system

Abstract

Provide a system. [Solution] A means of collecting information in real time from SNS, means for filtering and pre-processing the collected information; means for analyzing the preprocessed information using natural language processing technology; a means for detecting faults and generating and sending alerts based on the analysis results; The system includes a means for visualizing detected faults on a dashboard.

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] Conventional methods for monitoring mobile phone services rely on monitoring the devices themselves, making it difficult to detect outages actually experienced by users in real time. This can result in a longer time between the occurrence of an outage and a response, potentially worsening the user experience. Another problem is that it is difficult to identify the specific area where the outage occurred, making it difficult to respond to the outage efficiently. [Means for solving the problem]

[0005] The present invention provides a system that includes a means for collecting information from social networking sites in real time, a means for filtering and preprocessing the collected information, a means for analyzing the preprocessed information using natural language processing technology, a means for detecting outages based on the analysis results, a means for generating and sending alerts, and a means for visualizing the detected outages on a dashboard. This enables real-time outage detection and rapid response based on outage information reported by users on social networking sites. Furthermore, by visualizing outage information in map and graph format, users can intuitively grasp the area where the outage is occurring, enabling efficient outage response.

[0006] "SNS" is an abbreviation for social networking services, an online platform for users to share information and interact with each other.

[0007] "Filtering" is a processing method for selecting necessary information from large amounts of data and removing unnecessary information.

[0008] "Preprocessing" is a processing method for preparing data and converting it into a form suitable for analysis before data analysis.

[0009] "Natural language processing technology" is a technology for analyzing and processing human language using a computer, and includes technologies such as text analysis and voice recognition.

[0010] "Analysis" is the process of breaking down data, understanding its contents, and extracting useful information.

[0011] "Fault detection" is the automatic identification of problems in systems and services.

[0012] An "alert" is a warning message sent when a particular condition or event occurs.

[0013] A "dashboard" is an interface that visually displays data and allows users to quickly grasp the current situation.

[0014] "Visualization" is a technique that makes it easier to understand a situation by visually representing data using graphs and charts.

[0015] "Real-time" means that data and information are processed as they are generated, with results almost immediately available.

[0016] "Region information" is information relating to a specific geographical area, including location, name, and the like.

[0017] A "notification" is an act or message that informs interested parties about specific information or events.

[0018] A "trouble response team" is a specialized team that quickly resolves system or service problems.

[0019] "Investigation" is the activity of identifying the cause of a problem that has occurred and finding a solution to it. [Brief explanation of the drawings]

[0020] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7]FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0021] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

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

[0023] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0024] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0025] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0026] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0027] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0028] [First embodiment]

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

[0030] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0031] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0032] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0033] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0034] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0035] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

[0037] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0038] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0039] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0040] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0041] The present invention is a system that collects information in real time from SNS (social networking services) and analyzes it to detect and quickly respond to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[0042] System Overview

[0043] This system mainly consists of the following components:

[0044] 1. Data collection module: Collects tweets from social media in real time.

[0045] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[0046] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[0047] 4. Fault detection module: Detects faults based on the analysis results and generates alerts.

[0048] 5. Visualization module: Displays detected faults on a dashboard.

[0049] 6. Notification module: Sends notifications to relevant parties when a fault is detected.

[0050] Program processing procedure

[0051] Data collection and preprocessing

[0052] The server collects tweets in real time from social networking sites such as Twitter. Specifically, it uses the Twitter API to retrieve tweets related to specific keywords (e.g., "cell phone signal interference") and stores them in a temporary database.

[0053] The server preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[0054] Data analysis

[0055] The server analyzes the preprocessed tweet text using natural language processing technology. Specifically, it uses a morphological analysis tool to break it down into words and calculates the frequency of specific keywords. It then determines whether the tweet contains many keywords related to outages (e.g., "can't connect," "signal," "outage," etc.).

[0056] Fault detection and notification

[0057] The server determines if a keyword frequency exceeds a certain threshold, taking into account location and other relevant information. If an outage is detected, the server generates an alert and sends a notification to relevant parties via email, SMS, or a dedicated alert management system.

[0058] Information Visualization

[0059] The server displays detected fault information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[0060] Specific examples

[0061] For example, if mobile phone signals suddenly become unstable in a certain area, multiple users will post about the problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0062] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords from the tweets, such as "Shibuya," "signal," and "can't connect." If the frequency of these keywords exceeds a certain threshold, the server determines that a signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding. On-site investigations and remote diagnostics are conducted to restore service.

[0063] In this way, this system can improve the quality of mobile phone services more effectively than conventional methods by utilizing real-time information from SNS to detect failures early and enable rapid response.

[0064] The processing flow will be explained below.

