System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting information from the Internet based on an interest category of a user; means for summarizing the collected information using a natural language processing model; means for notifying a terminal of the user of the summarized information; and means for collecting feedback from the user and adjusting the natural language processing model for a next round of summary generation.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] In today's society, there is a vast amount of information available, including web articles, social media posts, news feeds, and videos. Users must expend a great deal of time and effort to find the information that is most relevant to them. This situation is particularly serious for busy modern people, and there is a need for systems that can efficiently collect and summarize information. Furthermore, current information gathering systems often fail to address users' individual interests and preferences, resulting in the inability to provide appropriate information. [Means for solving the problem]

[0005] The present invention solves the above-mentioned problems by providing a system that includes a means for collecting information from the Internet based on a user's interest categories and a means for summarizing the collected information using a natural language processing model. Specifically, the system uses a crawler to collect information from specific websites and social media platforms, and then summarizes the collected information using a deep learning model. Additionally, the system includes a means for notifying the user of the summarized information and collecting user feedback. This allows the natural language processing model to be adjusted the next time a summary is generated, allowing the system to provide information more suited to the user's preferences and interests.

[0006] Okay, below are definitions of important terms included in the claims.

[0007] "Interest categories" refer to the areas or topics of information that a user is interested in, and are used to determine the subject of information collection.

[0008] "Internet collection methods" refers to systems or technologies that automatically collect data related to user-specified interest categories.

[0009] A "natural language processing model" refers to a machine learning model or algorithm that analyzes collected text data, extracts important information, and generates summaries.

[0010] "Means for summarizing" refers to a system or technology that summarizes large amounts of information so that users can understand it quickly.

[0011] "Means for notification" refers to the system or protocol for transmitting the generated summary to the user's terminal.

[0012] "Means of collecting feedback" refers to the mechanisms and methods for collecting user ratings and opinions and using them to improve the system.

[0013] A "crawler" is a program or algorithm that automatically crawls and collects information from a particular website or social media platform.

[0014] A "deep learning model" is a type of natural language processing model that uses a multi-layer neural network to analyze text data and generate summaries.

[0015] A "terminal" is an electronic device that a user uses to receive and view information, including smartphones, tablets, and personal computers. [Brief explanation of the drawings]

[0016] [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

[0017] 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.

[0018] First, the terms used in the following description will be explained.

[0019] 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).

[0020] 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.

[0021] 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.

[0022] 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.

[0023] 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."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 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.

[0027] 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).

[0028] 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.

[0029] 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.

[0030] 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.

[0031] 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.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] 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."

[0037] The present invention provides a system for collecting information from the Internet based on the user's interest categories, summarizing the information, and providing it to the user. Specific embodiments of the present invention are described below.

[0038] Overall system overview

[0039] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means of notifying the user of the summarized information, and a mechanism for collecting user feedback and adjusting the system.

[0040] Information gathering

[0041] Server: Launches a crawler that collects data from specific websites and social media platforms based on the user's interest categories. The crawler periodically visits and collects new content to store in the database.

[0042] Example: If user A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[0043] Summary Generation

[0044] Server: The collected data is fed into a natural language processing model (e.g., a deep learning model). This model extracts important information from the text and generates a summary. The model is adjusted based on user feedback to improve its accuracy.

[0045] Example: When a long technical article is collected, it is summarized using an NLP model to extract only the main points, for example, a 5000-character article is converted into a 300-character summary.

[0046] Data provision

[0047] Server: Summarized information is organized by user and sent based on notification schedule.

[0048] On your device: The user's smartphone or tablet will receive a push notification that will appear on their screen. The user can then view the notification and read more if they are interested.

[0049] Example: If user A has configured their device to receive a summary at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information.

[0050] Feedback and Learning

[0051] Users: After viewing the summary, users can provide feedback on whether the information provided was useful, with options such as "helpful," "not helpful," or "I'd like more information."

[0052] Server: The server collects feedback and stores it in a database. This feedback data is used to improve the performance of the natural language processing model the next time a summary is generated.

[0053] Example: If user A gives feedback that the summary they received was very useful, that feedback can be used as training data for a model that can provide useful summaries to other users with similar interests.

[0054] ---

[0055] This allows users to efficiently gather important information from vast amounts of information and understand it in a short time. Furthermore, by utilizing user feedback, the system's accuracy can be continuously improved, enabling more personalized information to be provided.

[0056] The processing flow will be explained below.

[0057] ---

[0058] Step 1: User Registration and Setup

[0059] Users: When using the service for the first time, they create an account, enter their email address and password, and select the categories of their interests (e.g., "Technology," "Health," "Entertainment").

[0060] Server: Receives user input and stores user information and selected interest categories in a database.

[0061] Step 2: Gather information

[0062] Server: Based on the user's interest category, the query is passed to the crawler that is launched and configured to collect relevant information from specific websites and social media platforms.

[0063] Server: The crawler visits configured sites and gathers new content, such as articles from news sites or social media posts.

[0064] Server: Stores the collected data in a database.

[0065] Step 3: Summary generation

[0066] Server: Inputs the collected data into a natural language processing (NLP) model. The server invokes this NLP model to analyze the collected text data.

[0067] Server: The NLP model extracts important information from the text and generates a summary, for example, summarizing a 5,000-character article into 300 characters.

[0068] Server: Organizes the generated summaries for each user and stores them in a database.

[0069] Step 4: Provide data

[0070] Server: Once the summary is generated, it sends the summarized information to the user's device based on the user's notification schedule.

[0071] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[0072] Step 5: Feedback and learning

[0073] Users: After reviewing the summarized information, they can provide feedback by selecting from options such as "helpful," "not helpful," or "I'd like more information."

[0074] Server: Receives the provided feedback and stores it in a database.

[0075] Server: Uses the feedback data to adjust the parameters of the natural language processing model to improve the accuracy of summary generation, thereby improving the model so that the next summary is more suited to the user's preferences.

[0076] ---

[0077] This processing step allows users to efficiently gather information specific to their interests and quickly grasp the main points, while feedback is used to continuously improve the accuracy and user experience of the overall system.

[0078] Example 1

[0079] 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."

[0080] In today's Internet society, users are overwhelmed by the vast amount of information available, making it difficult for them to efficiently gather and understand important information tailored to their interests and needs. Furthermore, few systems provide feedback on the appropriateness of collected information to improve the quality of subsequent information provision. As a result, users often waste time and effort. To solve this problem, a method is needed to collect information based on users' interests, summarize it, notify them at the appropriate time, and utilize feedback to improve the accuracy of the system.

[0081] 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.

[0082] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for storing the collected data in a database, means for transmitting the summarized information based on a notification schedule, and means for collecting user feedback and adjusting the natural language processing model for the next summary generation. This allows users to efficiently collect important information and understand it in a short time. Furthermore, by utilizing user feedback, the accuracy of the system can be continuously improved, enabling more personalized information to be provided.

[0083] "User" means any person or entity that uses the System to gather interest-based information and receive summarized information.

[0084] "Interest categories" are classifications that indicate themes or topics that interest a user.

[0085] The "Internet" is a global network system that connects computers and servers around the world.

[0086] "Information collection methods" are the processes and tools used to obtain data from specific websites or social media platforms on the internet.

[0087] A "natural language processing model" is an artificial intelligence technology for understanding and analyzing text data, and specifically includes deep learning models.

[0088] A "summarization tool" is a process or tool that condenses collected information and extracts its main points.

[0089] A "terminal" is an electronic device used by a user to receive and view information, and specifically includes a smartphone or tablet.

[0090] "Means of Notification" means the method or mechanism for notifying the user of the summarized information, including push notifications.

[0091] "Feedback collection methods" are the methods and processes used to obtain user responses and ratings.

[0092] A "database" is a system for systematically storing and managing collected data and abstract data.

[0093] A "notification schedule" is a plan or setting that controls when information is notified to a user.

[0094] A "server" is a computing system for performing the processes of information gathering, summary generation, notification, feedback gathering, and data storage.

[0095] The present invention provides a system for collecting information from the Internet based on the user's interest categories, summarizing the information, and providing it to the user. Specific embodiments of the present invention are described below.

[0096] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, and a mechanism for collecting user feedback and adjusting the system.

[0097] Information gathering

[0098] The server launches a crawler based on the user's interest categories to collect data from specific websites and social media platforms. This crawler uses a data collection library, such as Scrapy. The server stores the collected data in a database (such as MySQL or MongoDB).

[0099] Specific examples

[0100] If User A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[0101] Summary Generation

[0102] The server inputs the collected data into a natural language processing model (e.g., a deep learning model like GPT-4 or BERT). The model extracts important information from the text and generates a summary. It then adjusts the model based on user feedback to improve its accuracy.

[0103] Specific examples

[0104] When long technical articles are collected, they are summarized using a natural language processing model to extract only the main points, for example, a 5,000-character article is converted into a 300-character summary.

[0105] Data provision

[0106] The server organizes the summarized information for each user and sends it based on a notification schedule, using scheduler software (e.g., a cron job), and the information is sent to the user's device via a push notification service (e.g., Firebase Cloud Messaging).

[0107] Specific examples

[0108] If user A has configured their device to receive a summary at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information.

[0109] Feedback and Learning

[0110] After reviewing the summarized information, users can provide feedback on whether the information was useful or not. Feedback consists of options such as "helpful," "not helpful," or "I'd like more information." The server stores this feedback in a database and uses it to improve the performance of the natural language processing model the next time a summary is generated.

[0111] Specific examples

[0112] If user A gives feedback that the summary they received was very useful, that feedback can be used as training data for the model, allowing it to provide useful summaries to other users with similar interests.

[0113] Example prompts to input to the generative AI model

[0114] "I would like to design a system that collects information from the Internet based on interest categories set by the user and summarizes that information. What are the specific processing steps and what hardware and software are used at each step? In particular, could you please provide more details about the steps of information collection, summary generation, data provision, and user feedback and learning?"

[0115] As a result, the present invention allows users to efficiently gather important information from vast amounts of information and understand it in a short time. Furthermore, by utilizing user feedback, the system's accuracy can be continuously improved, enabling more personalized information to be provided.

[0116] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0117] Step 1: User Interest Category Setting

[0118] User: The user sets the category of interest, for example, choosing a theme such as "Technology" or "Health."

[0119] Input: User selected interest categories.

[0120] Output: Interest category information passed to the server.

[0121] Step 2: Start the crawler

[0122] Server: The server launches a crawler based on the user's interest categories. The crawler is configured to visit specified websites and social media platforms. This crawler uses a library such as Scrapy.

[0123] Input: User interest category information.

[0124] Output: Collected article and post data.

[0125] Specific behavior:

[0126] The server looks up the user's interest categories and generates an appropriate URL list.

[0127] A crawler is launched based on the generated URL list to collect new articles and posts.

[0128] Step 3: Save to database

[0129] Server: Stores the collected data in a database, such as MySQL or MongoDB.

[0130] Input: The data retrieved by the crawler (articles, posts).

[0131] Output: Collected data stored in a database.

[0132] Specific behavior:

[0133] The crawler inserts the collected data into a database along with information such as text, URL, date and time.

[0134] Once the database has been saved, proceed to the next step.

[0135] Step 4: Summarize the data

[0136] Server: Inputs the collected data into a natural language processing (NLP) model, such as a deep learning model like GPT-4 or BERT.

[0137] Input: Article text extracted from the database.

[0138] Output: A summary sentence generated by the NLP model.

[0139] Specific behavior:

[0140] The server extracts new articles from the database and feeds the text data into the NLP model.

[0141] An NLP model extracts important information and generates a summary.

[0142] The generated summaries are stored in a database.

[0143] Step 5: Set up a notification schedule

[0144] Server: Organize the summarized information for each user and set up notifications to be sent based on a notification schedule, such as using a cron job.

[0145] Input: The notification schedule set by the user.

[0146] Output: Scheduled notification jobs.

[0147] Specific behavior:

[0148] The server refers to the user's notification schedule and sets the timing for sending the summary data.

[0149] When the scheduler reaches the set time, a push notification is triggered.

[0150] Step 6: Sending push notifications

[0151] Server: Posts summarized information.

[0152] Device: The user's smartphone or tablet receives the push notification and displays the information. Push notifications are sent using Firebase Cloud Messaging (FCM) or similar.

[0153] Input: Summary data, Notification schedule.

[0154] Output: The push notification sent to the user device.

[0155] Specific behavior:

[0156] The server sends a push notification, and the user device receives it.

[0157] A notification will appear on the device and users can click to view more information.

[0158] Step 7: Gather feedback

[0159] User: The user reviews the summarized information and provides feedback on whether the information was helpful. Feedback options include "helpful," "not helpful," and "I'd like more information."

[0160] Input: User ratings and comments.

[0161] Output: Feedback information sent to the server.

[0162] Specific behavior:

[0163] The user sees the notification and selects a feedback option.

[0164] The selected feedback is sent to the server.

[0165] Step 8: Save and learn from feedback

[0166] Server: The server collects the feedback and stores it in a database. This data is used to improve the performance of the natural language processing model the next time a summary is generated.

[0167] Input: Feedback information submitted by the user.

[0168] Output: Feedback stored in a database and a retrained NLP model.

[0169] Specific behavior:

[0170] The server stores the received feedback in a database.

[0171] Based on the stored feedback data, the NLP model is retrained to improve accuracy.

[0172] (Application example 1)

[0173] 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."

[0174] In virtual stores, it is difficult to efficiently obtain information about products and special offers that interest users. A method is needed to extract important information from a vast amount of information and provide it to users quickly. Another challenge is to understand whether the provided information was useful to users and collect feedback to improve the accuracy of the system.

[0175] 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.

[0176] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for collecting feedback from the user and adjusting the natural language processing model for the next summary generation, and means for collecting the latest information and special offers from virtual stores, summarizing them, and notifying the user of them. This enables the user to efficiently collect important information from a vast amount of information and understand it in a short time.

[0177] "User interest categories" indicate the types or areas of information in which a user is particularly interested.

[0178] "Methods of collecting information from the internet" refers to technologies used to obtain data from websites and social media platforms.

[0179] A "natural language processing model" refers to an artificial intelligence model for interpreting and processing human language.

[0180] "Means of summarizing" refers to the technique of extracting important points from collected information and summarizing them in short sentences.

[0181] "User device" refers to electronic devices such as smartphones, tablets, smart glasses, and head-mounted displays.

[0182] "Means for collecting feedback" refers to the technology that collects user ratings and opinions and stores them in a database.

[0183] "Means for adjusting the natural language processing model" refers to methods for improving the accuracy of the natural language processing model based on user feedback.

[0184] A "virtual store" refers to a virtual commercial space set up on the Internet.

[0185] "Means for collecting the latest information and special offers" refers to technology for efficiently collecting new product information and special offer information provided by virtual stores.

[0186] "Means of notification" refers to technology that sends information to the user's device so that it can be checked in real time.

[0187] A system for implementing this invention mainly comprises the following components: means for collecting information from the Internet based on a user's interest categories; means for summarizing the collected information using a natural language processing model; means for notifying the user of the summarized information on their terminal; means for collecting feedback from the user and adjusting the natural language processing model for the next summary generation; and means for collecting, summarizing, and notifying the user of the latest information and special offers from virtual stores.

[0188] Information Gathering Process

[0189] The server collects data from specific websites and social media platforms based on the user's defined interest categories. It uses a crawler to periodically visit these platforms and collect new content. For example, if a user defines their interest in "fashion," the server will crawl fashion-related news sites and blogs, collect newly published articles, and store them in a database.

[0190] Summary Generation Process

[0191] The server inputs the collected data into a natural language processing model (e.g., a deep learning model), extracts important information, and generates a summary. This model is adjusted based on user feedback, improving the accuracy of the next summary generation. As a specific example, if a 2,000-character fashion-related article is collected, the natural language processing model converts it into a 300-character summary. This allows users to grasp important information in a short amount of time.

[0192] Data provision process

[0193] The server organizes the summarized information for each user and sends it to the user's device based on the notification schedule. The device (such as a smartphone or smart glasses) receives the push notification and displays the summarized information on its screen. For example, if a user has configured their device to receive summary information at 8:00 a.m., the server will send a notification to the user's device every morning at 8:00 a.m. providing the latest fashion-related information.

[0194] Feedback Collection Process

[0195] After reviewing the summarized information, users can provide feedback on whether they found it useful. This feedback includes options such as "helpful," "not helpful," or "I'd like more information." The server collects this feedback, stores it in a database, and uses it the next time it generates a summary. This makes it possible to continuously improve the performance of the natural language processing model. For example, if a user provides feedback that "information on the latest fashion collections" was useful, that evaluation is used as training data for the model.

[0196] Hardware and software used

[0197] Natural Language Processing Models: Pipeline of the transformers library

[0198] Data collection: using the requests library

[0199] Crawlers: crawlers that collect data from websites and social media platforms

[0200] Prompt Sentence Examples

[0201] Input: "Latest fashion collection information"

[0202] Prompt: "Summarize an article about the latest fashion collection."

[0203] This system allows users to efficiently gather information that is important to them from a vast amount of information and understand it in a short time. Furthermore, the accuracy of the system is continuously improved through feedback, realizing more personalized information provision.

[0204] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0205] Step 1:

[0206] User interest category settings

[0207] The user specifies the categories of interest (e.g., fashion, technology, sports, etc.). The user's interest categories are sent as input to the server, which stores the interest categories in a database. The server uses this information for subsequent processing.

[0208] Step 2:

[0209] Information gathering

[0210] The server uses a crawler to visit specific websites and social media platforms based on the interest categories set by the user. As input, it takes the user's interest categories and a list of target URLs, and the crawler visits those URLs to collect new content. The collected data is stored in a database in text format.

[0211] Step 3:

[0212] Summary Generation

[0213] The server inputs the text data acquired in the information collection step into a natural language processing model (e.g., a deep learning model). As input, the collected text data is passed to a summary generation model, which extracts important information and summarizes it into a short sentence. For example, a 2,000-character article is converted into a 300-character summary using the prompt "Please summarize an article about the latest fashion collection." The generated summary is stored in a notification database.

[0214] Step 4:

[0215] Notification Schedule

[0216] The server manages the notification schedule specified for each user. It reads the user's notification setting time as input and registers summary information in the notification schedule according to that time. Notifications are sent at the specified time based on the registered schedule.

[0217] Step 5:

[0218] Push notifications

[0219] The server sends push notifications to the user's device based on the notification schedule. It uses the scheduled summary information and the user's device information as input. The device receives the push notification and displays the summary information on the screen. The user can check the notification and view more information if interested.

[0220] Step 6:

[0221] Feedback collection

[0222] Users provide feedback on the summary notification they receive. As input, the feedback provided by the users (e.g., "It was helpful," "It wasn't helpful," "I'd like more information") is collected and sent to the server. The feedback information is stored in a database.

