the server.

[2497] Step 3:

[2498] The server stores the received news browsing information and emotion data in a database. The input is the metadata and emotion data of the news article received as an HTTP request. These are stored in the database as records corresponding to each user. The output is a confirmation notification of the record stored in the database.

[2499] Step 4:

[2500] The server periodically analyzes the stored news browsing history and sentiment data. The input is the metadata and sentiment data of news articles stored in the database. A data analysis engine is used to identify user interests and sentiment trends and create the basis for generating personalized news reports. The output is the analysis results data.

[2501] Step 5:

[2502] The server uses a generative AI to generate a personalized news report. The input is the analysis result data obtained in step 4. The generative AI uses natural language processing technology to create a news report based on the user's interests and emotions. The output is a completed personalized news report.

[2503] Step 6:

[2504] The server notifies the terminal of the generated news report. The input is the completed personalized news report, which is sent to the terminal and notified to the user. The output is a notification display from the terminal.

[2505] Step 7:

[2506] The device allows users to browse news reports using voice commands or gestures. The input is a personalized news report sent from the server and the user's voice or gesture instructions. The device displays the news report through the interface of the smart glasses, and displays and changes the content based on the user's actions. The output is the content viewed or manipulated by the user.

[2507] Step 8:

[2508] While viewing a report, a user requests follow-up or related information on a specific topic. The input is a request made by the user through voice commands or gestures. The device receives this request and sends it to the server. The output is a notification of the request to the server.

[2509] Step 9:

[2510] The server receives requests and searches for relevant follow-up information. The input is a request from a user. The RAG model (Retrieval-Augmented Generation model) is used to search for relevant information from the Internet and other databases. The output is the relevant information found.

[2511] Step 10:

[2512] The server updates the news report using a generative AI. The input is the relevant information found and the existing news report. The generative AI adds the relevant information and creates a personalized news report again. The output is the updated news report.

[2513] Step 11:

[2514] The server notifies the terminal of the updated news report. The input is the updated news report. It is sent to the terminal and notified to the user. The output is an update notification display from the terminal.

[2515] Step 12:

[2516] The user browses the updated news report to check the latest updates and additional information. The input is the user's operation upon receiving the update notification. The user checks the news report through the smart glasses and obtains the necessary information. The output is the latest information obtained by the user.

[2517] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[2518] 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.

[2519] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[2520] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[2521] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2522] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[2523] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[2524] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[2525] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[2526] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[2527] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[2528] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[2529] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[2530] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[2531] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored. 【253...

Claims

1. A means for collecting a user's news browsing history; A means for storing the collected news browsing history in a database; means for analyzing the stored news browsing history to generate a personalized news report; means for providing the generated personalized news report to a user; A system including:

2. 10. The system of claim 1, further comprising means for retrieving and adding to the news report any follow-ups requested by the user.

3. 10. The system of claim 1, further comprising means for providing additional information related to the generated news report.

Images