[0065] Step 1:

[0066] The server connects to the API of the social media platform and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference") and uses the streaming API to receive data each time a new tweet is posted.

[0067] Step 2:

[0068] The server stores the received tweets in a temporary database that is kept for a short period of time for subsequent processing.

[0069] Step 3:

[0070] The server retrieves tweet data collected over a fixed period of time (e.g., 1 minute) for batch processing. The data is filtered as needed, for example, to remove irrelevant tweets such as advertisements and spam.

[0071] Step 4:

[0072] The server then breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the structure of the tweet is analyzed and meaningful words and phrases are extracted.

[0073] Step 5:

[0074] The server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[0075] Step 6:

[0076] If the frequency of a keyword exceeds a pre-defined threshold, the server determines that a particular region is experiencing an outage, based on location data from tweets and user profile information.

[0077] Step 7:

[0078] The server generates an alert when a failure is detected, which includes detailed information about the failure (e.g., the location where it occurred, the time of occurrence, and related keywords).

[0079] Step 8:

[0080] The server notifies designated parties (e.g., incident response teams) of generated alerts via email, SMS, or a dedicated alert management system.

[0081] Step 9:

[0082] The device (the fault response team) receives the notification and immediately begins to act. The notification includes detailed information about the fault and the specific area that needs to be addressed.

[0083] Step 10:

[0084] The device will use on-site inspections and remote diagnostics to identify the cause of the fault and take necessary measures, such as resetting the base station or updating the software.

[0085] Step 11:

[0086] The server displays the progress of the problem and the results of the response in real time on a dashboard, allowing administrators to intuitively grasp the progress of the problem response.

[0087] Step 12:

[0088] Once the issue is resolved, the device updates the dashboard to indicate that the problem has been resolved, so that other stakeholders are aware of the situation.

[0089] Through the above steps, the system of the present invention can utilize real-time information obtained from SNS to quickly and efficiently detect and respond to mobile phone service outages.

[0090] Example 1

[0091] 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."

[0092] Conventional device monitoring systems have had difficulty in early detection of failures in mobile phone services. Furthermore, while rapid response is required after failure detection, current systems lack sufficient real-time information collection and analysis, resulting in delayed response. Therefore, the present invention aims to provide a system that collects information from SNS in real time and quickly detects and responds to failures.

[0093] 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.

[0094] In this invention, the server includes means for collecting data from SNS in real time, means for filtering and preprocessing the collected data, means for analyzing the preprocessed data using natural language processing technology, means for detecting faults based on the analysis results and generating and sending alerts, means for visualizing the detected faults on a dashboard, and means for analyzing the collected data and determining the presence or absence of faults using a generative AI model. This makes it possible to analyze the data collected from SNS in real time and detect faults quickly and accurately.

[0095] "SNS" is an abbreviation for social networking service, an online platform where users share information and opinions.

[0096] "Data collection" is the process of obtaining tweets and posts in real time from social media sites and other sources.

[0097] "Filtering" is the process of removing noise and unnecessary information from collected data and converting it into a form suitable for analysis.

[0098] "Preprocessing" refers to the process of formatting data before analyzing it, making it analyzable. Examples of preprocessing include converting uppercase letters to lowercase letters and removing advertisements.

[0099] "Natural language processing technology" is a technology that enables computers to understand, interpret, and generate human language, and includes techniques such as morphological analysis and sentiment analysis.

[0100] "Analysis" is the process of examining pre-processed data using natural language processing techniques to find specific trends and patterns.

[0101] "Fault detection" is the process of determining whether a system or service has some kind of malfunction or problem based on the analysis results.

[0102] "Alert generation" is the process of creating a warning message to communicate real-time information when a fault is detected.

[0103] "Sending" is the process of notifying relevant parties and administrators of the generated alert.

[0104] A "dashboard" is an interface for visually displaying detected fault information, and is designed to make it easy for administrators to understand the situation.

[0105] A "generative AI model" is a model that uses artificial intelligence technology to analyze data and detect specific patterns or anomalies.

[0106] "Preprocessed data" refers to data that has undergone preprocessing such as filtering and normalization and has been converted into a form suitable for analysis.

[0107] "Frequency" refers to the number of times a particular event or keyword occurs, and this number serves as the basis for determining whether or not there is a problem.

[0108] A "threshold" is a boundary value for satisfying a specific condition, and when exceeded, an alert is generated.

[0109] The present invention is a system that collects information from SNS in real time, analyzes it, and thereby detects and responds quickly to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[0110] System Configuration

[0111] This system mainly consists of the following components:

[0112] 1. Data collection module: Collects tweets from social media in real time.

[0113] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[0114] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[0115] 4. Fault detection module: Detects faults based on the analysis results and generates alerts.

[0116] 5. Visualization module: Displays detected faults on a dashboard.