[0223] Step 7:

[0224] Model adjustment

[0225] The server adjusts the natural language processing model based on the collected feedback information. It takes the feedback data stored in the database as input and uses it as training data for the model. This improves the accuracy of the next summary generation. Specifically, it uses the feedback data to update the model parameters, enabling it to generate summaries that better match the user's interests.

[0226] These steps allow users to efficiently gather and quickly understand the latest information relevant to their interests, and also allow for continuous improvement of the system through feedback.

[0227] 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.

[0228] The present invention is a system that collects information from the Internet based on the user's interest categories, summarizes the information, and provides it to the user. It also recognizes the user's emotions and reflects them in the summary and feedback processing. Specific embodiments of the present invention are described below.

[0229] Overall system overview

[0230] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, and an emotion engine for recognizing the user's emotions. It also has a mechanism for collecting user feedback and reflecting it in the next summary generation.

[0231] Information gathering

[0232] Server: Launches a crawler that collects data from specific websites and social media platforms based on the user's interest categories. The crawler periodically visits and collects new content to store in the database.

[0233] Example: If user A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[0234] Summary Generation

[0235] Server: The collected data is fed into a natural language processing (NLP) model. This model extracts important information from the text and generates a summary. The model is adjusted based on user feedback to improve accuracy. User sentiment is also taken into account to adjust the summary.

[0236] Example: When a long technical article is collected, the article is summarized by an NLP model to extract only the main points. For example, a 5000-character article is converted into a 300-character summary. If the user is excited, a positive tone is added to the summary.

[0237] Data provision

[0238] Server: The summarized information is organized for each user and sent based on the notification schedule. An emotion engine is also used here to notify users based on their current emotional state.

[0239] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[0240] Example: If user A has configured their device to receive summary information at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information. If user A is feeling anxious, the information will be adjusted to be more easily understood.

[0241] Feedback and Learning

[0242] User: After reviewing the summarized information, the user can provide feedback by selecting from options such as "helpful," "not helpful," or "want more information." The user's emotional state at the time of providing feedback is also captured.

[0243] Server: Receives the provided feedback, stores it in a database, analyzes it using the emotion engine, and uses the results to adjust the NLP model.

[0244] Example: If user A gives feedback that the summary they received was very useful, that feedback and their emotional state at the time can be used as training data for the model to provide useful summaries to other users with similar interests. For example, we can strengthen the summarization method that receives a lot of positive feedback.

[0245] ---

[0246] This allows users to efficiently gather important information from a vast amount of information and understand it in a short time. Also, by taking into account the user's emotions, the present invention provides more personalized information and improves the user experience.

[0247] The processing flow will be explained below.

[0248] ---

[0249] Step 1: User Registration and Setup

[0250] Users: When using the service for the first time, they create an account, enter their email address and password, and select the categories of their interests (e.g., "Technology," "Health," "Entertainment").

[0251] Server: Receives user input and stores user information and selected interest categories in a database.

[0252] Step 2: Gather information

[0253] Server: Launches crawlers that collect data from specific websites and social media platforms based on user interest categories. The crawlers periodically traverse the site, collecting new content and storing it in a database.

[0254] What happens: The server instructs the crawler to visit health-related news sites and blogs based on User B's interest category "health." The crawler collects articles and posts and stores them in a database.

[0255] Step 3: Summary generation

[0256] Server: The collected data is fed into a natural language processing (NLP) model, which extracts key information from the text and generates a summary.

[0257] Server: The emotion engine analyzes the user's current emotional state and reflects the results in the summary generation process.

[0258] How it works: The server inputs a long health-related article into the NLP model and generates a summary. The emotion engine detects "positive emotions" from User B's recent feedback and applies positive language to the summary.

[0259] Step 4: Provide data

[0260] Server: The summarized information is organized for each user and sent based on the notification schedule. An emotion engine is also used here to notify users based on their current emotional state.

[0261] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[0262] Specific operation: The server pushes a summary of the latest health information to User B's device every morning at 8:00. User B's device receives the notification and displays a positive summary on the screen: "Latest health news: Simple exercises you can do starting today."

[0263] Step 5: Feedback and learning

[0264] User: After reviewing the summarized information, the user can provide feedback by selecting from options such as "helpful," "not helpful," or "want more information." The user's emotional state at the time of providing feedback is also captured.

[0265] Server: Receives the provided feedback, stores it in a database, analyzes it using the emotion engine, and uses the results to adjust the NLP model.

[0266] Specific operation: User B gives feedback that the service was "helpful," and the emotion engine determines the emotion at that time as "delight." This feedback and emotion data are stored in the database and used to optimize the NLP model.

[0267] ---

[0268] This processing step allows users to efficiently collect information specific to their interests and grasp the main points in a short time. Also, by taking into account the user's emotions, more personalized information can be provided and the user experience can be improved.

[0269] Example 2

[0270] 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."

[0271] Currently, there is a huge amount of information on the Internet, and it is difficult to efficiently collect, summarize, and provide it to users. Furthermore, there is no method for providing personalized information that takes user emotions into account, making it difficult to improve the user experience. Furthermore, there is a lack of a mechanism for reflecting user feedback to improve the accuracy of information provided.

[0272] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0273] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for recognizing the user's emotional state and adjusting the summary content and notification method, and means for collecting feedback and emotions from the user and adjusting the natural language processing model for next summary generation, thereby enabling the user to efficiently collect important information and understand it in a short time, as well as providing personalized information and improving the user experience.

[0274] "User" refers to the ultimate user who collects information and receives summarized data.

[0275] "Interest categories" refer to themes or genres in which a user is particularly interested.

[0276] The "Internet" refers to a huge information network that interconnects computer networks all over the world.

[0277] "Means of collecting information" refers to the technical methods and devices used to search for data on the Internet and obtain specific content.

[0278] A "natural language processing model" refers to a computational model for understanding and generating human language.

[0279] A "summary" is a piece of text that briefly summarizes the essence of a longer piece of text.

[0280] "Terminal" refers to a device used by a user, such as a computer or smartphone.

[0281] "Means of notification" refers to the communication technologies and protocols used to deliver the generated information to users.

[0282] "Emotional state" refers to the user's psychological state or mood.

[0283] "Adjustment" refers to technologies and algorithms that change the way information is presented in response to user sentiment and feedback.

[0284] "Feedback" refers to a user's evaluation or opinion of the information provided.

[0285] "Generative AI model" refers to an artificial intelligence model used for generative tasks (e.g., text generation, summary generation, etc.).

[0286] A "prompt" refers to an instruction given to a generative AI model to perform a specific task.

[0287] The present invention is a system that collects information from the Internet based on the user's interest categories, summarizes it, and provides it to the user. Furthermore, it recognizes the user's emotions, adjusts the summary content and notification method, and reflects feedback to improve the accuracy of the next summary generation. Specific embodiments are described below.

[0288] Overall system overview

[0289] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, an emotion engine for recognizing the user's emotions, and a means for collecting feedback from the user and reflecting it in the next summary generation.

[0290] Information gathering

[0291] The server launches a crawler based on the user's interest categories. The crawler uses software such as Scrapy or Beautiful Soup to periodically collect information from specific websites and social media. The collected data is then stored in a database such as MySQL.

[0292] Examples:

[0293] Assume that User A is interested in "technology." The server uses Scrapy to crawl technology-related news sites and blogs and collect newly published articles. The collected articles are stored in a MySQL database.

[0294] Summary Generation

[0295] The server inputs the collected data into a natural language processing model (such as OpenAI's GPT-3) to generate a summary. The summary is then adjusted based on the user's emotional state. User feedback and emotional data are also collected to help improve the model's accuracy.

[0296] Examples:

[0297] Once long technology-related articles are collected, they are summarized by GPT-3. For example, a 5,000-character article is summarized into 300 characters. If the user is excited, a positive tone is added to the summary. Example prompt: "Summarize the following technology article in 300 characters or less."

[0298] Data provision

[0299] The server organizes the summarized information for each user and sends it based on a notification schedule. Using an emotion engine, the content of notifications is adjusted according to the user's current emotional state.

[0300] The device receives a push notification and displays it on the user's smartphone or tablet, where the user can check the notification and read more if interested.

[0301] Examples:

[0302] If User A has set up to receive information every morning at 8:00, the server will send summarized technology-related information to User A's device at 8:00. The information will be tailored to make it easier for User A to understand, especially if User A is feeling anxious.

[0303] Feedback and Learning

[0304] After checking the summary information, the user can provide feedback such as "It was helpful," "It wasn't helpful," or "I'd like more detailed information." The user's emotional state at the time of providing feedback is also captured.

[0305] The server stores the provided feedback in a database and analyzes it using an emotion engine, using the results to adjust the retraining dataset for the natural language processing model and improve the accuracy of future summary generation.

[0306] Examples:

[0307] If user A reviews the summary and gives feedback that it was "helpful" and their sentiment at the time is "positive," this data will be used to refine the model next time, so that other users can also receive useful summaries.

[0308] This system allows users to efficiently gather important information from a vast amount of information and understand it in a short time, and also improves the user experience by providing personalization based on user emotions.

[0309] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0310] Step 1:

[0311] Setting user interest categories

[0312] Users set their own interest categories through the device interface. The input is the user-selected interest categories, and the output is stored on the server.

[0313] Specific behavior:

[0314] 1. The user opens the smartphone app.

[0315] 2. On the "Interest Category Settings" screen, select a category such as "Technology," "Health," or "Sports."

[0316] 3. Once you have completed your selection, tap the "Save" button.

[0317] 4. The input is the selection data of interest categories, which is sent to the server.

[0318] 5. The server stores the received interest categories in a database.

[0319] Step 2:

[0320] Collection of information

[0321] The server launches a dedicated crawler based on the interest categories set by the user. The input is interest category information, and the output is collected website data. The collected data is stored in a database.

[0322] Specific behavior:

[0323] 1. Run the crawler once a day using a cron job.

[0324] 2. Use Scrapy to collect new articles from news sites.

[0325] 3. Store the collected articles in the “raw_articles” table in the MySQL database.

[0326] 4. As input, you have the website URL and crawling schedule.

[0327] 5. As output, you will get the collected text data.

[0328] Step 3:

[0329] Summary Generation

[0330] The server inputs the collected data into a natural language processing model (GPT-3) to generate a summary, taking into account the user's emotional data, and outputs a summary sentence.

[0331] Specific behavior:

[0332] 1. Retrieve new articles from the database.

[0333] 2. Call the GPT-3 API and summarize the article using a prompt, e.g., "Summarize the following article in 300 characters or less:"

[0334] 3. Adjust the tone of the summary depending on the user's emotional state (excitement, anxiety, etc.).

[0335] 4. Save the summary in the "summarized_articles" table in the database.

[0336] 5. The inputs are the text data of the retrieved articles and the user's sentiment data.

[0337] 6. As an output, a properly adjusted summary sentence is obtained.

[0338] Step 4:

[0339] Summary Notification

[0340] The server organizes the summarized information for each user and sends it based on a notification schedule. Using an emotion engine, the notification content is tailored to the user's current emotional state.

[0341] The device receives the push notification and displays it on the user's smartphone or tablet. The input is the user's schedule and summary information, and the output is the push notification.

[0342] Specific behavior:

[0343] 1. Organize summary information by user.

[0344] 2. If the notification is set to occur at 8:00 AM, the push notification will be sent based on the notification schedule.

[0345] 3. Use Firebase Cloud Messaging (FCM) to send notifications to the user's device.

[0346] 4. The user checks the notification and, if interested, displays the details in the app.

[0347] 5. As input, there is organized summary information and a notification schedule.

[0348] 6. As an output, a push notification will be displayed on the user's device.

[0349] Step 5:

[0350] Get user feedback

[0351] After checking the summary information, the user provides feedback. The inputs are the evaluation items and emotional state, and the output is the feedback data sent to the server.

[0352] Specific behavior:

[0353] 1. After checking the summary, tap the feedback button.

[0354] 2. Select from the feedback options and tap the send button.

[0355] 3. The emotional state at the time of feedback is also recorded.

[0356] 4. The inputs are the feedback content and emotion data.

[0357] 5. As an output, the feedback data is sent to the server and stored.

[0358] Step 6:

[0359] Analyzing feedback and adjusting the model

[0360] The server stores the provided feedback in a database and analyzes it using an emotion engine. Based on the results, it adjusts the retraining dataset for the natural language processing model. The inputs are the feedback and the emotional state, and the output is the adjusted model parameters.

[0361] Specific behavior:

[0362] 1. Retrieve feedback information from the database.

[0363] 2. Analyze feedback data using a sentiment engine.

[0364] 3. Using the analysis results, update the training data for the GPT-3 model and retrain it.

[0365] 4. The model is adjusted and reflected in the next summary generation.

[0366] 5. The inputs are the feedback data and sentiment analysis results.

[0367] 6. As an output, the adjusted model parameters are obtained, which are used for the next summary generation.

[0368] (Application example 2)

[0369] 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."

[0370] In today's world, it is becoming increasingly difficult for users to find valuable information from the vast amount of information available. Furthermore, there are only a limited number of systems that provide information based on user sentiment and feedback, creating a need for improved user experience. In particular, it is challenging to efficiently collect, summarize, and provide information in a personalized manner based on the user's interest categories.

[0371] 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.

[0372] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for providing the summarized information to the user's device via push notification and adjusting the notification content based on the user's emotions, and means for collecting feedback from the user and their emotions at that time and reflecting them in the natural language processing model for the next summary generation. This allows the user to efficiently collect important information according to their interests and receive appropriate information according to their emotions.

[0373] "User interest categories" refer to specific themes or genres in which a user is personally interested.

[0374] "Means of collecting information from the internet" means methods of automatically obtaining information from certain websites and social media platforms on the internet.

[0375] A "natural language processing model" is a type of artificial intelligence used to understand and generate human language, such as summarizing and translating text.

[0376] "Summarization methods" refer to methods for extracting key information from large amounts of text data and presenting it in a shortened form.

[0377] "Providing push notifications to a user's device" means providing summarized information to a user's mobile device or computer in real time.

[0378] "Means for recognizing user emotions and adjusting notification content" refers to technology that assesses the user's emotional state and changes the way information is presented and the tone of the information based on that assessment.

[0379] "Collecting user feedback and emotions at the time" means collecting data on users' evaluations and impressions of the usefulness of the information provided, as well as their emotions at the time of evaluation.

[0380] "Means of informing the natural language processing model for the next summary generation" refers to a method of using the collected feedback and sentiment data to improve the future summarization process and generate more accurate summaries.

[0381] Overall system overview

[0382] This system collects information from the Internet based on the user's interest categories, summarizes it using a natural language processing model, and provides information based on the user's emotions. It also collects user feedback and emotion data and reflects it in the next summary generation. The system mainly consists of the following components:

[0383] 1. Information collection server

[0384] 2. Natural Language Processing Model for Summarization

[0385] 3. User terminal for providing summarized information

[0386] 4. Emotion Recognition Engine

[0387] 5. Feedback and learning modules

[0388] Information gathering

[0389] The server automatically collects information from the internet based on the interest categories set by the user. It uses a crawler to retrieve information from specific websites and social media platforms and stores that information in a database. For example, if a user sets their interest in "technology," the server will periodically crawl technology-related news sites and blogs to collect newly published articles.

[0390] Summary Generation

[0391] The data collected on the server is fed into a natural language processing model, which uses a deep learning-based generative AI model (such as BERT or GPT). The model extracts important information from the text and generates a summary. This summary is then adjusted according to the user's sentiment.

[0392] Data provision

[0393] The summarized information is sent to the user's device as a push notification based on the user's notification schedule, allowing the user to quickly obtain important information. The emotion engine recognizes the user's emotions and adapts the notification content to the user's emotions. For example, if the user is anxious, the summary will be presented in a more understandable and simple language.

[0394] Feedback and Learning

[0395] Users can provide feedback on the summaries they receive. This feedback can include the usefulness of the summary, requests for additional information, and the user's emotional state. The server collects this feedback and stores it in a database. The emotion recognition engine also analyzes the emotional state at the time of feedback and reflects this data in the next summary generation.

[0396] Hardware and software used

[0397] Server: A central processing unit for collecting information and generating summaries (e.g., cloud server, on-premise server)

[0398] Crawler: uses the requests and BeautifulSoup libraries to gather information from web pages

[0399] Natural language processing models: Implementing BERT and GPT models using the transformers library

[0400] Emotion Recognition Engine: Microsoft Azure Cognitive Services Sentiment Analysis API

[0401] Database: A relational database (e.g., MySQL, PostgreSQL) to store collected information and feedback.

[0402] Specific examples

[0403] If User B sets his interest in entertainment news, the server will crawl entertainment-related websites to collect new articles, which will be summarized through a natural language processing model and adjusted based on User B's emotional state. For example,

[0404] "Prompt: Summarize this article in 300 words or less.

[0405] Text:

[0406] "In reviews of the recently released Star Wars film, many fans expressed excitement at the sequel, particularly the visual effects and story development, and many commented on how well the film connects to the previous films..."

[0407] In this way, User B can quickly access the content of interest and understand the content of a new movie based on the summarized information. Through the feedback function, User B can also provide their opinion on how useful the summary they received was and whether more details are needed. This feedback will be used to generate the next summary.

[0408] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0409] Step 1:

[0410] The server collects information from the Internet based on the user's interest categories. As input, the user's interest categories (e.g., technology, entertainment) are given, and as output, a list of URLs for related web pages is obtained. Specifically, a crawler (using the requests and BeautifulSoup libraries) crawls websites, discovers newly published articles, and stores them in a database.

[0411] Step 2:

[0412] The server retrieves articles from the collected URL list and converts them into text data. The URL list is given as input, and the text data of the retrieved articles is obtained as output. Specifically, the crawler accesses each URL, analyzes and extracts the HTML content, and obtains the text portion of the article.

[0413] Step 3:

[0414] The server inputs the collected text data into a natural language processing model to generate a summary. Text data (e.g., a news article) is given as input, and a summary is obtained as output. Specifically, it runs BERT or GPT models using the transformers library to extract the main points of the article and generate a summary of up to 300 characters.

[0415] Step 4:

[0416] The server then runs the generated summary through an emotion recognition engine to adjust the tone and content of the summary. The summary and the user's most recent emotional data are given as input, and the adjusted summary is obtained as output. Specifically, the server uses the sentiment analysis API of Microsoft Azure Cognitive Services to evaluate the user's emotions and adjust the tone and expression of the summary based on that evaluation.

[0417] Step 5:

[0418] The server provides the adjusted summary to the user's device via a push notification. The adjusted summary and the user's notification schedule are given as input, and the server sends a push notification to the user's device as output. Specifically, the server uses a push notification service such as Firebase Cloud Messaging (FCM) to notify the summary to the user's smartphone or tablet.

[0419] Step 6:

[0420] The user checks the received summary and provides feedback. The user's feedback (e.g., "It was helpful" or "I'd like more information") is given as input, and the feedback data is sent to the server as output. Specifically, when the user presses the feedback button in the app, the evaluation is recorded in the database.

[0421] Step 7:

[0422] The server analyzes the collected feedback and user emotion data and reflects the results in the next summary generation. The feedback data and emotion data are given as input, and an adjusted natural language processing model is obtained as output. Specifically, the feedback and emotion data are used to update the model parameters and improve the accuracy of the next summary generation.

[0423] This series of processing steps allows users to efficiently collect and review information that interests them in a personalized way, and to receive even more accurate information through feedback.

[0424] 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.

[0425] 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.

[0426] 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.

[0427] [Second embodiment]

[0428] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0429] 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.

[0430] 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).

[0431] 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.

[0432] 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.

[0433] 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).

[0434] 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.

[0435] 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.

[0436] 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.

[0437] 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.

[0438] 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.

[0439] 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."

[0440] The present invention provides a system for collecting information from the Internet based on the user's interest categories, summarizing the information, and providing it to the user. Specific embodiments of the present invention are described below.

[0441] Overall system overview

[0442] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means of notifying the user of the summarized information, and a mechanism for collecting user feedback and adjusting the system.

[0443] Information gathering

[0444] Server: Launches a crawler that collects data from specific websites and social media platforms based on the user's interest categories. The crawler periodically visits and collects new content to store in the database.

[0445] Example: If user A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[0446] Summary Generation

[0447] Server: The collected data is fed into a natural language processing model (e.g., a deep learning model). This model extracts important information from the text and generates a summary. The model is adjusted based on user feedback to improve its accuracy.

[0448] Example: When a long technical article is collected, it is summarized using an NLP model to extract only the main points, for example, a 5000-character article is converted into a 300-character summary.

[0449] Data provision

[0450] Server: Summarized information is organized by user and sent based on notification schedule.

[0451] On your device: The user's smartphone or tablet will receive a push notification that will appear on their screen. The user can then view the notification and read more if they are interested.

[0452] Example: If user A has configured their device to receive a summary at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information.

[0453] Feedback and Learning

[0454] Users: After viewing the summary, users can provide feedback on whether the information provided was useful, with options such as "helpful," "not helpful," or "I'd like more information."

[0455] Server: The server collects feedback and stores it in a database. This feedback data is used to improve the performance of the natural language processing model the next time a summary is generated.

[0456] Example: If user A gives feedback that the summary they received was very useful, that feedback can be used as training data for a model that can provide useful summaries to other users with similar interests.

[0457] ---

[0458] This allows users to efficiently gather important information from vast amounts of information and understand it in a short time. Furthermore, by utilizing user feedback, the system's accuracy can be continuously improved, enabling more personalized information to be provided.

[0459] The processing flow will be explained below.

[0460] ---

[0461] Step 1: User Registration and Setup

[0462] Users: When using the service for the first time, they create an account, enter their email address and password, and select the categories of their interests (e.g., "Technology," "Health," "Entertainment").

[0463] Server: Receives user input and stores user information and selected interest categories in a database.

[0464] Step 2: Gather information

[0465] Server: Based on the user's interest category, the query is passed to the crawler that is launched and configured to collect relevant information from specific websites and social media platforms.

[0466] Server: The crawler visits configured sites and gathers new content, such as articles from news sites or social media posts.

[0467] Server: Stores the collected data in a database.

[0468] Step 3: Summary generation

[0469] Server: Inputs the collected data into a natural language processing (NLP) model. The server invokes this NLP model to analyze the collected text data.

[0470] Server: The NLP model extracts important information from the text and generates a summary, for example, summarizing a 5,000-character article into 300 characters.

[0471] Server: Organizes the generated summaries for each user and stores them in a database.

[0472] Step 4: Provide data

[0473] Server: Once the summary is generated, it sends the summarized information to the user's device based on the user's notification schedule.

[0474] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[0475] Step 5: Feedback and learning

[0476] Users: After reviewing the summarized information, they can provide feedback by selecting from options such as "helpful," "not helpful," or "I'd like more information."

[0477] Server: Receives the provided feedback and stores it in a database.

[0478] Server: Uses the feedback data to adjust the parameters of the natural language processing model to improve the accuracy of summary generation, thereby improving the model so that the next summary is more suited to the user's preferences.

[0479] ---

[0480] This processing step allows users to efficiently gather information specific to their interests and quickly grasp the main points, while feedback is used to continuously improve the accuracy and user experience of the overall system.

[0481] Example 1

[0482] 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."

[0483] In today's Internet society, users are overwhelmed by the vast amount of information available, making it difficult for them to efficiently gather and understand important information tailored to their interests and needs. Furthermore, few systems provide feedback on the appropriateness of collected information to improve the quality of subsequent information provision. As a result, users often waste time and effort. To solve this problem, a method is needed to collect information based on users' interests, summarize it, notify them at the appropriate time, and utilize feedback to improve the accuracy of the system.

[0484] 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.

[0485] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for storing the collected data in a database, means for transmitting the summarized information based on a notification schedule, and means for collecting user feedback and adjusting the natural language processing model for the next summary generation. This allows users to efficiently collect important information and understand it in a short time. Furthermore, by utilizing user feedback, the accuracy of the system can be continuously improved, enabling more personalized information to be provided.

[0486] "User" means any person or entity that uses the System to gather interest-based information and receive summarized information.

[0487] "Interest categories" are classifications that indicate themes or topics that interest a user.

[0488] The "Internet" is a global network system that connects computers and servers around the world.

[0489] "Information collection methods" are the processes and tools used to obtain data from specific websites or social media platforms on the internet.

[0490] A "natural language processing model" is an artificial intelligence technology for understanding and analyzing text data, and specifically includes deep learning models.

[0491] A "summarization tool" is a process or tool that condenses collected information and extracts its main points.

[0492] A "terminal" is an electronic device used by a user to receive and view information, and specifically includes a smartphone or tablet.

[0493] "Means of Notification" means the method or mechanism for notifying the user of the summarized information, including push notifications.

[0494] "Feedback collection methods" are the methods and processes used to obtain user responses and ratings.

[0495] A "database" is a system for systematically storing and managing collected data and abstract data.

[0496] A "notification schedule" is a plan or setting that controls when information is notified to a user.

[0497] A "server" is a computing system for performing the processes of information gathering, summary generation, notification, feedback gathering, and data storage.

[0498] The present invention provides a system for collecting information from the Internet based on the user's interest categories, summarizing the information, and providing it to the user. Specific embodiments of the present invention are described below.

[0499] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, and a mechanism for collecting user feedback and adjusting the system.

[0500] Information gathering

[0501] The server launches a crawler based on the user's interest categories to collect data from specific websites and social media platforms. This crawler uses a data collection library, such as Scrapy. The server stores the collected data in a database (such as MySQL or MongoDB).

[0502] Specific examples

[0503] If User A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[0504] Summary Generation

[0505] The server inputs the collected data into a natural language processing model (e.g., a deep learning model like GPT-4 or BERT). The model extracts important information from the text and generates a summary. It then adjusts the model based on user feedback to improve its accuracy.

[0506] Specific examples

[0507] When long technical articles are collected, they are summarized using a natural language processing model to extract only the main points, for example, a 5,000-character article is converted into a 300-character summary.

[0508] Data provision

[0509] The server organizes the summarized information for each user and sends it based on a notification schedule, using scheduler software (e.g., a cron job), and the information is sent to the user's device via a push notification service (e.g., Firebase Cloud Messaging).

[0510] Specific examples

[0511] If user A has configured their device to receive a summary at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information.

[0512] Feedback and Learning

[0513] After reviewing the summarized information, users can provide feedback on whether the information was useful or not. Feedback consists of options such as "helpful," "not helpful," or "I'd like more information." The server stores this feedback in a database and uses it to improve the performance of the natural language processing model the next time a summary is generated.

[0514] Specific examples

[0515] If user A gives feedback that the summary they received was very useful, that feedback can be used as training data for the model, allowing it to provide useful summaries to other users with similar interests.

[0516] Example prompts to input to the generative AI model

[0517] "I would like to design a system that collects information from the Internet based on interest categories set by the user and summarizes that information. What are the specific processing steps and what hardware and software are used at each step? In particular, could you please provide more details about the steps of information collection, summary generation, data provision, and user feedback and learning?"

[0518] As a result, the present invention allows users to efficiently gather important information from vast amounts of information and understand it in a short time. Furthermore, by utilizing user feedback, the system's accuracy can be continuously improved, enabling more personalized information to be provided.

[0519] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0520] Step 1: User Interest Category Setting

[0521] User: The user sets the category of interest, for example, choosing a theme such as "Technology" or "Health."

[0522] Input: User selected interest categories.

[0523] Output: Interest category information passed to the server.

[0524] Step 2: Start the crawler

[0525] Server: The server launches a crawler based on the user's interest categories. The crawler is configured to visit specified websites and social media platforms. This crawler uses a library such as Scrapy.

[0526] Input: User interest category information.

[0527] Output: Collected article and post data.

[0528] Specific behavior:

[0529] The server looks up the user's interest categories and generates an appropriate URL list.

[0530] A crawler is launched based on the generated URL list to collect new articles and posts.

[0531] Step 3: Save to database

[0532] Server: Stores the collected data in a database, such as MySQL or MongoDB.

[0533] Input: The data retrieved by the crawler (articles, posts).

[0534] Output: Collected data stored in a database.

[0535] Specific behavior:

[0536] The crawler inserts the collected data into a database along with information such as text, URL, date and time.

[0537] Once the database has been saved, proceed to the next step.

[0538] Step 4: Summarize the data

[0539] Server: Inputs the collected data into a natural language processing (NLP) model, such as a deep learning model like GPT-4 or BERT.

[0540] Input: Article text extracted from the database.

[0541] Output: A summary sentence generated by the NLP model.

[0542] Specific behavior:

[0543] The server extracts new articles from the database and feeds the text data into the NLP model.

[0544] An NLP model extracts important information and generates a summary.

[0545] The generated summaries are stored in a database.

[0546] Step 5: Set up a notification schedule

[0547] Server: Organize the summarized information for each user and set up notifications to be sent based on a notification schedule, such as using a cron job.

[0548] Input: The notification schedule set by the user.

[0549] Output: Scheduled notification jobs.

[0550] Specific behavior:

[0551] The server refers to the user's notification schedule and sets the timing for sending the summary data.

[0552] When the scheduler reaches the set time, a push notification is triggered.

[0553] Step 6: Sending push notifications

[0554] Server: Posts summarized information.

[0555] Device: The user's smartphone or tablet receives the push notification and displays the information. Push notifications are sent using Firebase Cloud Messaging (FCM) or similar.

[0556] Input: Summary data, Notification schedule.

[0557] Output: The push notification sent to the user device.

[0558] Specific behavior:

[0559] The server sends a push notification, and the user device receives it.

[0560] A notification will appear on the device and users can click to view more information.

[0561] Step 7: Gather feedback

[0562] User: The user reviews the summarized information and provides feedback on whether the information was helpful. Feedback options include "helpful," "not helpful," and "I'd like more information."

[0563] Input: User ratings and comments.

[0564] Output: Feedback information sent to the server.

[0565] Specific behavior:

[0566] The user sees the notification and selects a feedback option.

[0567] The selected feedback is sent to the server.

[0568] Step 8: Save and learn from feedback

[0569] Server: The server collects the feedback and stores it in a database. This data is used to improve the performance of the natural language processing model the next time a summary is generated.

[0570] Input: Feedback information submitted by the user.

[0571] Output: Feedback stored in a database and a retrained NLP model.

[0572] Specific behavior:

[0573] The server stores the received feedback in a database.

[0574] Based on the stored feedback data, the NLP model is retrained to improve accuracy.

[0575] (Application example 1)

[0576] 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."

[0577] In virtual stores, it is difficult to efficiently obtain information about products and special offers that interest users. A method is needed to extract important information from a vast amount of information and provide it to users quickly. Another challenge is to understand whether the provided information was useful to users and collect feedback to improve the accuracy of the system.

[0578] 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.

[0579] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for collecting feedback from the user and adjusting the natural language processing model for the next summary generation, and means for collecting the latest information and special offers from virtual stores, summarizing them, and notifying the user of them. This enables the user to efficiently collect important information from a vast amount of information and understand it in a short time.

[0580] "User interest categories" indicate the types or areas of information in which a user is particularly interested.

[0581] "Methods of collecting information from the internet" refers to technologies used to obtain data from websites and social media platforms.

[0582] A "natural language processing model" refers to an artificial intelligence model for interpreting and processing human language.

[0583] "Means of summarizing" refers to the technique of extracting important points from collected information and summarizing them in short sentences.

[0584] "User device" refers to electronic devices such as smartphones, tablets, smart glasses, and head-mounted displays.

[0585] "Means for collecting feedback" refers to the technology that collects user ratings and opinions and stores them in a database.

[0586] "Means for adjusting the natural language processing model" refers to methods for improving the accuracy of the natural language processing model based on user feedback.

[0587] A "virtual store" refers to a virtual commercial space set up on the Internet.

[0588] "Means for collecting the latest information and special offers" refers to technology for efficiently collecting new product information and special offer information provided by virtual stores.

[0589] "Means of notification" refers to technology that sends information to the user's device so that it can be checked in real time.

[0590] A system for implementing this invention mainly comprises the following components: means for collecting information from the Internet based on a user's interest categories; means for summarizing the collected information using a natural language processing model; means for notifying the user of the summarized information on their terminal; means for collecting feedback from the user and adjusting the natural language processing model for the next summary generation; and means for collecting, summarizing, and notifying the user of the latest information and special offers from virtual stores.

[0591] Information Gathering Process

[0592] The server collects data from specific websites and social media platforms based on the user's defined interest categories. It uses a crawler to periodically visit these platforms and collect new content. For example, if a user defines their interest in "fashion," the server will crawl fashion-related news sites and blogs, collect newly published articles, and store them in a database.

[0593] Summary Generation Process

[0594] The server inputs the collected data into a natural language processing model (e.g., a deep learning model), extracts important information, and generates a summary. This model is adjusted based on user feedback, improving the accuracy of the next summary generation. As a specific example, if a 2,000-character fashion-related article is collected, the natural language processing model converts it into a 300-character summary. This allows users to grasp important information in a short amount of time.

[0595] Data provision process

[0596] The server organizes the summarized information for each user and sends it to the user's device based on the notification schedule. The device (such as a smartphone or smart glasses) receives the push notification and displays the summarized information on its screen. For example, if a user has configured their device to receive summary information at 8:00 a.m., the server will send a notification to the user's device every morning at 8:00 a.m. providing the latest fashion-related information.

[0597] Feedback Collection Process

[0598] After reviewing the summarized information, users can provide feedback on whether they found it useful. This feedback includes options such as "helpful," "not helpful," or "I'd like more information." The server collects this feedback, stores it in a database, and uses it the next time it generates a summary. This makes it possible to continuously improve the performance of the natural language processing model. For example, if a user provides feedback that "information on the latest fashion collections" was useful, that evaluation is used as training data for the model.

[0599] Hardware and software used

[0600] Natural Language Processing Models: Pipeline of the transformers library

[0601] Data collection: using the requests library

[0602] Crawlers: crawlers that collect data from websites and social media platforms

[0603] Prompt Sentence Examples

[0604] Input: "Latest fashion collection information"

[0605] Prompt: "Summarize an article about the latest fashion collection."

[0606] This system allows users to efficiently gather information that is important to them from a vast amount of information and understand it in a short time. Furthermore, the accuracy of the system is continuously improved through feedback, realizing more personalized information provision.

[0607] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0608] Step 1:

[0609] User interest category settings

[0610] The user specifies the categories of interest (e.g., fashion, technology, sports, etc.). The user's interest categories are sent as input to the server, which stores the interest categories in a database. The server uses this information for subsequent processing.

[0611] Step 2:

[0612] Information gathering

[0613] The server uses a crawler to visit specific websites and social media platforms based on the interest categories set by the user. As input, it takes the user's interest categories and a list of target URLs, and the crawler visits those URLs to collect new content. The collected data is stored in a database in text format.

[0614] Step 3:

[0615] Summary Generation

[0616] The server inputs the text data acquired in the information collection step into a natural language processing model (e.g., a deep learning model). As input, the collected text data is passed to a summary generation model, which extracts important information and summarizes it into a short sentence. For example, a 2,000-character article is converted into a 300-character summary using the prompt "Please summarize an article about the latest fashion collection." The generated summary is stored in a notification database.

[0617] Step 4:

[0618] Notification Schedule

[0619] The server manages the notification schedule specified for each user. It reads the user's notification setting time as input and registers summary information in the notification schedule according to that time. Notifications are sent at the specified time based on the registered schedule.

[0620] Step 5:

[0621] Push notifications

[0622] The server sends push notifications to the user's device based on the notification schedule. It uses the scheduled summary information and the user's device information as input. The device receives the push notification and displays the summary information on the screen. The user can check the notification and view more information if interested.

[0623] Step 6:

[0624] Feedback collection

[0625] Users provide feedback on the summary notification they receive. As input, the feedback provided by the users (e.g., "It was helpful," "It wasn't helpful," "I'd like more information") is collected and sent to the server. The feedback information is stored in a database.

[0626] Step 7:

[0627] Model adjustment

[0628] The server adjusts the natural language processing model based on the collected feedback information. It takes the feedback data stored in the database as input and uses it as training data for the model. This improves the accuracy of the next summary generation. Specifically, it uses the feedback data to update the model parameters, enabling it to generate summaries that better match the user's interests.

[0629] These steps allow users to efficiently gather and quickly understand the latest information relevant to their interests, and also allow for continuous improvement of the system through feedback.

[0630] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0631] The present invention is a system that collects information from the Internet based on the user's interest categories, summarizes the information, and provides it to the user. It also recognizes the user's emotions and reflects them in the summary and feedback processing. Specific embodiments of the present invention are described below.

[0632] Overall system overview

[0633] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, and an emotion engine for recognizing the user's emotions. It also has a mechanism for collecting user feedback and reflecting it in the next summary generation.

[0634] Information gathering

[0635] Server: Launches a crawler that collects data from specific websites and social media platforms based on the user's interest categories. The crawler periodically visits and collects new content to store in the database.

[0636] Example: If user A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[0637] Summary Generation

[0638] Server: The collected data is fed into a natural language processing (NLP) model. This model extracts important information from the text and generates a summary. The model is adjusted based on user feedback to improve accuracy. User sentiment is also taken into account to adjust the summary.

[0639] Example: When a long technical article is collected, the article is summarized by an NLP model to extract only the main points. For example, a 5000-character article is converted into a 300-character summary. If the user is excited, a positive tone is added to the summary.

[0640] Data provision

[0641] Server: The summarized information is organized for each user and sent based on the notification schedule. An emotion engine is also used here to notify users based on their current emotional state.

[0642] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[0643] Example: If user A has configured their device to receive summary information at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information. If user A is feeling anxious, the information will be adjusted to be more easily understood.