[0117] 6. Notification module: Sends notifications to relevant parties when a fault is detected.

[0118] 7. Generative AI model: Analyzes collected data and determines whether a fault exists.

[0119] Hardware and software used

[0120] The server uses Twitter API and other SNS APIs to collect data from SNS in real time. The collected data is temporarily stored in a database. For data preprocessing, Python libraries (e.g., Pandas, Numpy) are used to remove noise and normalize text.

[0121] For data analysis, natural language processing techniques such as morphological analysis tools (e.g., MeCab) are used, along with generative AI models (e.g., TENSORFLOW (registered trademark) and PyTorch). Based on the analysis results, failures are detected, and alerts and notifications are generated using the SendGrid API and other email sending services.

[0122] Dashboard visualization uses a web application framework (e.g., Flask or Django) and a front-end visualization library (e.g., D3.js or Chart.js).

[0123] Specific examples

[0124] For example, if mobile phone signals suddenly become unstable in a certain area, multiple users will post about the problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0125] The server collects these tweets via the SNS API and stores them in a database. The server then extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets. If the frequency of these keywords exceeds a certain threshold, the server determines that a signal interference is occurring in the Shibuya area and generates an alert.

[0126] The generated alert is notified to the fault response team using the SendGrid API, and the device (trouble response team) immediately begins responding, conducting on-site investigations and remote diagnosis to restore service.

[0127] Examples of prompt statements

[0128] The following is a specific example of a prompt sentence to be input to the generative AI model:

[0129] "Analyze tweets containing keywords such as 'signal wave,' 'interference,' 'cannot connect,' and 'unstable' in the specified area and time period, and determine whether the frequency of the relevant keywords exceeds a certain threshold. Example: Tweets in Shibuya from 12:00 to 14:00 on October 12, 2023."

[0130] By using this prompt, data collected from social media can be analyzed by a generative AI model to quickly and accurately determine whether or not a disability exists.

[0131] As described above, this system analyzes data collected from SNS in real time, enabling rapid fault detection and response, thereby improving the quality of mobile phone services more effectively than conventional methods.

[0132] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0133] Step 1:

[0134] Data collection

[0135] The server collects tweets in real time from social networking sites such as Twitter. Specifically, it uses the Twitter API to retrieve tweets related to specific keywords (e.g., "cell phone signal interference"). The Twitter API key and search query are given as input, and the raw data of the retrieved tweets is stored in a database as output. Specifically, it periodically sends API requests to retrieve new tweets.

[0136] Step 2:

[0137] Data Preprocessing

[0138] The server pre-processes the collected data. As input, it takes raw data and as output, it gets pre-processed, clean data. The pre-processing includes the following specific operations:

[0139] Noise reduction: Utilize blacklists to remove advertising and spam tweets.

[0140] Text normalization: convert uppercase to lowercase and remove punctuation marks.

[0141] Language filtering: Filter out tweets in languages ​​other than Japanese.

[0142] Step 3:

[0143] Data analysis

[0144] The server analyzes the preprocessed data using natural language processing technology. The preprocessed clean data is given as input, and the analyzed data is obtained as output. Specifically, it breaks down the data into words using a morphological analysis tool (e.g., MeCab), extracts relevant keywords (e.g., "can't connect," "signal," "interference"), and calculates their frequency.

[0145] Step 4:

[0146] Fault Detection

[0147] The server detects a failure based on the analysis results. The analyzed data and a set threshold are given as input, and a flag indicating whether a failure has occurred and alert information are obtained as output. Specifically, if the frequency of a specific keyword exceeds the threshold, it determines that a failure has occurred in that area and generates an alert. Location information is also analyzed to identify the affected area.

[0148] Step 5:

[0149] Alert Generation

[0150] The server generates an alert when a failure is detected. The input is a failure detection flag and its detailed information (e.g., the area where the failure occurred, the time of occurrence), and the output is an alert message. Specifically, the server creates an alert message containing detailed information such as the type of failure, the area where the failure occurred, and the time of occurrence, and saves it in the database.

[0151] Step 6:

[0152] Send notifications

[0153] The server notifies the relevant parties of the generated alert. The alert message is given as input, and the sending result is obtained as output. Specifically, the server sends the alert using an API for sending email or SMS (e.g., SendGrid API), and sends the information to a dedicated alert management system.

[0154] Step 7:

[0155] Information Visualization

[0156] The server displays the detected outage information on a dashboard. The outage information is given as input, and the updated dashboard is obtained as output. Specifically, the outage information is displayed in real time using a web application framework (e.g., Flask or Django) and visualized using maps and graphs.

[0157] (Application example 1)

[0158] 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."