[0644] Feedback and Learning

[0645] User: After reviewing the summarized information, the user can provide feedback by selecting from options such as "helpful," "not helpful," or "want more information." The user's emotional state at the time of providing feedback is also captured.

[0646] Server: Receives the provided feedback, stores it in a database, analyzes it using the emotion engine, and uses the results to adjust the NLP model.

[0647] Example: If user A gives feedback that the summary they received was very useful, that feedback and their emotional state at the time can be used as training data for the model to provide useful summaries to other users with similar interests. For example, we can strengthen the summarization method that receives a lot of positive feedback.

[0648] ---

[0649] This allows users to efficiently gather important information from a vast amount of information and understand it in a short time. Also, by taking into account the user's emotions, the present invention provides more personalized information and improves the user experience.

[0650] The processing flow will be explained below.

[0651] ---

[0652] Step 1: User Registration and Setup

[0653] Users: When using the service for the first time, they create an account, enter their email address and password, and select the categories of their interests (e.g., "Technology," "Health," "Entertainment").

[0654] Server: Receives user input and stores user information and selected interest categories in a database.

[0655] Step 2: Gather information

[0656] Server: Launches crawlers that collect data from specific websites and social media platforms based on user interest categories. The crawlers periodically traverse the site, collecting new content and storing it in a database.

[0657] What happens: The server instructs the crawler to visit health-related news sites and blogs based on User B's interest category "health." The crawler collects articles and posts and stores them in a database.

[0658] Step 3: Summary generation

[0659] Server: The collected data is fed into a natural language processing (NLP) model, which extracts key information from the text and generates a summary.

[0660] Server: The emotion engine analyzes the user's current emotional state and reflects the results in the summary generation process.

[0661] How it works: The server inputs a long health-related article into the NLP model and generates a summary. The emotion engine detects "positive emotions" from User B's recent feedback and applies positive language to the summary.

[0662] Step 4: Provide data

[0663] Server: The summarized information is organized for each user and sent based on the notification schedule. An emotion engine is also used here to notify users based on their current emotional state.

[0664] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[0665] Specific operation: The server pushes a summary of the latest health information to User B's device every morning at 8:00. User B's device receives the notification and displays a positive summary on the screen: "Latest health news: Simple exercises you can do starting today."

[0666] Step 5: Feedback and learning

[0667] User: After reviewing the summarized information, the user can provide feedback by selecting from options such as "helpful," "not helpful," or "want more information." The user's emotional state at the time of providing feedback is also captured.

[0668] Server: Receives the provided feedback, stores it in a database, analyzes it using the emotion engine, and uses the results to adjust the NLP model.

[0669] Specific operation: User B gives feedback that the service was "helpful," and the emotion engine determines the emotion at that time as "delight." This feedback and emotion data are stored in the database and used to optimize the NLP model.

[0670] ---

[0671] This processing step allows users to efficiently collect information specific to their interests and grasp the main points in a short time. Also, by taking into account the user's emotions, more personalized information can be provided and the user experience can be improved.

[0672] Example 2

[0673] 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."

[0674] Currently, there is a huge amount of information on the Internet, and it is difficult to efficiently collect, summarize, and provide it to users. Furthermore, there is no method for providing personalized information that takes user emotions into account, making it difficult to improve the user experience. Furthermore, there is a lack of a mechanism for reflecting user feedback to improve the accuracy of information provided.

[0675] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0676] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for recognizing the user's emotional state and adjusting the summary content and notification method, and means for collecting feedback and emotions from the user and adjusting the natural language processing model for next summary generation, thereby enabling the user to efficiently collect important information and understand it in a short time, as well as providing personalized information and improving the user experience.

[0677] "User" refers to the ultimate user who collects information and receives summarized data.

[0678] "Interest categories" refer to themes or genres in which a user is particularly interested.

[0679] The "Internet" refers to a huge information network that interconnects computer networks all over the world.

[0680] "Means of collecting information" refers to the technical methods and devices used to search for data on the Internet and obtain specific content.

[0681] A "natural language processing model" refers to a computational model for understanding and generating human language.

[0682] A "summary" is a piece of text that briefly summarizes the essence of a longer piece of text.

[0683] "Terminal" refers to a device used by a user, such as a computer or smartphone.

[0684] "Means of notification" refers to the communication technologies and protocols used to deliver the generated information to users.

[0685] "Emotional state" refers to the user's psychological state or mood.

[0686] "Adjustment" refers to technologies and algorithms that change the way information is presented in response to user sentiment and feedback.

[0687] "Feedback" refers to a user's evaluation or opinion of the information provided.

[0688] "Generative AI model" refers to an artificial intelligence model used for generative tasks (e.g., text generation, summary generation, etc.).

[0689] A "prompt" refers to an instruction given to a generative AI model to perform a specific task.

[0690] The present invention is a system that collects information from the Internet based on the user's interest categories, summarizes it, and provides it to the user. Furthermore, it recognizes the user's emotions, adjusts the summary content and notification method, and reflects feedback to improve the accuracy of the next summary generation. Specific embodiments are described below.

[0691] Overall system overview

[0692] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, an emotion engine for recognizing the user's emotions, and a means for collecting feedback from the user and reflecting it in the next summary generation.

[0693] Information gathering

[0694] The server launches a crawler based on the user's interest categories. The crawler uses software such as Scrapy or Beautiful Soup to periodically collect information from specific websites and social media. The collected data is then stored in a database such as MySQL.

[0695] Examples:

[0696] Assume that User A is interested in "technology." The server uses Scrapy to crawl technology-related news sites and blogs and collect newly published articles. The collected articles are stored in a MySQL database.

[0697] Summary Generation

[0698] The server inputs the collected data into a natural language processing model (such as OpenAI's GPT-3) to generate a summary. The summary is then adjusted based on the user's emotional state. User feedback and emotional data are also collected to help improve the model's accuracy.

[0699] Examples:

[0700] Once long technology-related articles are collected, they are summarized by GPT-3. For example, a 5,000-character article is summarized into 300 characters. If the user is excited, a positive tone is added to the summary. Example prompt: "Summarize the following technology article in 300 characters or less."

[0701] Data provision

[0702] The server organizes the summarized information for each user and sends it based on a notification schedule. Using an emotion engine, the content of notifications is adjusted according to the user's current emotional state.

[0703] The device receives a push notification and displays it on the user's smartphone or tablet, where the user can check the notification and read more if interested.

[0704] Examples:

[0705] If User A has set up to receive information every morning at 8:00, the server will send summarized technology-related information to User A's device at 8:00. The information will be tailored to make it easier for User A to understand, especially if User A is feeling anxious.

[0706] Feedback and Learning

[0707] After checking the summary information, the user can provide feedback such as "It was helpful," "It wasn't helpful," or "I'd like more detailed information." The user's emotional state at the time of providing feedback is also captured.

[0708] The server stores the provided feedback in a database and analyzes it using an emotion engine, using the results to adjust the retraining dataset for the natural language processing model and improve the accuracy of future summary generation.

[0709] Examples:

[0710] If user A reviews the summary and gives feedback that it was "helpful" and their sentiment at the time is "positive," this data will be used to refine the model next time, so that other users can also receive useful summaries.

[0711] This system allows users to efficiently gather important information from a vast amount of information and understand it in a short time, and also improves the user experience by providing personalization based on user emotions.

[0712] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0713] Step 1:

[0714] Setting user interest categories

[0715] Users set their own interest categories through the device interface. The input is the user-selected interest categories, and the output is stored on the server.

[0716] Specific behavior:

[0717] 1. The user opens the smartphone app.

[0718] 2. On the "Interest Category Settings" screen, select a category such as "Technology," "Health," or "Sports."

[0719] 3. Once you have completed your selection, tap the "Save" button.

[0720] 4. The input is the selection data of interest categories, which is sent to the server.

[0721] 5. The server stores the received interest categories in a database.

[0722] Step 2:

[0723] Collection of information

[0724] The server launches a dedicated crawler based on the interest categories set by the user. The input is interest category information, and the output is collected website data. The collected data is stored in a database.

[0725] Specific behavior:

[0726] 1. Run the crawler once a day using a cron job.

[0727] 2. Use Scrapy to collect new articles from news sites.

[0728] 3. Store the collected articles in the “raw_articles” table in the MySQL database.

[0729] 4. As input, you have the website URL and crawling schedule.

[0730] 5. As output, you will get the collected text data.

[0731] Step 3:

[0732] Summary Generation

[0733] The server inputs the collected data into a natural language processing model (GPT-3) to generate a summary, taking into account the user's emotional data, and outputs a summary sentence.

[0734] Specific behavior:

[0735] 1. Retrieve new articles from the database.

[0736] 2. Call the GPT-3 API and summarize the article using a prompt, e.g., "Summarize the following article in 300 characters or less:"

[0737] 3. Adjust the tone of the summary depending on the user's emotional state (excitement, anxiety, etc.).

[0738] 4. Save the summary in the "summarized_articles" table in the database.

[0739] 5. The inputs are the text data of the retrieved articles and the user's sentiment data.

[0740] 6. As an output, a properly adjusted summary sentence is obtained.

[0741] Step 4:

[0742] Summary Notification

[0743] The server organizes the summarized information for each user and sends it based on a notification schedule. Using an emotion engine, the notification content is tailored to the user's current emotional state.

[0744] The device receives the push notification and displays it on the user's smartphone or tablet. The input is the user's schedule and summary information, and the output is the push notification.

[0745] Specific behavior:

[0746] 1. Organize summary information by user.

[0747] 2. If the notification is set to occur at 8:00 AM, the push notification will be sent based on the notification schedule.

[0748] 3. Use Firebase Cloud Messaging (FCM) to send notifications to the user's device.

[0749] 4. The user checks the notification and, if interested, displays the details in the app.

[0750] 5. As input, there is organized summary information and a notification schedule.

[0751] 6. As an output, a push notification will be displayed on the user's device.

[0752] Step 5:

[0753] Get user feedback

[0754] After checking the summary information, the user provides feedback. The inputs are the evaluation items and emotional state, and the output is the feedback data sent to the server.

[0755] Specific behavior:

[0756] 1. After checking the summary, tap the feedback button.

[0757] 2. Select from the feedback options and tap the send button.

[0758] 3. The emotional state at the time of feedback is also recorded.

[0759] 4. The inputs are the feedback content and emotion data.

[0760] 5. As an output, the feedback data is sent to the server and stored.

[0761] Step 6:

[0762] Analyzing feedback and adjusting the model

[0763] The server stores the provided feedback in a database and analyzes it using an emotion engine. Based on the results, it adjusts the retraining dataset for the natural language processing model. The inputs are the feedback and the emotional state, and the output is the adjusted model parameters.

[0764] Specific behavior:

[0765] 1. Retrieve feedback information from the database.

[0766] 2. Analyze feedback data using a sentiment engine.

[0767] 3. Using the analysis results, update the training data for the GPT-3 model and retrain it.

[0768] 4. The model is adjusted and reflected in the next summary generation.

[0769] 5. The inputs are the feedback data and sentiment analysis results.

[0770] 6. As an output, the adjusted model parameters are obtained, which are used for the next summary generation.

[0771] (Application example 2)

[0772] 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."

[0773] In today's world, it is becoming increasingly difficult for users to find valuable information from the vast amount of information available. Furthermore, there are only a limited number of systems that provide information based on user sentiment and feedback, creating a need for improved user experience. In particular, it is challenging to efficiently collect, summarize, and provide information in a personalized manner based on the user's interest categories.

[0774] 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.

[0775] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for providing the summarized information to the user's device via push notification and adjusting the notification content based on the user's emotions, and means for collecting feedback from the user and their emotions at that time and reflecting them in the natural language processing model for the next summary generation. This allows the user to efficiently collect important information according to their interests and receive appropriate information according to their emotions.

[0776] "User interest categories" refer to specific themes or genres in which a user is personally interested.

[0777] "Means of collecting information from the internet" means methods of automatically obtaining information from certain websites and social media platforms on the internet.

[0778] A "natural language processing model" is a type of artificial intelligence used to understand and generate human language, such as summarizing and translating text.

[0779] "Summarization methods" refer to methods for extracting key information from large amounts of text data and presenting it in a shortened form.

[0780] "Providing push notifications to a user's device" means providing summarized information to a user's mobile device or computer in real time.

[0781] "Means for recognizing user emotions and adjusting notification content" refers to technology that assesses the user's emotional state and changes the way information is presented and the tone of the information based on that assessment.

[0782] "Collecting user feedback and emotions at the time" means collecting data on users' evaluations and impressions of the usefulness of the information provided, as well as their emotions at the time of evaluation.

[0783] "Means of informing the natural language processing model for the next summary generation" refers to a method of using the collected feedback and sentiment data to improve the future summarization process and generate more accurate summaries.

[0784] Overall system overview

[0785] This system collects information from the Internet based on the user's interest categories, summarizes it using a natural language processing model, and provides information based on the user's emotions. It also collects user feedback and emotion data and reflects it in the next summary generation. The system mainly consists of the following components:

[0786] 1. Information collection server

[0787] 2. Natural Language Processing Model for Summarization

[0788] 3. User terminal for providing summarized information

[0789] 4. Emotion Recognition Engine

[0790] 5. Feedback and learning modules

[0791] Information gathering

[0792] The server automatically collects information from the internet based on the interest categories set by the user. It uses a crawler to retrieve information from specific websites and social media platforms and stores that information in a database. For example, if a user sets their interest in "technology," the server will periodically crawl technology-related news sites and blogs to collect newly published articles.

[0793] Summary Generation

[0794] The data collected on the server is fed into a natural language processing model, which uses a deep learning-based generative AI model (such as BERT or GPT). The model extracts important information from the text and generates a summary. This summary is then adjusted according to the user's sentiment.

[0795] Data provision

[0796] The summarized information is sent to the user's device as a push notification based on the user's notification schedule, allowing the user to quickly obtain important information. The emotion engine recognizes the user's emotions and adapts the notification content to the user's emotions. For example, if the user is anxious, the summary will be presented in a more understandable and simple language.

[0797] Feedback and Learning

[0798] Users can provide feedback on the summaries they receive. This feedback can include the usefulness of the summary, requests for additional information, and the user's emotional state. The server collects this feedback and stores it in a database. The emotion recognition engine also analyzes the emotional state at the time of feedback and reflects this data in the next summary generation.

[0799] Hardware and software used

[0800] Server: A central processing unit for collecting information and generating summaries (e.g., cloud server, on-premise server)

[0801] Crawler: uses the requests and BeautifulSoup libraries to gather information from web pages

[0802] Natural language processing models: Implementing BERT and GPT models using the transformers library

[0803] Emotion Recognition Engine: Microsoft Azure Cognitive Services Sentiment Analysis API

[0804] Database: A relational database (e.g., MySQL, PostgreSQL) to store collected information and feedback.

[0805] Specific examples

[0806] If User B sets his interest in entertainment news, the server will crawl entertainment-related websites to collect new articles, which will be summarized through a natural language processing model and adjusted based on User B's emotional state. For example,

[0807] "Prompt: Summarize this article in 300 words or less.

[0808] Text:

[0809] "In reviews of the recently released Star Wars film, many fans expressed excitement at the sequel, particularly the visual effects and story development, and many commented on how well the film connects to the previous films..."

[0810] In this way, User B can quickly access the content of interest and understand the content of a new movie based on the summarized information. Through the feedback function, User B can also provide their opinion on how useful the summary they received was and whether more details are needed. This feedback will be used to generate the next summary.

[0811] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0812] Step 1:

[0813] The server collects information from the Internet based on the user's interest categories. As input, the user's interest categories (e.g., technology, entertainment) are given, and as output, a list of URLs for related web pages is obtained. Specifically, a crawler (using the requests and BeautifulSoup libraries) crawls websites, discovers newly published articles, and stores them in a database.

[0814] Step 2:

[0815] The server retrieves articles from the collected URL list and converts them into text data. The URL list is given as input, and the text data of the retrieved articles is obtained as output. Specifically, the crawler accesses each URL, analyzes and extracts the HTML content, and obtains the text portion of the article.

[0816] Step 3:

[0817] The server inputs the collected text data into a natural language processing model to generate a summary. Text data (e.g., a news article) is given as input, and a summary is obtained as output. Specifically, it runs BERT or GPT models using the transformers library to extract the main points of the article and generate a summary of up to 300 characters.

[0818] Step 4:

[0819] The server then runs the generated summary through an emotion recognition engine to adjust the tone and content of the summary. The summary and the user's most recent emotional data are given as input, and the adjusted summary is obtained as output. Specifically, the server uses the sentiment analysis API of Microsoft Azure Cognitive Services to evaluate the user's emotions and adjust the tone and expression of the summary based on that evaluation.

[0820] Step 5:

[0821] The server provides the adjusted summary to the user's device via a push notification. The adjusted summary and the user's notification schedule are given as input, and the server sends a push notification to the user's device as output. Specifically, the server uses a push notification service such as Firebase Cloud Messaging (FCM) to notify the summary to the user's smartphone or tablet.

[0822] Step 6:

[0823] The user checks the received summary and provides feedback. The user's feedback (e.g., "It was helpful" or "I'd like more information") is given as input, and the feedback data is sent to the server as output. Specifically, when the user presses the feedback button in the app, the evaluation is recorded in the database.

[0824] Step 7:

[0825] The server analyzes the collected feedback and user emotion data and reflects the results in the next summary generation. The feedback data and emotion data are given as input, and an adjusted natural language processing model is obtained as output. Specifically, the feedback and emotion data are used to update the model parameters and improve the accuracy of the next summary generation.

[0826] This series of processing steps allows users to efficiently collect and review information that interests them in a personalized way, and to receive even more accurate information through feedback.

[0827] 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.

[0828] 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.

[0829] 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.

[0830] [Third embodiment]

[0831] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0832] 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.

[0833] 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).

[0834] 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.

[0835] 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.

[0836] 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).

[0837] 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.

[0838] 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.

[0839] 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.

[0840] 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.

[0841] 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.

[0842] 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."

[0843] The present invention provides a system for collecting information from the Internet based on the user's interest categories, summarizing the information, and providing it to the user. Specific embodiments of the present invention are described below.

[0844] Overall system overview

[0845] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means of notifying the user of the summarized information, and a mechanism for collecting user feedback and adjusting the system.

[0846] Information gathering

[0847] Server: Launches a crawler that collects data from specific websites and social media platforms based on the user's interest categories. The crawler periodically visits and collects new content to store in the database.

[0848] Example: If user A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[0849] Summary Generation

[0850] Server: The collected data is fed into a natural language processing model (e.g., a deep learning model). This model extracts important information from the text and generates a summary. The model is adjusted based on user feedback to improve its accuracy.

[0851] Example: When a long technical article is collected, it is summarized using an NLP model to extract only the main points, for example, a 5000-character article is converted into a 300-character summary.

[0852] Data provision

[0853] Server: Summarized information is organized by user and sent based on notification schedule.

[0854] On your device: The user's smartphone or tablet will receive a push notification that will appear on their screen. The user can then view the notification and read more if they are interested.

[0855] Example: If user A has configured their device to receive a summary at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information.

[0856] Feedback and Learning

[0857] Users: After viewing the summary, users can provide feedback on whether the information provided was useful, with options such as "helpful," "not helpful," or "I'd like more information."

[0858] Server: The server collects feedback and stores it in a database. This feedback data is used to improve the performance of the natural language processing model the next time a summary is generated.

[0859] Example: If user A gives feedback that the summary they received was very useful, that feedback can be used as training data for a model that can provide useful summaries to other users with similar interests.

[0860] ---

[0861] This allows users to efficiently gather important information from vast amounts of information and understand it in a short time. Furthermore, by utilizing user feedback, the system's accuracy can be continuously improved, enabling more personalized information to be provided.

[0862] The processing flow will be explained below.

[0863] ---

[0864] Step 1: User Registration and Setup

[0865] Users: When using the service for the first time, they create an account, enter their email address and password, and select the categories of their interests (e.g., "Technology," "Health," "Entertainment").

[0866] Server: Receives user input and stores user information and selected interest categories in a database.

[0867] Step 2: Gather information

[0868] Server: Based on the user's interest category, the query is passed to the crawler that is launched and configured to collect relevant information from specific websites and social media platforms.

[0869] Server: The crawler visits configured sites and gathers new content, such as articles from news sites or social media posts.

[0870] Server: Stores the collected data in a database.

[0871] Step 3: Summary generation

[0872] Server: Inputs the collected data into a natural language processing (NLP) model. The server invokes this NLP model to analyze the collected text data.

[0873] Server: The NLP model extracts important information from the text and generates a summary, for example, summarizing a 5,000-character article into 300 characters.

[0874] Server: Organizes the generated summaries for each user and stores them in a database.

[0875] Step 4: Provide data

[0876] Server: Once the summary is generated, it sends the summarized information to the user's device based on the user's notification schedule.

[0877] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[0878] Step 5: Feedback and learning

[0879] Users: After reviewing the summarized information, they can provide feedback by selecting from options such as "helpful," "not helpful," or "I'd like more information."

[0880] Server: Receives the provided feedback and stores it in a database.

[0881] Server: Uses the feedback data to adjust the parameters of the natural language processing model to improve the accuracy of summary generation, thereby improving the model so that the next summary is more suited to the user's preferences.

[0882] ---

[0883] This processing step allows users to efficiently gather information specific to their interests and quickly grasp the main points, while feedback is used to continuously improve the accuracy and user experience of the overall system.

[0884] Example 1

[0885] 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."

[0886] In today's Internet society, users are overwhelmed by the vast amount of information available, making it difficult for them to efficiently gather and understand important information tailored to their interests and needs. Furthermore, few systems provide feedback on the appropriateness of collected information to improve the quality of subsequent information provision. As a result, users often waste time and effort. To solve this problem, a method is needed to collect information based on users' interests, summarize it, notify them at the appropriate time, and utilize feedback to improve the accuracy of the system.

[0887] 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.

[0888] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for storing the collected data in a database, means for transmitting the summarized information based on a notification schedule, and means for collecting user feedback and adjusting the natural language processing model for the next summary generation. This allows users to efficiently collect important information and understand it in a short time. Furthermore, by utilizing user feedback, the accuracy of the system can be continuously improved, enabling more personalized information to be provided.

[0889] "User" means any person or entity that uses the System to gather interest-based information and receive summarized information.

[0890] "Interest categories" are classifications that indicate themes or topics that interest a user.

[0891] The "Internet" is a global network system that connects computers and servers around the world.

[0892] "Information collection methods" are the processes and tools used to obtain data from specific websites or social media platforms on the internet.

[0893] A "natural language processing model" is an artificial intelligence technology for understanding and analyzing text data, and specifically includes deep learning models.

[0894] A "summarization tool" is a process or tool that condenses collected information and extracts its main points.

[0895] A "terminal" is an electronic device used by a user to receive and view information, and specifically includes a smartphone or tablet.

[0896] "Means of Notification" means the method or mechanism for notifying the user of the summarized information, including push notifications.

[0897] "Feedback collection methods" are the methods and processes used to obtain user responses and ratings.

[0898] A "database" is a system for systematically storing and managing collected data and abstract data.

[0899] A "notification schedule" is a plan or setting that controls when information is notified to a user.

[0900] A "server" is a computing system for performing the processes of information gathering, summary generation, notification, feedback gathering, and data storage.

[0901] The present invention provides a system for collecting information from the Internet based on the user's interest categories, summarizing the information, and providing it to the user. Specific embodiments of the present invention are described below.

[0902] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, and a mechanism for collecting user feedback and adjusting the system.

[0903] Information gathering

[0904] The server launches a crawler based on the user's interest categories to collect data from specific websites and social media platforms. This crawler uses a data collection library, such as Scrapy. The server stores the collected data in a database (such as MySQL or MongoDB).

[0905] Specific examples

[0906] If User A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[0907] Summary Generation

[0908] The server inputs the collected data into a natural language processing model (e.g., a deep learning model like GPT-4 or BERT). The model extracts important information from the text and generates a summary. It then adjusts the model based on user feedback to improve its accuracy.

[0909] Specific examples

[0910] When long technical articles are collected, they are summarized using a natural language processing model to extract only the main points, for example, a 5,000-character article is converted into a 300-character summary.

[0911] Data provision

[0912] The server organizes the summarized information for each user and sends it based on a notification schedule, using scheduler software (e.g., a cron job), and the information is sent to the user's device via a push notification service (e.g., Firebase Cloud Messaging).

[0913] Specific examples

[0914] If user A has configured their device to receive a summary at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information.

[0915] Feedback and Learning

[0916] After reviewing the summarized information, users can provide feedback on whether the information was useful or not. Feedback consists of options such as "helpful," "not helpful," or "I'd like more information." The server stores this feedback in a database and uses it to improve the performance of the natural language processing model the next time a summary is generated.

[0917] Specific examples

[0918] If user A gives feedback that the summary they received was very useful, that feedback can be used as training data for the model, allowing it to provide useful summaries to other users with similar interests.

[0919] Example prompts to input to the generative AI model

[0920] "I would like to design a system that collects information from the Internet based on interest categories set by the user and summarizes that information. What are the specific processing steps and what hardware and software are used at each step? In particular, could you please provide more details about the steps of information collection, summary generation, data provision, and user feedback and learning?"

[0921] As a result, the present invention allows users to efficiently gather important information from vast amounts of information and understand it in a short time. Furthermore, by utilizing user feedback, the system's accuracy can be continuously improved, enabling more personalized information to be provided.

[0922] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0923] Step 1: User Interest Category Setting

[0924] User: The user sets the category of interest, for example, choosing a theme such as "Technology" or "Health."

[0925] Input: User selected interest categories.

[0926] Output: Interest category information passed to the server.

[0927] Step 2: Start the crawler

[0928] Server: The server launches a crawler based on the user's interest categories. The crawler is configured to visit specified websites and social media platforms. This crawler uses a library such as Scrapy.

[0929] Input: User interest category information.

[0930] Output: Collected article and post data.

[0931] Specific behavior:

[0932] The server looks up the user's interest categories and generates an appropriate URL list.

[0933] A crawler is launched based on the generated URL list to collect new articles and posts.

[0934] Step 3: Save to database

[0935] Server: Stores the collected data in a database, such as MySQL or MongoDB.

[0936] Input: The data retrieved by the crawler (articles, posts).

[0937] Output: Collected data stored in a database.

[0938] Specific behavior:

[0939] The crawler inserts the collected data into a database along with information such as text, URL, date and time.

[0940] Once the database has been saved, proceed to the next step.

[0941] Step 4: Summarize the data

[0942] Server: Inputs the collected data into a natural language processing (NLP) model, such as a deep learning model like GPT-4 or BERT.

[0943] Input: Article text extracted from the database.

[0944] Output: A summary sentence generated by the NLP model.

[0945] Specific behavior:

[0946] The server extracts new articles from the database and feeds the text data into the NLP model.

[0947] An NLP model extracts important information and generates a summary.

[0948] The generated summaries are stored in a database.

[0949] Step 5: Set up a notification schedule

[0950] Server: Organize the summarized information for each user and set up notifications to be sent based on a notification schedule, such as using a cron job.

[0951] Input: The notification schedule set by the user.

[0952] Output: Scheduled notification jobs.

[0953] Specific behavior:

[0954] The server refers to the user's notification schedule and sets the timing for sending the summary data.

[0955] When the scheduler reaches the set time, a push notification is triggered.

[0956] Step 6: Sending push notifications

[0957] Server: Posts summarized information.

[0958] Device: The user's smartphone or tablet receives the push notification and displays the information. Push notifications are sent using Firebase Cloud Messaging (FCM) or similar.

[0959] Input: Summary data, Notification schedule.

[0960] Output: The push notification sent to the user device.

[0961] Specific behavior:

[0962] The server sends a push notification, and the user device receives it.

[0963] A notification will appear on the device and users can click to view more information.

[0964] Step 7: Gather feedback

[0965] User: The user reviews the summarized information and provides feedback on whether the information was helpful. Feedback options include "helpful," "not helpful," and "I'd like more information."

[0966] Input: User ratings and comments.

[0967] Output: Feedback information sent to the server.

[0968] Specific behavior:

[0969] The user sees the notification and selects a feedback option.

[0970] The selected feedback is sent to the server.

[0971] Step 8: Save and learn from feedback

[0972] Server: The server collects the feedback and stores it in a database. This data is used to improve the performance of the natural language processing model the next time a summary is generated.

[0973] Input: Feedback information submitted by the user.

[0974] Output: Feedback stored in a database and a retrained NLP model.

[0975] Specific behavior:

[0976] The server stores the received feedback in a database.

[0977] Based on the stored feedback data, the NLP model is retrained to improve accuracy.

[0978] (Application example 1)

[0979] 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."

[0980] In virtual stores, it is difficult to efficiently obtain information about products and special offers that interest users. A method is needed to extract important information from a vast amount of information and provide it to users quickly. Another challenge is to understand whether the provided information was useful to users and collect feedback to improve the accuracy of the system.

[0981] 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.

[0982] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for collecting feedback from the user and adjusting the natural language processing model for the next summary generation, and means for collecting the latest information and special offers from virtual stores, summarizing them, and notifying the user of them. This enables the user to efficiently collect important information from a vast amount of information and understand it in a short time.

[0983] "User interest categories" indicate the types or areas of information in which a user is particularly interested.

[0984] "Methods of collecting information from the internet" refers to technologies used to obtain data from websites and social media platforms.

[0985] A "natural language processing model" refers to an artificial intelligence model for interpreting and processing human language.

[0986] "Means of summarizing" refers to the technique of extracting important points from collected information and summarizing them in short sentences.

[0987] "User device" refers to electronic devices such as smartphones, tablets, smart glasses, and head-mounted displays.

[0988] "Means for collecting feedback" refers to the technology that collects user ratings and opinions and stores them in a database.

[0989] "Means for adjusting the natural language processing model" refers to methods for improving the accuracy of the natural language processing model based on user feedback.

[0990] A "virtual store" refers to a virtual commercial space set up on the Internet.

[0991] "Means for collecting the latest information and special offers" refers to technology for efficiently collecting new product information and special offer information provided by virtual stores.

[0992] "Means of notification" refers to technology that sends information to the user's device so that it can be checked in real time.

[0993] A system for implementing this invention mainly comprises the following components: means for collecting information from the Internet based on a user's interest categories; means for summarizing the collected information using a natural language processing model; means for notifying the user of the summarized information on their terminal; means for collecting feedback from the user and adjusting the natural language processing model for the next summary generation; and means for collecting, summarizing, and notifying the user of the latest information and special offers from virtual stores.

[0994] Information Gathering Process

[0995] The server collects data from specific websites and social media platforms based on the user's defined interest categories. It uses a crawler to periodically visit these platforms and collect new content. For example, if a user defines their interest in "fashion," the server will crawl fashion-related news sites and blogs, collect newly published articles, and store them in a database.

[0996] Summary Generation Process

[0997] The server inputs the collected data into a natural language processing model (e.g., a deep learning model), extracts important information, and generates a summary. This model is adjusted based on user feedback, improving the accuracy of the next summary generation. As a specific example, if a 2,000-character fashion-related article is collected, the natural language processing model converts it into a 300-character summary. This allows users to grasp important information in a short amount of time.

[0998] Data provision process

[0999] The server organizes the summarized information for each user and sends it to the user's device based on the notification schedule. The device (such as a smartphone or smart glasses) receives the push notification and displays the summarized information on its screen. For example, if a user has configured their device to receive summary information at 8:00 a.m., the server will send a notification to the user's device every morning at 8:00 a.m. providing the latest fashion-related information.

[1000] Feedback Collection Process

[1001] After reviewing the summarized information, users can provide feedback on whether they found it useful. This feedback includes options such as "helpful," "not helpful," or "I'd like more information." The server collects this feedback, stores it in a database, and uses it the next time it generates a summary. This makes it possible to continuously improve the performance of the natural language processing model. For example, if a user provides feedback that "information on the latest fashion collections" was useful, that evaluation is used as training data for the model.

[1002] Hardware and software used

[1003] Natural Language Processing Models: Pipeline of the transformers library

[1004] Data collection: using the requests library

[1005] Crawlers: crawlers that collect data from websites and social media platforms

[1006] Prompt Sentence Examples

[1007] Input: "Latest fashion collection information"

[1008] Prompt: "Summarize an article about the latest fashion collection."

[1009] This system allows users to efficiently gather information that is important to them from a vast amount of information and understand it in a short time. Furthermore, the accuracy of the system is continuously improved through feedback, realizing more personalized information provision.

[1010] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1011] Step 1:

[1012] User interest category settings

[1013] The user specifies the categories of interest (e.g., fashion, technology, sports, etc.). The user's interest categories are sent as input to the server, which stores the interest categories in a database. The server uses this information for subsequent processing.

[1014] Step 2:

[1015] Information gathering

[1016] The server uses a crawler to visit specific websites and social media platforms based on the interest categories set by the user. As input, it takes the user's interest categories and a list of target URLs, and the crawler visits those URLs to collect new content. The collected data is stored in a database in text format.

[1017] Step 3:

[1018] Summary Generation

[1019] The server inputs the text data acquired in the information collection step into a natural language processing model (e.g., a deep learning model). As input, the collected text data is passed to a summary generation model, which extracts important information and summarizes it into a short sentence. For example, a 2,000-character article is converted into a 300-character summary using the prompt "Please summarize an article about the latest fashion collection." The generated summary is stored in a notification database.

[1020] Step 4:

[1021] Notification Schedule

[1022] The server manages the notification schedule specified for each user. It reads the user's notification setting time as input and registers summary information in the notification schedule according to that time. Notifications are sent at the specified time based on the registered schedule.

[1023] Step 5:

[1024] Push notifications

[1025] The server sends push notifications to the user's device based on the notification schedule. It uses the scheduled summary information and the user's device information as input. The device receives the push notification and displays the summary information on the screen. The user can check the notification and view more information if interested.

[1026] Step 6:

[1027] Feedback collection

[1028] Users provide feedback on the summary notification they receive. As input, the feedback provided by the users (e.g., "It was helpful," "It wasn't helpful," "I'd like more information") is collected and sent to the server. The feedback information is stored in a database.

[1029] Step 7:

[1030] Model adjustment

[1031] The server adjusts the natural language processing model based on the collected feedback information. It takes the feedback data stored in the database as input and uses it as training data for the model. This improves the accuracy of the next summary generation. Specifically, it uses the feedback data to update the model parameters, enabling it to generate summaries that better match the user's interests.

[1032] These steps allow users to efficiently gather and quickly understand the latest information relevant to their interests, and also allow for continuous improvement of the system through feedback.

[1033] 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.

[1034] The present invention is a system that collects information from the Internet based on the user's interest categories, summarizes the information, and provides it to the user. It also recognizes the user's emotions and reflects them in the summary and feedback processing. Specific embodiments of the present invention are described below.

[1035] Overall system overview

[1036] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, and an emotion engine for recognizing the user's emotions. It also has a mechanism for collecting user feedback and reflecting it in the next summary generation.

[1037] Information gathering

[1038] Server: Launches a crawler that collects data from specific websites and social media platforms based on the user's interest categories. The crawler periodically visits and collects new content to store in the database.

[1039] Example: If user A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[1040] Summary Generation

[1041] Server: The collected data is fed into a natural language processing (NLP) model. This model extracts important information from the text and generates a summary. The model is adjusted based on user feedback to improve accuracy. User sentiment is also taken into account to adjust the summary.

[1042] Example: When a long technical article is collected, the article is summarized by an NLP model to extract only the main points. For example, a 5000-character article is converted into a 300-character summary. If the user is excited, a positive tone is added to the summary.

[1043] Data provision

[1044] Server: The summarized information is organized for each user and sent based on the notification schedule. An emotion engine is also used here to notify users based on their current emotional state.

[1045] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[1046] Example: If user A has configured their device to receive summary information at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information. If user A is feeling anxious, the information will be adjusted to be more easily understood.

[1047] Feedback and Learning

[1048] User: After reviewing the summarized information, the user can provide feedback by selecting from options such as "helpful," "not helpful," or "want more information." The user's emotional state at the time of providing feedback is also captured.

[1049] Server: Receives the provided feedback, stores it in a database, analyzes it using the emotion engine, and uses the results to adjust the NLP model.

[1050] Example: If user A gives feedback that the summary they received was very useful, that feedback and their emotional state at the time can be used as training data for the model to provide useful summaries to other users with similar interests. For example, we can strengthen the summarization method that receives a lot of positive feedback.

[1051] ---

[1052] This allows users to efficiently gather important information from a vast amount of information and understand it in a short time. Also, by taking into account the user's emotions, the present invention provides more personalized information and improves the user experience.

[1053] The processing flow will be explained below.

[1054] ---

[1055] Step 1: User Registration and Setup

[1056] Users: When using the service for the first time, they create an account, enter their email address and password, and select the categories of their interests (e.g., "Technology," "Health," "Entertainment").

[1057] Server: Receives user input and stores user information and selected interest categories in a database.

[1058] Step 2: Gather information

[1059] Server: Launches crawlers that collect data from specific websites and social media platforms based on user interest categories. The crawlers periodically traverse the site, collecting new content and storing it in a database.

[1060] What happens: The server instructs the crawler to visit health-related news sites and blogs based on User B's interest category "health." The crawler collects articles and posts and stores them in a database.

[1061] Step 3: Summary generation

[1062] Server: The collected data is fed into a natural language processing (NLP) model, which extracts key information from the text and generates a summary.

[1063] Server: The emotion engine analyzes the user's current emotional state and reflects the results in the summary generation process.

[1064] How it works: The server inputs a long health-related article into the NLP model and generates a summary. The emotion engine detects "positive emotions" from User B's recent feedback and applies positive language to the summary.