[0159] Conventional fault detection and ad delivery systems have difficulty collecting and analyzing information in real time, making it difficult to deliver ads effectively or quickly resolve faults. Furthermore, it is difficult to select optimal ads based on consumer opinions and emotions depending on the region and time of day, limiting the effectiveness of advertising.

[0160] 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.

[0161] In this invention, the server includes means for collecting information from SNS in real time, means for filtering and preprocessing the collected information, means for analyzing the preprocessed information using natural language processing technology, means for calculating region and emotion scores based on the analysis results and selecting advertisements, and means for sending the selected advertisements to users by push notification. This makes it possible to analyze consumer opinions and emotions in real time and quickly deliver optimal advertisements based on them.

[0162] "SNS" refers to social networking services, which are online platforms that allow users to interact with each other over the Internet.

[0163] "Real time" refers to processing that occurs simultaneously with actual time.

[0164] "Data filtering" is the process of selecting only necessary information from collected information.

[0165] "Preprocessing" refers to a series of steps that prepare information for analysis, including noise removal and text normalization.

[0166] "Natural language processing technology" refers to the technology that enables computers to understand and analyze human language, and is capable of performing a variety of language processing.

[0167] An "emotion score" is a numerical representation of the intensity of emotions contained in text data, classified as positive, negative, or neutral.

[0168] "Ad selection" refers to the process of selecting the most suitable advertisement based on the analysis results.

[0169] "Push notifications" are the ability for an application to send messages to a user's device in real time.

[0170] "User" refers to the general consumer who uses this system.

[0171] MODE FOR CARRYING OUT THE INVENTION

[0172] This invention is a system that collects and analyzes real-time information from social media and sends optimal advertisements to users via push notifications based on consumer opinions and emotions. The specific configuration and operation of this system are described below.

[0173] System Configuration

[0174] The system includes the following components:

[0175] 1. Data collection module: Collects information from social media in real time.

[0176] 2. Data preprocessing module: Filters and preprocesses the collected information.

[0177] 3. Analysis module: Analyzes the preprocessed information using natural language processing techniques and calculates the sentiment score.

[0178] 4. Advertisement selection module: Selects the most suitable advertisement based on the analysis results.

[0179] 5. Notification module: Send selected ads to users via push notifications.

[0180] Hardware and software used

[0181] Hardware: Servers, smartphones

[0182] software:

[0183] Twitter API, other SNS APIs

[0184] Database (SQLite)

[0185] Natural language processing libraries (MeCab, NLTK, BERT)

[0186] Push notification service (Firebase Cloud Messaging)

[0187] Operation process

[0188] 1. Data Collection and Preprocessing

[0189] The server collects tweets related to specific keywords (e.g., "new product," "sale," "recommended," etc.) from social media (e.g., Twitter) in real time. This is done using social media APIs such as the Twitter API. The collected tweet data is stored in a temporary database on the server.

[0190] The server then preprocesses the collected tweet data, which includes noise removal (e.g., removing spam messages) and text normalization (e.g., converting uppercase to lowercase).

[0191] 2. Data Analysis

[0192] The preprocessed tweet text is analyzed on the server using natural language processing techniques. Specifically, a morphological analysis tool (e.g., MeCab) is used to break it down into words and calculate the frequency of specific keywords. A sentiment analysis model (e.g., NLTK or BERT) is also used to calculate positive, negative, and neutral sentiment scores.

[0193] 3. Ad selection and delivery

[0194] Based on the analysis results, the ad selection module selects the most suitable ad, based on the sentiment score and location information, and sends it to the user via smartphone push notification.

[0195] Specific examples

[0196] For example, suppose that users in Shibuya post an increasing number of tweets about a new smartphone on social media, many of which are positive. The collected tweet data contains keywords such as "Shibuya," "smartphone," and "great." If the analysis confirms that the tweets have a high positive sentiment score, a promotional ad for the new smartphone will be sent to the user's smartphone during that time period.

[0197] Prompt Sentence Examples

[0198] "Collect the latest positive tweets about new smartphones in Shibuya, calculate the sentiment score, and select ads."

[0199] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0200] Step 1:

[0201] The server collects tweets related to specified keywords from social media (e.g., Twitter) in real time. Here, the Twitter API is used to obtain tweets that match keywords such as "new product" or "sale." These collected tweets are the input data, and the output is that they are stored in a temporary database on the server.

[0202] Step 2:

[0203] The server preprocesses the collected tweet data. This preprocessing includes removing advertisements and spam messages and normalizing the text (e.g., converting uppercase to lowercase). This process outputs clean text data suitable for analysis. This process filters out noise data, resulting in a clear database.