[1065] Step 4: Provide data

[1066] Server: The summarized information is organized for each user and sent based on the notification schedule. An emotion engine is also used here to notify users based on their current emotional state.

[1067] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[1068] Specific operation: The server pushes a summary of the latest health information to User B's device every morning at 8:00. User B's device receives the notification and displays a positive summary on the screen: "Latest health news: Simple exercises you can do starting today."

[1069] Step 5: Feedback and learning

[1070] User: After reviewing the summarized information, the user can provide feedback by selecting from options such as "helpful," "not helpful," or "want more information." The user's emotional state at the time of providing feedback is also captured.

[1071] Server: Receives the provided feedback, stores it in a database, analyzes it using the emotion engine, and uses the results to adjust the NLP model.

[1072] Specific operation: User B gives feedback that the service was "helpful," and the emotion engine determines the emotion at that time as "delight." This feedback and emotion data are stored in the database and used to optimize the NLP model.

[1073] ---

[1074] This processing step allows users to efficiently collect information specific to their interests and grasp the main points in a short time. Also, by taking into account the user's emotions, more personalized information can be provided and the user experience can be improved.

[1075] Example 2

[1076] 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."

[1077] Currently, there is a huge amount of information on the Internet, and it is difficult to efficiently collect, summarize, and provide it to users. Furthermore, there is no method for providing personalized information that takes user emotions into account, making it difficult to improve the user experience. Furthermore, there is a lack of a mechanism for reflecting user feedback to improve the accuracy of information provided.

[1078] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1079] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for recognizing the user's emotional state and adjusting the summary content and notification method, and means for collecting feedback and emotions from the user and adjusting the natural language processing model for next summary generation, thereby enabling the user to efficiently collect important information and understand it in a short time, as well as providing personalized information and improving the user experience.

[1080] "User" refers to the ultimate user who collects information and receives summarized data.

[1081] "Interest categories" refer to themes or genres in which a user is particularly interested.

[1082] The "Internet" refers to a huge information network that interconnects computer networks all over the world.

[1083] "Means of collecting information" refers to the technical methods and devices used to search for data on the Internet and obtain specific content.

[1084] A "natural language processing model" refers to a computational model for understanding and generating human language.

[1085] A "summary" is a piece of text that briefly summarizes the essence of a longer piece of text.

[1086] "Terminal" refers to a device used by a user, such as a computer or smartphone.

[1087] "Means of notification" refers to the communication technologies and protocols used to deliver the generated information to users.

[1088] "Emotional state" refers to the user's psychological state or mood.

[1089] "Adjustment" refers to technologies and algorithms that change the way information is presented in response to user sentiment and feedback.

[1090] "Feedback" refers to a user's evaluation or opinion of the information provided.

[1091] "Generative AI model" refers to an artificial intelligence model used for generative tasks (e.g., text generation, summary generation, etc.).

[1092] A "prompt" refers to an instruction given to a generative AI model to perform a specific task.

[1093] The present invention is a system that collects information from the Internet based on the user's interest categories, summarizes it, and provides it to the user. Furthermore, it recognizes the user's emotions, adjusts the summary content and notification method, and reflects feedback to improve the accuracy of the next summary generation. Specific embodiments are described below.

[1094] Overall system overview

[1095] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, an emotion engine for recognizing the user's emotions, and a means for collecting feedback from the user and reflecting it in the next summary generation.

[1096] Information gathering

[1097] The server launches a crawler based on the user's interest categories. The crawler uses software such as Scrapy or Beautiful Soup to periodically collect information from specific websites and social media. The collected data is then stored in a database such as MySQL.

[1098] Examples:

[1099] Assume that User A is interested in "technology." The server uses Scrapy to crawl technology-related news sites and blogs and collect newly published articles. The collected articles are stored in a MySQL database.

[1100] Summary Generation

[1101] The server inputs the collected data into a natural language processing model (such as OpenAI's GPT-3) to generate a summary. The summary is then adjusted based on the user's emotional state. User feedback and emotional data are also collected to help improve the model's accuracy.

[1102] Examples:

[1103] Once long technology-related articles are collected, they are summarized by GPT-3. For example, a 5,000-character article is summarized into 300 characters. If the user is excited, a positive tone is added to the summary. Example prompt: "Summarize the following technology article in 300 characters or less."

[1104] Data provision

[1105] The server organizes the summarized information for each user and sends it based on a notification schedule. Using an emotion engine, the content of notifications is adjusted according to the user's current emotional state.

[1106] The device receives a push notification and displays it on the user's smartphone or tablet, where the user can check the notification and read more if interested.

[1107] Examples:

[1108] If User A has set up to receive information every morning at 8:00, the server will send summarized technology-related information to User A's device at 8:00. The information will be tailored to make it easier for User A to understand, especially if User A is feeling anxious.

[1109] Feedback and Learning

[1110] After checking the summary information, the user can provide feedback such as "It was helpful," "It wasn't helpful," or "I'd like more detailed information." The user's emotional state at the time of providing feedback is also captured.

[1111] The server stores the provided feedback in a database and analyzes it using an emotion engine, using the results to adjust the retraining dataset for the natural language processing model and improve the accuracy of future summary generation.

[1112] Examples:

[1113] If user A reviews the summary and gives feedback that it was "helpful" and their sentiment at the time is "positive," this data will be used to refine the model next time, so that other users can also receive useful summaries.

[1114] This system allows users to efficiently gather important information from a vast amount of information and understand it in a short time, and also improves the user experience by providing personalization based on user emotions.

[1115] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1116] Step 1:

[1117] Setting user interest categories

[1118] Users set their own interest categories through the device interface. The input is the user-selected interest categories, and the output is stored on the server.

[1119] Specific behavior:

[1120] 1. The user opens the smartphone app.

[1121] 2. On the "Interest Category Settings" screen, select a category such as "Technology," "Health," or "Sports."

[1122] 3. Once you have completed your selection, tap the "Save" button.

[1123] 4. The input is the selection data of interest categories, which is sent to the server.

[1124] 5. The server stores the received interest categories in a database.

[1125] Step 2:

[1126] Collection of information

[1127] The server launches a dedicated crawler based on the interest categories set by the user. The input is interest category information, and the output is collected website data. The collected data is stored in a database.

[1128] Specific behavior:

[1129] 1. Run the crawler once a day using a cron job.

[1130] 2. Use Scrapy to collect new articles from news sites.

[1131] 3. Store the collected articles in the “raw_articles” table in the MySQL database.

[1132] 4. As input, you have the website URL and crawling schedule.

[1133] 5. As output, you will get the collected text data.

[1134] Step 3:

[1135] Summary Generation

[1136] The server inputs the collected data into a natural language processing model (GPT-3) to generate a summary, taking into account the user's emotional data, and outputs a summary sentence.

[1137] Specific behavior:

[1138] 1. Retrieve new articles from the database.

[1139] 2. Call the GPT-3 API and summarize the article using a prompt, e.g., "Summarize the following article in 300 characters or less:"

[1140] 3. Adjust the tone of the summary depending on the user's emotional state (excitement, anxiety, etc.).

[1141] 4. Save the summary in the "summarized_articles" table in the database.

[1142] 5. The inputs are the text data of the retrieved articles and the user's sentiment data.

[1143] 6. As an output, a properly adjusted summary sentence is obtained.

[1144] Step 4:

[1145] Summary Notification

[1146] The server organizes the summarized information for each user and sends it based on a notification schedule. Using an emotion engine, the notification content is tailored to the user's current emotional state.

[1147] The device receives the push notification and displays it on the user's smartphone or tablet. The input is the user's schedule and summary information, and the output is the push notification.

[1148] Specific behavior:

[1149] 1. Organize summary information by user.

[1150] 2. If the notification is set to occur at 8:00 AM, the push notification will be sent based on the notification schedule.

[1151] 3. Use Firebase Cloud Messaging (FCM) to send notifications to the user's device.

[1152] 4. The user checks the notification and, if interested, displays the details in the app.

[1153] 5. As input, there is organized summary information and a notification schedule.

[1154] 6. As an output, a push notification will be displayed on the user's device.

[1155] Step 5:

[1156] Get user feedback

[1157] After checking the summary information, the user provides feedback. The inputs are the evaluation items and emotional state, and the output is the feedback data sent to the server.

[1158] Specific behavior:

[1159] 1. After checking the summary, tap the feedback button.

[1160] 2. Select from the feedback options and tap the send button.

[1161] 3. The emotional state at the time of feedback is also recorded.

[1162] 4. The inputs are the feedback content and emotion data.

[1163] 5. As an output, the feedback data is sent to the server and stored.

[1164] Step 6:

[1165] Analyzing feedback and adjusting the model

[1166] The server stores the provided feedback in a database and analyzes it using an emotion engine. Based on the results, it adjusts the retraining dataset for the natural language processing model. The inputs are the feedback and the emotional state, and the output is the adjusted model parameters.

[1167] Specific behavior:

[1168] 1. Retrieve feedback information from the database.

[1169] 2. Analyze feedback data using a sentiment engine.

[1170] 3. Using the analysis results, update the training data for the GPT-3 model and retrain it.

[1171] 4. The model is adjusted and reflected in the next summary generation.

[1172] 5. The inputs are the feedback data and sentiment analysis results.

[1173] 6. As an output, the adjusted model parameters are obtained, which are used for the next summary generation.

[1174] (Application example 2)

[1175] 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."

[1176] In today's world, it is becoming increasingly difficult for users to find valuable information from the vast amount of information available. Furthermore, there are only a limited number of systems that provide information based on user sentiment and feedback, creating a need for improved user experience. In particular, it is challenging to efficiently collect, summarize, and provide information in a personalized manner based on the user's interest categories.

[1177] 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.

[1178] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for providing the summarized information to the user's device via push notification and adjusting the notification content based on the user's emotions, and means for collecting feedback from the user and their emotions at that time and reflecting them in the natural language processing model for the next summary generation. This allows the user to efficiently collect important information according to their interests and receive appropriate information according to their emotions.

[1179] "User interest categories" refer to specific themes or genres in which a user is personally interested.

[1180] "Means of collecting information from the internet" means methods of automatically obtaining information from certain websites and social media platforms on the internet.

[1181] A "natural language processing model" is a type of artificial intelligence used to understand and generate human language, such as summarizing and translating text.

[1182] "Summarization methods" refer to methods for extracting key information from large amounts of text data and presenting it in a shortened form.

[1183] "Providing push notifications to a user's device" means providing summarized information to a user's mobile device or computer in real time.

[1184] "Means for recognizing user emotions and adjusting notification content" refers to technology that assesses the user's emotional state and changes the way information is presented and the tone of the information based on that assessment.

[1185] "Collecting user feedback and emotions at the time" means collecting data on users' evaluations and impressions of the usefulness of the information provided, as well as their emotions at the time of evaluation.

[1186] "Means of informing the natural language processing model for the next summary generation" refers to a method of using the collected feedback and sentiment data to improve the future summarization process and generate more accurate summaries.

[1187] Overall system overview

[1188] This system collects information from the Internet based on the user's interest categories, summarizes it using a natural language processing model, and provides information based on the user's emotions. It also collects user feedback and emotion data and reflects it in the next summary generation. The system mainly consists of the following components:

[1189] 1. Information collection server

[1190] 2. Natural Language Processing Model for Summarization

[1191] 3. User terminal for providing summarized information

[1192] 4. Emotion Recognition Engine

[1193] 5. Feedback and learning modules

[1194] Information gathering

[1195] The server automatically collects information from the internet based on the interest categories set by the user. It uses a crawler to retrieve information from specific websites and social media platforms and stores that information in a database. For example, if a user sets their interest in "technology," the server will periodically crawl technology-related news sites and blogs to collect newly published articles.

[1196] Summary Generation

[1197] The data collected on the server is fed into a natural language processing model, which uses a deep learning-based generative AI model (such as BERT or GPT). The model extracts important information from the text and generates a summary. This summary is then adjusted according to the user's sentiment.

[1198] Data provision

[1199] The summarized information is sent to the user's device as a push notification based on the user's notification schedule, allowing the user to quickly obtain important information. The emotion engine recognizes the user's emotions and adapts the notification content to the user's emotions. For example, if the user is anxious, the summary will be presented in a more understandable and simple language.

[1200] Feedback and Learning

[1201] Users can provide feedback on the summaries they receive. This feedback can include the usefulness of the summary, requests for additional information, and the user's emotional state. The server collects this feedback and stores it in a database. The emotion recognition engine also analyzes the emotional state at the time of feedback and reflects this data in the next summary generation.

[1202] Hardware and software used

[1203] Server: A central processing unit for collecting information and generating summaries (e.g., cloud server, on-premise server)

[1204] Crawler: uses the requests and BeautifulSoup libraries to gather information from web pages

[1205] Natural language processing models: Implementing BERT and GPT models using the transformers library

[1206] Emotion Recognition Engine: Microsoft Azure Cognitive Services Sentiment Analysis API

[1207] Database: A relational database (e.g., MySQL, PostgreSQL) to store collected information and feedback.

[1208] Specific examples

[1209] If User B sets his interest in entertainment news, the server will crawl entertainment-related websites to collect new articles, which will be summarized through a natural language processing model and adjusted based on User B's emotional state. For example,

[1210] "Prompt: Summarize this article in 300 words or less.

[1211] Text:

[1212] "In reviews of the recently released Star Wars film, many fans expressed excitement at the sequel, particularly the visual effects and story development, and many commented on how well the film connects to the previous films..."

[1213] In this way, User B can quickly access the content of interest and understand the content of a new movie based on the summarized information. Through the feedback function, User B can also provide their opinion on how useful the summary they received was and whether more details are needed. This feedback will be used to generate the next summary.

[1214] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1215] Step 1:

[1216] The server collects information from the Internet based on the user's interest categories. As input, the user's interest categories (e.g., technology, entertainment) are given, and as output, a list of URLs for related web pages is obtained. Specifically, a crawler (using the requests and BeautifulSoup libraries) crawls websites, discovers newly published articles, and stores them in a database.

[1217] Step 2:

[1218] The server retrieves articles from the collected URL list and converts them into text data. The URL list is given as input, and the text data of the retrieved articles is obtained as output. Specifically, the crawler accesses each URL, analyzes and extracts the HTML content, and obtains the text portion of the article.

[1219] Step 3:

[1220] The server inputs the collected text data into a natural language processing model to generate a summary. Text data (e.g., a news article) is given as input, and a summary is obtained as output. Specifically, it runs BERT or GPT models using the transformers library to extract the main points of the article and generate a summary of up to 300 characters.

[1221] Step 4:

[1222] The server then runs the generated summary through an emotion recognition engine to adjust the tone and content of the summary. The summary and the user's most recent emotional data are given as input, and the adjusted summary is obtained as output. Specifically, the server uses the sentiment analysis API of Microsoft Azure Cognitive Services to evaluate the user's emotions and adjust the tone and expression of the summary based on that evaluation.

[1223] Step 5:

[1224] The server provides the adjusted summary to the user's device via a push notification. The adjusted summary and the user's notification schedule are given as input, and the server sends a push notification to the user's device as output. Specifically, the server uses a push notification service such as Firebase Cloud Messaging (FCM) to notify the summary to the user's smartphone or tablet.

[1225] Step 6:

[1226] The user checks the received summary and provides feedback. The user's feedback (e.g., "It was helpful" or "I'd like more information") is given as input, and the feedback data is sent to the server as output. Specifically, when the user presses the feedback button in the app, the evaluation is recorded in the database.

[1227] Step 7:

[1228] The server analyzes the collected feedback and user emotion data and reflects the results in the next summary generation. The feedback data and emotion data are given as input, and an adjusted natural language processing model is obtained as output. Specifically, the feedback and emotion data are used to update the model parameters and improve the accuracy of the next summary generation.

[1229] This series of processing steps allows users to efficiently collect and review information that interests them in a personalized way, and to receive even more accurate information through feedback.

[1230] 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.

[1231] 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.

[1232] 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.

[1233] [Fourth embodiment]

[1234] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1235] 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.

[1236] 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).

[1237] 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.

[1238] 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.

[1239] 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).

[1240] 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.

[1241] 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.

[1242] 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.

[1243] 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.

[1244] 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.

[1245] 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.

[1246] 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."

[1247] The present invention provides a system for collecting information from the Internet based on the user's interest categories, summarizing the information, and providing it to the user. Specific embodiments of the present invention are described below.

[1248] Overall system overview

[1249] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means of notifying the user of the summarized information, and a mechanism for collecting user feedback and adjusting the system.

[1250] Information gathering

[1251] Server: Launches a crawler that collects data from specific websites and social media platforms based on the user's interest categories. The crawler periodically visits and collects new content to store in the database.

[1252] Example: If user A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[1253] Summary Generation

[1254] Server: The collected data is fed into a natural language processing model (e.g., a deep learning model). This model extracts important information from the text and generates a summary. The model is adjusted based on user feedback to improve its accuracy.

[1255] Example: When a long technical article is collected, it is summarized using an NLP model to extract only the main points, for example, a 5000-character article is converted into a 300-character summary.

[1256] Data provision

[1257] Server: Summarized information is organized by user and sent based on notification schedule.

[1258] On your device: The user's smartphone or tablet will receive a push notification that will appear on their screen. The user can then view the notification and read more if they are interested.

[1259] Example: If user A has configured their device to receive a summary at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information.

[1260] Feedback and Learning

[1261] Users: After viewing the summary, users can provide feedback on whether the information provided was useful, with options such as "helpful," "not helpful," or "I'd like more information."

[1262] Server: The server collects feedback and stores it in a database. This feedback data is used to improve the performance of the natural language processing model the next time a summary is generated.

[1263] Example: If user A gives feedback that the summary they received was very useful, that feedback can be used as training data for a model that can provide useful summaries to other users with similar interests.

[1264] ---

[1265] This allows users to efficiently gather important information from vast amounts of information and understand it in a short time. Furthermore, by utilizing user feedback, the system's accuracy can be continuously improved, enabling more personalized information to be provided.

[1266] The processing flow will be explained below.

[1267] ---

[1268] Step 1: User Registration and Setup

[1269] Users: When using the service for the first time, they create an account, enter their email address and password, and select the categories of their interests (e.g., "Technology," "Health," "Entertainment").

[1270] Server: Receives user input and stores user information and selected interest categories in a database.

[1271] Step 2: Gather information

[1272] Server: Based on the user's interest category, the query is passed to the crawler that is launched and configured to collect relevant information from specific websites and social media platforms.

[1273] Server: The crawler visits configured sites and gathers new content, such as articles from news sites or social media posts.

[1274] Server: Stores the collected data in a database.

[1275] Step 3: Summary generation

[1276] Server: Inputs the collected data into a natural language processing (NLP) model. The server invokes this NLP model to analyze the collected text data.

[1277] Server: The NLP model extracts important information from the text and generates a summary, for example, summarizing a 5,000-character article into 300 characters.

[1278] Server: Organizes the generated summaries for each user and stores them in a database.

[1279] Step 4: Provide data

[1280] Server: Once the summary is generated, it sends the summarized information to the user's device based on the user's notification schedule.

[1281] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[1282] Step 5: Feedback and learning

[1283] Users: After reviewing the summarized information, they can provide feedback by selecting from options such as "helpful," "not helpful," or "I'd like more information."

[1284] Server: Receives the provided feedback and stores it in a database.

[1285] Server: Uses the feedback data to adjust the parameters of the natural language processing model to improve the accuracy of summary generation, thereby improving the model so that the next summary is more suited to the user's preferences.

[1286] ---

[1287] This processing step allows users to efficiently gather information specific to their interests and quickly grasp the main points, while feedback is used to continuously improve the accuracy and user experience of the overall system.

[1288] Example 1

[1289] 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."

[1290] In today's Internet society, users are overwhelmed by the vast amount of information available, making it difficult for them to efficiently gather and understand important information tailored to their interests and needs. Furthermore, few systems provide feedback on the appropriateness of collected information to improve the quality of subsequent information provision. As a result, users often waste time and effort. To solve this problem, a method is needed to collect information based on users' interests, summarize it, notify them at the appropriate time, and utilize feedback to improve the accuracy of the system.

[1291] 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.

[1292] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for storing the collected data in a database, means for transmitting the summarized information based on a notification schedule, and means for collecting user feedback and adjusting the natural language processing model for the next summary generation. This allows users to efficiently collect important information and understand it in a short time. Furthermore, by utilizing user feedback, the accuracy of the system can be continuously improved, enabling more personalized information to be provided.

[1293] "User" means any person or entity that uses the System to gather interest-based information and receive summarized information.

[1294] "Interest categories" are classifications that indicate themes or topics that interest a user.

[1295] The "Internet" is a global network system that connects computers and servers around the world.

[1296] "Information collection methods" are the processes and tools used to obtain data from specific websites or social media platforms on the internet.

[1297] A "natural language processing model" is an artificial intelligence technology for understanding and analyzing text data, and specifically includes deep learning models.

[1298] A "summarization tool" is a process or tool that condenses collected information and extracts its main points.

[1299] A "terminal" is an electronic device used by a user to receive and view information, and specifically includes a smartphone or tablet.

[1300] "Means of Notification" means the method or mechanism for notifying the user of the summarized information, including push notifications.

[1301] "Feedback collection methods" are the methods and processes used to obtain user responses and ratings.

[1302] A "database" is a system for systematically storing and managing collected data and abstract data.

[1303] A "notification schedule" is a plan or setting that controls when information is notified to a user.

[1304] A "server" is a computing system for performing the processes of information gathering, summary generation, notification, feedback gathering, and data storage.

[1305] The present invention provides a system for collecting information from the Internet based on the user's interest categories, summarizing the information, and providing it to the user. Specific embodiments of the present invention are described below.

[1306] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, and a mechanism for collecting user feedback and adjusting the system.

[1307] Information gathering

[1308] The server launches a crawler based on the user's interest categories to collect data from specific websites and social media platforms. This crawler uses a data collection library, such as Scrapy. The server stores the collected data in a database (such as MySQL or MongoDB).

[1309] Specific examples

[1310] If User A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[1311] Summary Generation

[1312] The server inputs the collected data into a natural language processing model (e.g., a deep learning model like GPT-4 or BERT). The model extracts important information from the text and generates a summary. It then adjusts the model based on user feedback to improve its accuracy.

[1313] Specific examples

[1314] When long technical articles are collected, they are summarized using a natural language processing model to extract only the main points, for example, a 5,000-character article is converted into a 300-character summary.

[1315] Data provision

[1316] The server organizes the summarized information for each user and sends it based on a notification schedule, using scheduler software (e.g., a cron job), and the information is sent to the user's device via a push notification service (e.g., Firebase Cloud Messaging).

[1317] Specific examples

[1318] If user A has configured their device to receive a summary at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information.

[1319] Feedback and Learning

[1320] After reviewing the summarized information, users can provide feedback on whether the information was useful or not. Feedback consists of options such as "helpful," "not helpful," or "I'd like more information." The server stores this feedback in a database and uses it to improve the performance of the natural language processing model the next time a summary is generated.

[1321] Specific examples

[1322] If user A gives feedback that the summary they received was very useful, that feedback can be used as training data for the model, allowing it to provide useful summaries to other users with similar interests.

[1323] Example prompts to input to the generative AI model

[1324] "I would like to design a system that collects information from the Internet based on interest categories set by the user and summarizes that information. What are the specific processing steps and what hardware and software are used at each step? In particular, could you please provide more details about the steps of information collection, summary generation, data provision, and user feedback and learning?"

[1325] As a result, the present invention allows users to efficiently gather important information from vast amounts of information and understand it in a short time. Furthermore, by utilizing user feedback, the system's accuracy can be continuously improved, enabling more personalized information to be provided.

[1326] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1327] Step 1: User Interest Category Setting

[1328] User: The user sets the category of interest, for example, choosing a theme such as "Technology" or "Health."

[1329] Input: User selected interest categories.

[1330] Output: Interest category information passed to the server.

[1331] Step 2: Start the crawler

[1332] Server: The server launches a crawler based on the user's interest categories. The crawler is configured to visit specified websites and social media platforms. This crawler uses a library such as Scrapy.

[1333] Input: User interest category information.

[1334] Output: Collected article and post data.

[1335] Specific behavior:

[1336] The server looks up the user's interest categories and generates an appropriate URL list.

[1337] A crawler is launched based on the generated URL list to collect new articles and posts.

[1338] Step 3: Save to database

[1339] Server: Stores the collected data in a database, such as MySQL or MongoDB.

[1340] Input: The data retrieved by the crawler (articles, posts).

[1341] Output: Collected data stored in a database.

[1342] Specific behavior:

[1343] The crawler inserts the collected data into a database along with information such as text, URL, date and time.

[1344] Once the database has been saved, proceed to the next step.

[1345] Step 4: Summarize the data

[1346] Server: Inputs the collected data into a natural language processing (NLP) model, such as a deep learning model like GPT-4 or BERT.

[1347] Input: Article text extracted from the database.

[1348] Output: A summary sentence generated by the NLP model.

[1349] Specific behavior:

[1350] The server extracts new articles from the database and feeds the text data into the NLP model.

[1351] An NLP model extracts important information and generates a summary.

[1352] The generated summaries are stored in a database.

[1353] Step 5: Set up a notification schedule

[1354] Server: Organize the summarized information for each user and set up notifications to be sent based on a notification schedule, such as using a cron job.

[1355] Input: The notification schedule set by the user.

[1356] Output: Scheduled notification jobs.

[1357] Specific behavior:

[1358] The server refers to the user's notification schedule and sets the timing for sending the summary data.

[1359] When the scheduler reaches the set time, a push notification is triggered.

[1360] Step 6: Sending push notifications

[1361] Server: Posts summarized information.

[1362] Device: The user's smartphone or tablet receives the push notification and displays the information. Push notifications are sent using Firebase Cloud Messaging (FCM) or similar.

[1363] Input: Summary data, Notification schedule.

[1364] Output: The push notification sent to the user device.

[1365] Specific behavior:

[1366] The server sends a push notification, and the user device receives it.

[1367] A notification will appear on the device and users can click to view more information.

[1368] Step 7: Gather feedback

[1369] User: The user reviews the summarized information and provides feedback on whether the information was helpful. Feedback options include "helpful," "not helpful," and "I'd like more information."

[1370] Input: User ratings and comments.

[1371] Output: Feedback information sent to the server.

[1372] Specific behavior:

[1373] The user sees the notification and selects a feedback option.

[1374] The selected feedback is sent to the server.

[1375] Step 8: Save and learn from feedback

[1376] Server: The server collects the feedback and stores it in a database. This data is used to improve the performance of the natural language processing model the next time a summary is generated.

[1377] Input: Feedback information submitted by the user.

[1378] Output: Feedback stored in a database and a retrained NLP model.

[1379] Specific behavior:

[1380] The server stores the received feedback in a database.

[1381] Based on the stored feedback data, the NLP model is retrained to improve accuracy.

[1382] (Application example 1)

[1383] 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."

[1384] In virtual stores, it is difficult to efficiently obtain information about products and special offers that interest users. A method is needed to extract important information from a vast amount of information and provide it to users quickly. Another challenge is to understand whether the provided information was useful to users and collect feedback to improve the accuracy of the system.

[1385] 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.

[1386] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for collecting feedback from the user and adjusting the natural language processing model for the next summary generation, and means for collecting the latest information and special offers from virtual stores, summarizing them, and notifying the user of them. This enables the user to efficiently collect important information from a vast amount of information and understand it in a short time.

[1387] "User interest categories" indicate the types or areas of information in which a user is particularly interested.

[1388] "Methods of collecting information from the internet" refers to technologies used to obtain data from websites and social media platforms.

[1389] A "natural language processing model" refers to an artificial intelligence model for interpreting and processing human language.

[1390] "Means of summarizing" refers to the technique of extracting important points from collected information and summarizing them in short sentences.

[1391] "User device" refers to electronic devices such as smartphones, tablets, smart glasses, and head-mounted displays.

[1392] "Means for collecting feedback" refers to the technology that collects user ratings and opinions and stores them in a database.

[1393] "Means for adjusting the natural language processing model" refers to methods for improving the accuracy of the natural language processing model based on user feedback.

[1394] A "virtual store" refers to a virtual commercial space set up on the Internet.

[1395] "Means for collecting the latest information and special offers" refers to technology for efficiently collecting new product information and special offer information provided by virtual stores.

[1396] "Means of notification" refers to technology that sends information to the user's device so that it can be checked in real time.

[1397] A system for implementing this invention mainly comprises the following components: means for collecting information from the Internet based on a user's interest categories; means for summarizing the collected information using a natural language processing model; means for notifying the user of the summarized information on their terminal; means for collecting feedback from the user and adjusting the natural language processing model for the next summary generation; and means for collecting, summarizing, and notifying the user of the latest information and special offers from virtual stores.

[1398] Information Gathering Process

[1399] The server collects data from specific websites and social media platforms based on the user's defined interest categories. It uses a crawler to periodically visit these platforms and collect new content. For example, if a user defines their interest in "fashion," the server will crawl fashion-related news sites and blogs, collect newly published articles, and store them in a database.

[1400] Summary Generation Process

[1401] The server inputs the collected data into a natural language processing model (e.g., a deep learning model), extracts important information, and generates a summary. This model is adjusted based on user feedback, improving the accuracy of the next summary generation. As a specific example, if a 2,000-character fashion-related article is collected, the natural language processing model converts it into a 300-character summary. This allows users to grasp important information in a short amount of time.

[1402] Data provision process

[1403] The server organizes the summarized information for each user and sends it to the user's device based on the notification schedule. The device (such as a smartphone or smart glasses) receives the push notification and displays the summarized information on its screen. For example, if a user has configured their device to receive summary information at 8:00 a.m., the server will send a notification to the user's device every morning at 8:00 a.m. providing the latest fashion-related information.

[1404] Feedback Collection Process

[1405] After reviewing the summarized information, users can provide feedback on whether they found it useful. This feedback includes options such as "helpful," "not helpful," or "I'd like more information." The server collects this feedback, stores it in a database, and uses it the next time it generates a summary. This makes it possible to continuously improve the performance of the natural language processing model. For example, if a user provides feedback that "information on the latest fashion collections" was useful, that evaluation is used as training data for the model.

[1406] Hardware and software used

[1407] Natural Language Processing Models: Pipeline of the transformers library

[1408] Data collection: using the requests library

[1409] Crawlers: crawlers that collect data from websites and social media platforms

[1410] Prompt Sentence Examples

[1411] Input: "Latest fashion collection information"

[1412] Prompt: "Summarize an article about the latest fashion collection."

[1413] This system allows users to efficiently gather information that is important to them from a vast amount of information and understand it in a short time. Furthermore, the accuracy of the system is continuously improved through feedback, realizing more personalized information provision.

[1414] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1415] Step 1:

[1416] User interest category settings

[1417] The user specifies the categories of interest (e.g., fashion, technology, sports, etc.). The user's interest categories are sent as input to the server, which stores the interest categories in a database. The server uses this information for subsequent processing.

[1418] Step 2:

[1419] Information gathering

[1420] The server uses a crawler to visit specific websites and social media platforms based on the interest categories set by the user. As input, it takes the user's interest categories and a list of target URLs, and the crawler visits those URLs to collect new content. The collected data is stored in a database in text format.

[1421] Step 3:

[1422] Summary Generation

[1423] The server inputs the text data acquired in the information collection step into a natural language processing model (e.g., a deep learning model). As input, the collected text data is passed to a summary generation model, which extracts important information and summarizes it into a short sentence. For example, a 2,000-character article is converted into a 300-character summary using the prompt "Please summarize an article about the latest fashion collection." The generated summary is stored in a notification database.

[1424] Step 4:

[1425] Notification Schedule

[1426] The server manages the notification schedule specified for each user. It reads the user's notification setting time as input and registers summary information in the notification schedule according to that time. Notifications are sent at the specified time based on the registered schedule.

[1427] Step 5:

[1428] Push notifications

[1429] The server sends push notifications to the user's device based on the notification schedule. It uses the scheduled summary information and the user's device information as input. The device receives the push notification and displays the summary information on the screen. The user can check the notification and view more information if interested.

[1430] Step 6:

[1431] Feedback collection

[1432] Users provide feedback on the summary notification they receive. As input, the feedback provided by the users (e.g., "It was helpful," "It wasn't helpful," "I'd like more information") is collected and sent to the server. The feedback information is stored in a database.

[1433] Step 7:

[1434] Model adjustment

[1435] The server adjusts the natural language processing model based on the collected feedback information. It takes the feedback data stored in the database as input and uses it as training data for the model. This improves the accuracy of the next summary generation. Specifically, it uses the feedback data to update the model parameters, enabling it to generate summaries that better match the user's interests.

[1436] These steps allow users to efficiently gather and quickly understand the latest information relevant to their interests, and also allow for continuous improvement of the system through feedback.

[1437] 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.

[1438] The present invention is a system that collects information from the Internet based on the user's interest categories, summarizes the information, and provides it to the user. It also recognizes the user's emotions and reflects them in the summary and feedback processing. Specific embodiments of the present invention are described below.

[1439] Overall system overview

[1440] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, and an emotion engine for recognizing the user's emotions. It also has a mechanism for collecting user feedback and reflecting it in the next summary generation.

[1441] Information gathering

[1442] Server: Launches a crawler that collects data from specific websites and social media platforms based on the user's interest categories. The crawler periodically visits and collects new content to store in the database.

[1443] Example: If user A specifies that he is interested in "technology," the server will crawl technology-related news sites and blogs to collect newly published articles, which will then be stored in the database.

[1444] Summary Generation

[1445] Server: The collected data is fed into a natural language processing (NLP) model. This model extracts important information from the text and generates a summary. The model is adjusted based on user feedback to improve accuracy. User sentiment is also taken into account to adjust the summary.

[1446] Example: When a long technical article is collected, the article is summarized by an NLP model to extract only the main points. For example, a 5000-character article is converted into a 300-character summary. If the user is excited, a positive tone is added to the summary.

[1447] Data provision

[1448] Server: The summarized information is organized for each user and sent based on the notification schedule. An emotion engine is also used here to notify users based on their current emotional state.

[1449] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[1450] Example: If user A has configured their device to receive summary information at 8:00 AM, the server will send a notification to user A's device every morning at 8:00 AM. This notification will contain a summary of the latest technology information. If user A is feeling anxious, the information will be adjusted to be more easily understood.

[1451] Feedback and Learning

[1452] User: After reviewing the summarized information, the user can provide feedback by selecting from options such as "helpful," "not helpful," or "want more information." The user's emotional state at the time of providing feedback is also captured.

[1453] Server: Receives the provided feedback, stores it in a database, analyzes it using the emotion engine, and uses the results to adjust the NLP model.

[1454] Example: If user A gives feedback that the summary they received was very useful, that feedback and their emotional state at the time can be used as training data for the model to provide useful summaries to other users with similar interests. For example, we can strengthen the summarization method that receives a lot of positive feedback.

[1455] ---

[1456] This allows users to efficiently gather important information from a vast amount of information and understand it in a short time. Also, by taking into account the user's emotions, the present invention provides more personalized information and improves the user experience.

[1457] The processing flow will be explained below.

[1458] ---

[1459] Step 1: User Registration and Setup

[1460] Users: When using the service for the first time, they create an account, enter their email address and password, and select the categories of their interests (e.g., "Technology," "Health," "Entertainment").

[1461] Server: Receives user input and stores user information and selected interest categories in a database.

[1462] Step 2: Gather information

[1463] Server: Launches crawlers that collect data from specific websites and social media platforms based on user interest categories. The crawlers periodically traverse the site, collecting new content and storing it in a database.

[1464] What happens: The server instructs the crawler to visit health-related news sites and blogs based on User B's interest category "health." The crawler collects articles and posts and stores them in a database.

[1465] Step 3: Summary generation

[1466] Server: The collected data is fed into a natural language processing (NLP) model, which extracts key information from the text and generates a summary.

[1467] Server: The emotion engine analyzes the user's current emotional state and reflects the results in the summary generation process.

[1468] How it works: The server inputs a long health-related article into the NLP model and generates a summary. The emotion engine detects "positive emotions" from User B's recent feedback and applies positive language to the summary.

[1469] Step 4: Provide data

[1470] Server: The summarized information is organized for each user and sent based on the notification schedule. An emotion engine is also used here to notify users based on their current emotional state.

[1471] Device: The user's smartphone or tablet receives the push notification and displays it on the screen. The user can then view the notification and read more if they are interested.

[1472] Specific operation: The server pushes a summary of the latest health information to User B's device every morning at 8:00. User B's device receives the notification and displays a positive summary on the screen: "Latest health news: Simple exercises you can do starting today."

[1473] Step 5: Feedback and learning

[1474] User: After reviewing the summarized information, the user can provide feedback by selecting from options such as "helpful," "not helpful," or "want more information." The user's emotional state at the time of providing feedback is also captured.

[1475] Server: Receives the provided feedback, stores it in a database, analyzes it using the emotion engine, and uses the results to adjust the NLP model.

[1476] Specific operation: User B gives feedback that the service was "helpful," and the emotion engine determines the emotion at that time as "delight." This feedback and emotion data are stored in the database and used to optimize the NLP model.

[1477] ---

[1478] This processing step allows users to efficiently collect information specific to their interests and grasp the main points in a short time. Also, by taking into account the user's emotions, more personalized information can be provided and the user experience can be improved.

[1479] Example 2

[1480] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1481] Currently, there is a huge amount of information on the Internet, and it is difficult to efficiently collect, summarize, and provide it to users. Furthermore, there is no method for providing personalized information that takes user emotions into account, making it difficult to improve the user experience. Furthermore, there is a lack of a mechanism for reflecting user feedback to improve the accuracy of information provided.