[0204] Step 3:

[0205] The server analyzes the preprocessed tweet text using natural language processing techniques (e.g., MeCab, NLTK, BERT). It performs morphological analysis, breaks it down to the word level, and calculates the frequency of keywords. At the same time, it uses a sentiment analysis model to calculate a positive, negative, or neutral sentiment score for each tweet. The output of this analysis process is a list of keyword frequencies and a list of sentiment scores.

[0206] Step 4:

[0207] The server selects an advertisement based on the analysis results from step 3. Using the sentiment score and keyword frequency obtained from the analysis results, it selects the most suitable advertisement (e.g., new product promotion or sale information). This selected advertisement is the output data.

[0208] Step 5:

[0209] The server sends the selected advertisement to the smartphone user via push notification. This process uses a push notification service such as Firebase Cloud Messaging to deliver the advertisement content to the user's smartphone. The user's receiving device acts as the push notification receiver, and the advertisement message is output.

[0210] Step 6:

[0211] The user can then view the push notification and respond to the ad if necessary. This final step involves the user's actions becoming the final form of input and output, as well as the collection of data to measure the effectiveness of the ad.

[0212] 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.

[0213] The present invention is a system that collects information from SNS (social networking services) in real time, analyzes the information, and recognizes emotions, thereby detecting and quickly responding to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[0214] System Overview

[0215] This system mainly consists of the following components:

[0216] 1. Data collection module: Collects tweets from social media in real time.

[0217] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[0218] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[0219] 4. Emotion Engine: Recognize user emotions from preprocessed tweet text.

[0220] 5. Fault detection module: Detects faults based on the analysis results and information from the emotion engine and generates alerts.

[0221] 6. Visualization module: Displays detected faults on a dashboard.

[0222] 7. Notification module: Sends notifications to relevant parties when a fault is detected.

[0223] Program processing procedure

[0224] Data collection and preprocessing

[0225] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"). Using the streaming API, it receives data as new tweets are posted. The received tweets are stored in a temporary database.

[0226] The server preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[0227] Data analysis and emotion recognition

[0228] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, it analyzes the structure of the tweet and extracts meaningful words and phrases.

[0229] Next, the server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[0230] Furthermore, the server uses an emotion engine to recognize the user's emotions (e.g., anger, frustration, sadness, etc.) from the tweet text, which provides data to evaluate the severity of the problem and the user's dissatisfaction level.

[0231] Fault detection and notification

[0232] The server determines that an outage is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold, taking into account location and other relevant information. If an outage is detected, the server generates an alert and sends a notification to relevant parties. This notification is sent via email, SMS, or a dedicated alert management system.

[0233] Information Visualization

[0234] The server displays detected fault information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[0235] Specific examples

[0236] For example, if mobile phone signals suddenly become unstable in a certain area, many users will post about this problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0237] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and finds that many of them contain feelings of dissatisfaction and anger.

[0238] Based on this information, the server determines that a serious signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding by conducting an on-site investigation and remote diagnosis to restore service.

[0239] The server displays the progress of the incident and the results of the response in real time on a dashboard, and also visualizes user emotional information, allowing administrators to understand the severity of the incident in more detail.

[0240] In this way, this system can utilize real-time information obtained from social media and emotion recognition technology to quickly and effectively detect and respond to mobile phone service outages.

[0241] The processing flow will be explained below.

[0242] Step 1:

[0243] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"), and uses the streaming API to receive data each time a new tweet is posted.

[0244] Step 2:

[0245] The server stores the received tweets in a temporary database that is kept for a short period of time for subsequent processing.

[0246] Step 3:

[0247] The server retrieves tweet data collected over a fixed time period (e.g., 1 minute) for batch processing. If necessary, the data is gallery-filtered to remove irrelevant tweets such as advertisements and spam.

[0248] Step 4:

[0249] The server then breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the structure of the tweet is analyzed and meaningful words and phrases are extracted.

[0250] Step 5:

[0251] The server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[0252] Step 6:

[0253] The server recognizes the user's emotions from the preprocessed tweet text using an emotion engine, which analyzes the text using natural language processing techniques to identify emotions such as anger, frustration, and sadness.

[0254] Step 7:

[0255] The server determines that an outage is occurring in a region when the frequency of keywords and user sentiment exceed a certain threshold. The region information is determined based on tweet location data and user profile information.

[0256] Step 8:

[0257] The server generates an alert when a failure is detected, which includes detailed information about the failure (e.g., the area where the failure occurred, the time of occurrence, and related keywords) and the user's emotional state.

[0258] Step 9:

[0259] The server notifies designated parties (e.g., incident response teams) of generated alerts via email, SMS, or a dedicated alert management system.

[0260] Step 10:

[0261] The device (the fault response team) receives the notification and immediately begins to act. The notification includes detailed information about the fault and the specific area that needs to be addressed.