[1482] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1483] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for notifying the user of the summarized information, means for recognizing the user's emotional state and adjusting the summary content and notification method, and means for collecting feedback and emotions from the user and adjusting the natural language processing model for next summary generation, thereby enabling the user to efficiently collect important information and understand it in a short time, as well as providing personalized information and improving the user experience.

[1484] "User" refers to the ultimate user who collects information and receives summarized data.

[1485] "Interest categories" refer to themes or genres in which a user is particularly interested.

[1486] The "Internet" refers to a huge information network that interconnects computer networks all over the world.

[1487] "Means of collecting information" refers to the technical methods and devices used to search for data on the Internet and obtain specific content.

[1488] A "natural language processing model" refers to a computational model for understanding and generating human language.

[1489] A "summary" is a piece of text that briefly summarizes the essence of a longer piece of text.

[1490] "Terminal" refers to a device used by a user, such as a computer or smartphone.

[1491] "Means of notification" refers to the communication technologies and protocols used to deliver the generated information to users.

[1492] "Emotional state" refers to the user's psychological state or mood.

[1493] "Adjustment" refers to technologies and algorithms that change the way information is presented in response to user sentiment and feedback.

[1494] "Feedback" refers to a user's evaluation or opinion of the information provided.

[1495] "Generative AI model" refers to an artificial intelligence model used for generative tasks (e.g., text generation, summary generation, etc.).

[1496] A "prompt" refers to an instruction given to a generative AI model to perform a specific task.

[1497] The present invention is a system that collects information from the Internet based on the user's interest categories, summarizes it, and provides it to the user. Furthermore, it recognizes the user's emotions, adjusts the summary content and notification method, and reflects feedback to improve the accuracy of the next summary generation. Specific embodiments are described below.

[1498] Overall system overview

[1499] This system mainly consists of the following components: a server for collecting information from the Internet, a natural language processing model for summarizing the information, a means for notifying the user of the summarized information, an emotion engine for recognizing the user's emotions, and a means for collecting feedback from the user and reflecting it in the next summary generation.

[1500] Information gathering

[1501] The server launches a crawler based on the user's interest categories. The crawler uses software such as Scrapy or Beautiful Soup to periodically collect information from specific websites and social media. The collected data is then stored in a database such as MySQL.

[1502] Examples:

[1503] Assume that User A is interested in "technology." The server uses Scrapy to crawl technology-related news sites and blogs and collect newly published articles. The collected articles are stored in a MySQL database.

[1504] Summary Generation

[1505] The server inputs the collected data into a natural language processing model (such as OpenAI's GPT-3) to generate a summary. The summary is then adjusted based on the user's emotional state. User feedback and emotional data are also collected to help improve the model's accuracy.

[1506] Examples:

[1507] Once long technology-related articles are collected, they are summarized by GPT-3. For example, a 5,000-character article is summarized into 300 characters. If the user is excited, a positive tone is added to the summary. Example prompt: "Summarize the following technology article in 300 characters or less."

[1508] Data provision

[1509] The server organizes the summarized information for each user and sends it based on a notification schedule. Using an emotion engine, the content of notifications is adjusted according to the user's current emotional state.

[1510] The device receives a push notification and displays it on the user's smartphone or tablet, where the user can check the notification and read more if interested.

[1511] Examples:

[1512] If User A has set up to receive information every morning at 8:00, the server will send summarized technology-related information to User A's device at 8:00. The information will be tailored to make it easier for User A to understand, especially if User A is feeling anxious.

[1513] Feedback and Learning

[1514] After checking the summary information, the user can provide feedback such as "It was helpful," "It wasn't helpful," or "I'd like more detailed information." The user's emotional state at the time of providing feedback is also captured.

[1515] The server stores the provided feedback in a database and analyzes it using an emotion engine, using the results to adjust the retraining dataset for the natural language processing model and improve the accuracy of future summary generation.

[1516] Examples:

[1517] If user A reviews the summary and gives feedback that it was "helpful" and their sentiment at the time is "positive," this data will be used to refine the model next time, so that other users can also receive useful summaries.

[1518] This system allows users to efficiently gather important information from a vast amount of information and understand it in a short time, and also improves the user experience by providing personalization based on user emotions.

[1519] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1520] Step 1:

[1521] Setting user interest categories

[1522] Users set their own interest categories through the device interface. The input is the user-selected interest categories, and the output is stored on the server.

[1523] Specific behavior:

[1524] 1. The user opens the smartphone app.

[1525] 2. On the "Interest Category Settings" screen, select a category such as "Technology," "Health," or "Sports."

[1526] 3. Once you have completed your selection, tap the "Save" button.

[1527] 4. The input is the selection data of interest categories, which is sent to the server.

[1528] 5. The server stores the received interest categories in a database.

[1529] Step 2:

[1530] Collection of information

[1531] The server launches a dedicated crawler based on the interest categories set by the user. The input is interest category information, and the output is collected website data. The collected data is stored in a database.

[1532] Specific behavior:

[1533] 1. Run the crawler once a day using a cron job.

[1534] 2. Use Scrapy to collect new articles from news sites.

[1535] 3. Store the collected articles in the “raw_articles” table in the MySQL database.

[1536] 4. As input, you have the website URL and crawling schedule.

[1537] 5. As output, you will get the collected text data.

[1538] Step 3:

[1539] Summary Generation

[1540] The server inputs the collected data into a natural language processing model (GPT-3) to generate a summary, taking into account the user's emotional data, and outputs a summary sentence.

[1541] Specific behavior:

[1542] 1. Retrieve new articles from the database.

[1543] 2. Call the GPT-3 API and summarize the article using a prompt, e.g., "Summarize the following article in 300 characters or less:"

[1544] 3. Adjust the tone of the summary depending on the user's emotional state (excitement, anxiety, etc.).

[1545] 4. Save the summary in the "summarized_articles" table in the database.

[1546] 5. The inputs are the text data of the retrieved articles and the user's sentiment data.

[1547] 6. As an output, a properly adjusted summary sentence is obtained.

[1548] Step 4:

[1549] Summary Notification

[1550] The server organizes the summarized information for each user and sends it based on a notification schedule. Using an emotion engine, the notification content is tailored to the user's current emotional state.

[1551] The device receives the push notification and displays it on the user's smartphone or tablet. The input is the user's schedule and summary information, and the output is the push notification.

[1552] Specific behavior:

[1553] 1. Organize summary information by user.

[1554] 2. If the notification is set to occur at 8:00 AM, the push notification will be sent based on the notification schedule.

[1555] 3. Use Firebase Cloud Messaging (FCM) to send notifications to the user's device.

[1556] 4. The user checks the notification and, if interested, displays the details in the app.

[1557] 5. As input, there is organized summary information and a notification schedule.

[1558] 6. As an output, a push notification will be displayed on the user's device.

[1559] Step 5:

[1560] Get user feedback

[1561] After checking the summary information, the user provides feedback. The inputs are the evaluation items and emotional state, and the output is the feedback data sent to the server.

[1562] Specific behavior:

[1563] 1. After checking the summary, tap the feedback button.

[1564] 2. Select from the feedback options and tap the send button.

[1565] 3. The emotional state at the time of feedback is also recorded.

[1566] 4. The inputs are the feedback content and emotion data.

[1567] 5. As an output, the feedback data is sent to the server and stored.

[1568] Step 6:

[1569] Analyzing feedback and adjusting the model

[1570] The server stores the provided feedback in a database and analyzes it using an emotion engine. Based on the results, it adjusts the retraining dataset for the natural language processing model. The inputs are the feedback and the emotional state, and the output is the adjusted model parameters.

[1571] Specific behavior:

[1572] 1. Retrieve feedback information from the database.

[1573] 2. Analyze feedback data using a sentiment engine.

[1574] 3. Using the analysis results, update the training data for the GPT-3 model and retrain it.

[1575] 4. The model is adjusted and reflected in the next summary generation.

[1576] 5. The inputs are the feedback data and sentiment analysis results.

[1577] 6. As an output, the adjusted model parameters are obtained, which are used for the next summary generation.

[1578] (Application example 2)

[1579] 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."

[1580] In today's world, it is becoming increasingly difficult for users to find valuable information from the vast amount of information available. Furthermore, there are only a limited number of systems that provide information based on user sentiment and feedback, creating a need for improved user experience. In particular, it is challenging to efficiently collect, summarize, and provide information in a personalized manner based on the user's interest categories.

[1581] 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.

[1582] In this invention, the server includes means for collecting information from the Internet based on the user's interest categories, means for summarizing the collected information using a natural language processing model, means for providing the summarized information to the user's device via push notification and adjusting the notification content based on the user's emotions, and means for collecting feedback from the user and their emotions at that time and reflecting them in the natural language processing model for the next summary generation. This allows the user to efficiently collect important information according to their interests and receive appropriate information according to their emotions.

[1583] "User interest categories" refer to specific themes or genres in which a user is personally interested.

[1584] "Means of collecting information from the internet" means methods of automatically obtaining information from certain websites and social media platforms on the internet.

[1585] A "natural language processing model" is a type of artificial intelligence used to understand and generate human language, such as summarizing and translating text.

[1586] "Summarization methods" refer to methods for extracting key information from large amounts of text data and presenting it in a shortened form.

[1587] "Providing push notifications to a user's device" means providing summarized information to a user's mobile device or computer in real time.

[1588] "Means for recognizing user emotions and adjusting notification content" refers to technology that assesses the user's emotional state and changes the way information is presented and the tone of the information based on that assessment.

[1589] "Collecting user feedback and emotions at the time" means collecting data on users' evaluations and impressions of the usefulness of the information provided, as well as their emotions at the time of evaluation.

[1590] "Means of informing the natural language processing model for the next summary generation" refers to a method of using the collected feedback and sentiment data to improve the future summarization process and generate more accurate summaries.

[1591] Overall system overview

[1592] This system collects information from the Internet based on the user's interest categories, summarizes it using a natural language processing model, and provides information based on the user's emotions. It also collects user feedback and emotion data and reflects it in the next summary generation. The system mainly consists of the following components:

[1593] 1. Information collection server

[1594] 2. Natural Language Processing Model for Summarization

[1595] 3. User terminal for providing summarized information

[1596] 4. Emotion Recognition Engine

[1597] 5. Feedback and learning modules

[1598] Information gathering

[1599] The server automatically collects information from the internet based on the interest categories set by the user. It uses a crawler to retrieve information from specific websites and social media platforms and stores that information in a database. For example, if a user sets their interest in "technology," the server will periodically crawl technology-related news sites and blogs to collect newly published articles.

[1600] Summary Generation

[1601] The data collected on the server is fed into a natural language processing model, which uses a deep learning-based generative AI model (such as BERT or GPT). The model extracts important information from the text and generates a summary. This summary is then adjusted according to the user's sentiment.

[1602] Data provision

[1603] The summarized information is sent to the user's device as a push notification based on the user's notification schedule, allowing the user to quickly obtain important information. The emotion engine recognizes the user's emotions and adapts the notification content to the user's emotions. For example, if the user is anxious, the summary will be presented in a more understandable and simple language.

[1604] Feedback and Learning

[1605] Users can provide feedback on the summaries they receive. This feedback can include the usefulness of the summary, requests for additional information, and the user's emotional state. The server collects this feedback and stores it in a database. The emotion recognition engine also analyzes the emotional state at the time of feedback and reflects this data in the next summary generation.

[1606] Hardware and software used

[1607] Server: A central processing unit for collecting information and generating summaries (e.g., cloud server, on-premise server)

[1608] Crawler: uses the requests and BeautifulSoup libraries to gather information from web pages

[1609] Natural language processing models: Implementing BERT and GPT models using the transformers library

[1610] Emotion Recognition Engine: Microsoft Azure Cognitive Services Sentiment Analysis API

[1611] Database: A relational database (e.g., MySQL, PostgreSQL) to store collected information and feedback.

[1612] Specific examples

[1613] If User B sets his interest in entertainment news, the server will crawl entertainment-related websites to collect new articles, which will be summarized through a natural language processing model and adjusted based on User B's emotional state. For example,

[1614] "Prompt: Summarize this article in 300 words or less.

[1615] Text:

[1616] "In reviews of the recently released Star Wars film, many fans expressed excitement at the sequel, particularly the visual effects and story development, and many commented on how well the film connects to the previous films..."

[1617] In this way, User B can quickly access the content of interest and understand the content of a new movie based on the summarized information. Through the feedback function, User B can also provide their opinion on how useful the summary they received was and whether more details are needed. This feedback will be used to generate the next summary.

[1618] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1619] Step 1:

[1620] The server collects information from the Internet based on the user's interest categories. As input, the user's interest categories (e.g., technology, entertainment) are given, and as output, a list of URLs for related web pages is obtained. Specifically, a crawler (using the requests and BeautifulSoup libraries) crawls websites, discovers newly published articles, and stores them in a database.

[1621] Step 2:

[1622] The server retrieves articles from the collected URL list and converts them into text data. The URL list is given as input, and the text data of the retrieved articles is obtained as output. Specifically, the crawler accesses each URL, analyzes and extracts the HTML content, and obtains the text portion of the article.

[1623] Step 3:

[1624] The server inputs the collected text data into a natural language processing model to generate a summary. Text data (e.g., a news article) is given as input, and a summary is obtained as output. Specifically, it runs BERT or GPT models using the transformers library to extract the main points of the article and generate a summary of up to 300 characters.

[1625] Step 4:

[1626] The server then runs the generated summary through an emotion recognition engine to adjust the tone and content of the summary. The summary and the user's most recent emotional data are given as input, and the adjusted summary is obtained as output. Specifically, the server uses the sentiment analysis API of Microsoft Azure Cognitive Services to evaluate the user's emotions and adjust the tone and expression of the summary based on that evaluation.

[1627] Step 5:

[1628] The server provides the adjusted summary to the user's device via a push notification. The adjusted summary and the user's notification schedule are given as input, and the server sends a push notification to the user's device as output. Specifically, the server uses a push notification service such as Firebase Cloud Messaging (FCM) to notify the summary to the user's smartphone or tablet.

[1629] Step 6:

[1630] The user checks the received summary and provides feedback. The user's feedback (e.g., "It was helpful" or "I'd like more information") is given as input, and the feedback data is sent to the server as output. Specifically, when the user presses the feedback button in the app, the evaluation is recorded in the database.

[1631] Step 7:

[1632] The server analyzes the collected feedback and user emotion data and reflects the results in the next summary generation. The feedback data and emotion data are given as input, and an adjusted natural language processing model is obtained as output. Specifically, the feedback and emotion data are used to update the model parameters and improve the accuracy of the next summary generation.

[1633] This series of processing steps allows users to efficiently collect and review information that interests them in a personalized way, and to receive even more accurate information through feedback.

[1634] 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.

[1635] 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.

[1636] 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.

[1637] 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.

[1638] 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.

[1639] 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.

[1640] 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).

[1641] 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.

[1642] 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."

[1643] 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.

[1644] 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).

[1645] 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.

[1646] 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.

[1647] 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.

[1648] 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.

[1649] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1650] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1651] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1652] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1653] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1654] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1655] The following is further disclosed regarding the above embodiment.

[1656] (Claim 1)

[1657] A means of collecting information from the internet based on user interest categories;

[1658] a means for summarizing the collected information using a natural language processing model;

[1659] a means for notifying the user of the summarized information on the terminal;

[1660] A means of collecting user feedback and adjusting the natural language processing model for the next summary generation; and

[1661] A system including:

[1662] (Claim 2)

[1663] 10. The system of claim 1, wherein the system uses a crawler to collect information from specific websites and social media platforms.

[1664] (Claim 3)

[1665] 10. The system of claim 1, wherein the natural language processing model is a deep learning model.

[1666] "Example 1"

[1667] (Claim 1)

[1668] A means of collecting information from the internet based on user interest categories;

[1669] a means for summarizing the collected information using a natural language processing model;

[1670] a means for notifying the user of the summarized information on the terminal;

[1671] A means of collecting user feedback and adjusting the natural language processing model for the next summary generation; and

[1672] a means for storing the collected data in a database;

[1673] means for transmitting summarized information based on a notification schedule;

[1674] A system including:

[1675] (Claim 2)

[1676] 10. The system of claim 1, wherein the system uses a crawler to collect information from specific websites and social media platforms.

[1677] (Claim 3)

[1678] 10. The system of claim 1, wherein the natural language processing model is a deep learning model.

[1679] "Application Example 1"

[1680] (Claim 1)

[1681] A means of collecting information from the internet based on user interest categories;

[1682] a means for summarizing the collected information using a natural language processing model;

[1683] a means for notifying the user of the summarized information on the terminal;

[1684] A means of collecting user feedback and adjusting the natural language processing model for the next summary generation; and

[1685] A means of collecting, summarizing and notifying users of the latest information and special offers from the virtual store;

[1686] A system including:

[1687] (Claim 2)

[1688] 10. The system of claim 1, wherein the system uses a crawler to collect information from specific websites and social media platforms.

[1689] (Claim 3)

[1690] 10. The system of claim 1, wherein the natural language processing model is a deep learning model.

[1691] "Example 2: Combining Emotion Engines"

[1692] (Claim 1)

[1693] A means of collecting information from the internet based on user interest categories;

[1694] a means for summarizing the collected information using a natural language processing model;

[1695] a means for notifying the user of the summarized information on the terminal;

[1696] A means of recognizing the user's emotional state and adjusting the summary content and notification method;

[1697] a means of collecting user feedback and sentiment to adjust the natural language processing model for next summary generation;

[1698] A system including:

[1699] (Claim 2)

[1700] 10. The system of claim 1, wherein the system uses a crawler to collect information from specific websites and social media platforms.

[1701] (Claim 3)

[1702] 10. The system of claim 1, wherein the system uses a generative AI model as a natural language processing model and generates a summary using a prompt sentence.

[1703] "Application example 2 when combining emotion engines"

[1704] (Claim 1)

[1705] A means of collecting information from the internet based on user interest categories;

[1706] a means for summarizing the collected information using a natural language processing model;

[1707] A means for providing summarized information to a user's device via push notification and adjusting the notification content based on the user's emotions;

[1708] A means to collect user feedback and emotions and feed them back into the natural language processing model for the next summary generation.

[1709] A system including:

[1710] (Claim 2)

[1711] 10. The system of claim 1, wherein the system uses a crawler to collect information from specific websites and social media platforms.

[1712] (Claim 3)

[1713] 10. The system of claim 1, wherein the natural language processing model is a deep learning model. [Explanation of symbols]

[1714] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of collecting information from the internet based on user interest categories; a means for summarizing the collected information using a natural language processing model; a means for notifying the user of the summarized information on the terminal; A means of collecting user feedback and adjusting the natural language processing model for the next summary generation; and A system including:

2. The system of claim 1 , wherein the system uses a crawler to collect information from specific websites and social media platforms.

3. The system of claim 1 , wherein the natural language processing model is a deep learning model.

Images