[0262] Step 11:

[0263] The device will use on-site inspections and remote diagnostics to identify the cause of the fault and take necessary measures, such as resetting the base station or updating the software.

[0264] Step 12:

[0265] The server displays the progress and results of the problem in real time on a dashboard, which also visualizes user sentiment, allowing administrators to better understand the severity of the problem.

[0266] Step 13:

[0267] Once the issue is resolved, the device updates the dashboard to indicate that the problem has been resolved, so that other stakeholders are aware of the situation.

[0268] Through the above steps, the system of the present invention can utilize real-time information obtained from SNS and emotion recognition technology to quickly and effectively detect and respond to mobile phone service outages.

[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] Conventional fault detection systems for mobile phone services have had problems with detecting faults in real time and being unable to assess the severity of faults based on user emotional information. As a result, early detection and rapid response to faults have not been fully realized, leading to growing user dissatisfaction.

[0272] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for collecting data from SNS in real time, means for filtering and preprocessing the collected data, means for analyzing the preprocessed data using natural language processing technology, means for calculating the frequency of appearance of keywords from the analyzed data, means for recognizing emotions from the analyzed data, means for detecting faults based on the frequency of keywords and emotional information, and means for generating and transmitting alerts, and means for visualizing the detected faults on a dashboard. This enables real-time fault detection and assessment of the severity of faults based on user emotions, thereby realizing a prompt and effective response.

[0273] "SNS" refers to social networking services, which are online platforms where users can post information and share it with other users in real time.

[0274] "Real-time" refers to a state in which data and information are processed and provided immediately without delay.

[0275] "Data collection" refers to the process of gathering specific information or data according to specified criteria.

[0276] "Filtering" refers to the process of removing unnecessary data and noise and selecting data that is suitable for the purpose.

[0277] "Preprocessing" refers to a series of steps that transform data into a form suitable for analysis and processing.

[0278] "Natural language processing technology" refers to the technology that allows computers to understand and analyze human language.

[0279] "Analysis" refers to the process of breaking down given data or information into detail and understanding its meaning and structure.

[0280] "Keyword frequency" refers to the number of times a particular word or phrase appears in text data.

[0281] "Emotion recognition" refers to determining a user's emotional state from data such as text and voice.

[0282] "Fault detection" refers to the process of identifying and identifying anomalies or problems in a system or service.

[0283] An "alert" is a notification sent to warn you when an abnormality or emergency occurs.

[0284] A "dashboard" is an interface that displays information and data visually and intuitively, allowing administrators to understand the situation.

[0285] The present invention is a system that collects information from SNS (social networking services) in real time, analyzes the information, and recognizes emotions in the information, thereby detecting and quickly responding to problems in mobile phone services. The following describes in detail an embodiment of the present invention.

[0286] Data collection

[0287] The server uses a data collection API of a social networking platform, such as the Twitter API, to collect real-time tweets filtered by specific keywords (e.g., "cell phone signal interference"). This real-time data collection is achieved by a program that implements Twitter's streaming API. The data is stored in a temporary database on the server.

[0288] Data Preprocessing

[0289] The server preprocesses the collected tweet data, including noise removal (removing advertisements and spam) and text normalization (converting uppercase letters to lowercase), converting the data into a form suitable for analysis.

[0290] Data analysis

[0291] The server breaks down the preprocessed text data into words using a morphological analysis tool (e.g., MeCab). Morphological analysis analyzes the structure of tweets and extracts meaningful words and phrases. The server then analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.).

[0292] emotion recognition

[0293] The server uses an emotion engine to recognize the user's emotion (e.g., anger, frustration, sadness, etc.) from the tweet text. The emotion engine uses a pre-trained model (e.g., BERT). This analysis determines the intensity and type of the user's emotion, and evaluates the severity of the problem and the user's level of dissatisfaction.

[0294] Fault Detection and Notification

[0295] The server determines that an outage is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold. Location and other relevant information are also taken into account. When an outage is detected, the server generates an alert and sends a notification to relevant parties. Notifications are sent via email, SMS, or a dedicated alert management system.

[0296] Information Visualization

[0297] The server displays detected fault information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[0298] Specific examples

[0299] For example, if mobile phone signals suddenly become unstable in a certain area, many users will post about the problem on social media. Specifically, multiple tweets will be posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0300] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and finds that many of them contain feelings of dissatisfaction and anger.

[0301] Based on this information, the server determines that a serious signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding by conducting an on-site investigation and remote diagnosis to restore service.

[0302] The server displays the progress of the problem and the results of the response in real time on a dashboard, along with visualizing user sentiment. Administrators can use the dashboard to gain a more detailed understanding of the severity of the problem and take prompt and appropriate action.

[0303] Example prompts for generative AI models

[0304] By inputting the following prompts into the generative AI model, we can obtain a concrete explanation of the system's behavior:

[0305] Based on the following passage, please explain in detail what happens when cell phone signals become unreliable in an area:

[0306] "The server connects to the social media platform's API and sets up a real-time tweet stream using specific keywords (e.g., "cell signal interference")..."

[0307] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0308] Step 1:

[0309] The server connects to the API of the social networking platform and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"). The input is the social networking platform's API key and the monitored keyword, and the output is the collection of real-time tweet data containing the specific keyword. These tweets are stored in a temporary database on the server.

[0310] Step 2:

[0311] The server preprocesses the collected tweet data. The input is the collected tweet data, and the output is filtered, clean data. This preprocessing step includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase to lowercase). Specific operations include data cleaning using regular expressions and text formatting.

[0312] Step 3:

[0313] The server analyzes the preprocessed data using natural language processing technology. The input is preprocessed clean text data, and the output is analyzed text data. Specific operations include breaking down the text into words using a morphological analysis tool (e.g., MeCab) and analyzing its structure. This analysis extracts meaningful words and phrases contained in the tweet content.

[0314] Step 4:

[0315] The server calculates the frequency of occurrence of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) from the analyzed data. The input is the analyzed data broken down into words, and the output is the frequency of occurrence of each keyword. The specific operation in this step is to count the frequency of keywords based on the extracted word list.

[0316] Step 5:

[0317] The server uses an emotion engine to recognize user emotions from tweet text. The input is the analyzed text data, and the output is emotion information for each tweet (e.g., anger, frustration, sadness). The emotion engine uses a pre-trained model (e.g., BERT) and includes a process to determine the intensity and type of user emotion.

[0318] Step 6:

[0319] If the keyword frequency and emotion information exceed a certain threshold, the server determines that a fault has occurred in that area and generates and sends an alert. The input is keyword frequency data and emotion recognition results, and the output is generated alerts and notifications. Specific operations include the process of sending notifications to relevant parties via email or SMS when data exceeding the set threshold is detected.

[0320] Step 7:

[0321] The server visualizes detected fault information and emotion information on a dashboard in real time. The input is alert information and emotion data, and the output is visualized information displayed on the dashboard screen. Specifically, it provides information visually using maps and graphs, allowing administrators to intuitively understand the situation.

[0322] (Application example 2)

[0323] 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."

[0324] In recent years, with the spread of the Internet and social networking sites (SNS), cyber threats and attacks against businesses and individuals have been increasing. However, existing security systems have difficulty detecting and responding to threats in real time. Furthermore, there is a need to quickly identify and respond to truly significant threats from the vast amount of information available on SNS. The present invention aims to provide a system that uses real-time information obtained from SNS to quickly and effectively detect and respond to cyber threats.

[0325] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting information from SNS in real time, means for filtering and pre-processing the collected information, means for analyzing the pre-processed information using natural language processing technology, means for detecting threats based on the analysis results and generating and sending alerts, means for visualizing the detected threats on a dashboard, means for recognizing emotions from the collected information and evaluating the severity of the threats based on the emotion information, and means for notifying relevant parties of information regarding the detected threats. This makes it possible to quickly and effectively detect and respond to cyber threats by utilizing real-time information on SNS.

[0326] "SNS" means social networking service, an online service that allows users to share information and interact with each other via the Internet.

[0327] "Real-time" refers to a situation in which the latest information is collected and processed without momentary delay.

[0328] "Information collection means" refers to a mechanism or method for obtaining specific data or information, and in this invention refers to a means for extracting necessary data from SNS in real time.

[0329] "Filtering" refers to the process of removing unnecessary information from collected data and preparing it in a form suitable for analysis.

[0330] "Preprocessing" refers to the initial processing steps taken to prepare data in a certain format or condition, and includes processes such as text normalization and noise removal.

[0331] "Natural language processing technology" refers to technology for processing human language using computers, and includes text analysis and emotion recognition.

[0332] "Analysis tools" refers to methods and tools used to analyze collected and preprocessed data and extract meaning and patterns.

[0333] "Threat detection measures" refer to the means of identifying and responding to cyber attacks and dangerous behavior based on the results of analysis.

[0334] An "alert" is a warning message sent to notify relevant parties when an abnormality or threat is detected.

[0335] "Notification means" refers to mechanisms and methods for quickly communicating information about detected threats or anomalies to relevant parties.

[0336] "Dashboard" refers to an interface that visually displays an overview of the system status and data, and in the present invention refers to an interface that displays detected threat information.

[0337] "Emotion recognition technology" refers to technology for inferring a user's emotions from text, voice, etc.

[0338] The "importance assessment means" refers to a means for assessing the importance of a threat based on the recognized emotional information and determining the priority of the response.

[0339] The present invention provides a system for detecting and responding to cyber threats by collecting information from social networking sites in real time, filtering and preprocessing the information, and analyzing it using natural language processing technology. The following describes in detail an embodiment of the present invention.

[0340] System Configuration

[0341] This system mainly consists of the following components:

[0342] 1. Data collection module: A means of collecting information from social media in real time.

[0343] 2. Data Preprocessing Module: A means of filtering and preprocessing the collected information.

[0344] 3. Analysis module: A means of analyzing the preprocessed information using natural language processing techniques.

[0345] 4. Emotion Engine: A means of recognizing emotions from pre-processed information and assessing the severity of threats.

[0346] 5. Threat Detection Module: A means of detecting threats based on the analysis results and generating and sending alerts.

[0347] 6. Visualization module: A means of visualizing detected threats on a dashboard.

[0348] 7. Notification module: A means of notifying relevant parties about detected threats.

[0349] Program processing

[0350] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using specific keywords (e.g., "company name cyber attack"). Using the streaming API, it receives data each time a new tweet is posted. The received tweets are stored in a temporary database.

[0351] The server then preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[0352] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). This process analyzes the structure of the tweet and extracts meaningful words and phrases. The extracted word list is then analyzed to calculate the frequency of specific keywords (e.g., "attack," "threat," "danger," etc.). This determines whether the tweet is related to a threat.

[0353] Furthermore, the server uses an emotion engine to recognize the user's emotions (e.g., anger, frustration, sadness, etc.) from the tweet text, which provides data to evaluate the severity of the threat and the user's level of dissatisfaction.

[0354] The server determines that a threat is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold, taking into account location and other relevant information. If a threat is detected, the server generates an alert and sends a notification to relevant parties. This notification can be sent via email, SMS, or a dedicated alert management system.

[0355] Finally, the server displays the detected threat information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides visual information using maps and graphs, allowing administrators to intuitively grasp the situation.

[0356] Specific examples

[0357] For example, if a company faces an increasing threat of cyberattacks, many users will post about this issue on social media. Specifically, suppose multiple tweets are posted along the lines of, "It looks like an important company is under cyberattack right now. I wonder if they're okay?"

[0358] The server collects these tweets via the SNS API and stores them in a database. The server then extracts related keywords such as "company," "attack," and "danger" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and determines that they contain a large number of emotions such as fear and anxiety.

[0359] Based on this information, the server determines that a serious cyber threat has occurred to the company and generates an alert. A notification is sent to the threat response team, which then promptly begins responding by conducting on-site inspections and remote diagnostics to confirm the safety of the service.

[0360] The server displays the progress of threats and response results in real time on a dashboard, and also visualizes user sentiment, allowing administrators to gain a more detailed understanding of the severity of threats.

[0361] In this way, this system utilizes real-time information obtained from social media and emotion recognition technology to quickly and effectively detect and respond to cyber threats.

[0362] Examples of prompt statements

[0363] Please analyze the following text for any threats to your business:

[0364] 1. "It seems like an important company is under cyber attack. Is everything okay?"

[0365] 2. "The credibility of companies has been at risk recently."

[0366] 3. "The company network was attacked and I can't work."

[0367] As a result, please tell us which posts present a significant threat.

[0368] The above is an embodiment of the present invention.

[0369] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0370] Step 1:

[0371] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using specific keywords (e.g., "corporate cyberattack"). Using the streaming API, it receives data as new tweets are posted. The input here is the raw data provided by the API, and the output is the tweet information temporarily stored in a database.

[0372] Step 2:

[0373] The server preprocesses the collected tweet data. Preprocessing includes normalization such as deleting URLs and usernames, removing spaces and unnecessary characters, and converting uppercase to lowercase. The input is the tweet data collected in step 1, and the output is the preprocessed clean text data.

[0374] Step 3:

[0375] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the input is the preprocessed clean text data, and the output is a word list or tokenized text. The frequency of occurrence of specific keywords (e.g., "attack" and "threat") is calculated to determine threat relevance.

[0376] Step 4:

[0377] The server uses an emotion engine to recognize user emotions (e.g., anger, frustration, fear, etc.) from tweet text. The input is a word list or tokenized text, and the output is an emotion score. Emotion recognition can also use generative AI models to provide more accurate emotion information.

[0378] Step 5:

[0379] The server determines that a threat is occurring in a region if the keyword frequency and sentiment score exceed a certain threshold. The input is the keyword frequency analysis result and sentiment score, and the output is the threat detection result. Geographical information and other related information are also taken into account in the evaluation.

[0380] Step 6:

[0381] If a threat is detected, the server generates an alert and sends a notification to the relevant parties via email, SMS, a dedicated alert management system, etc. The input is the threat detection result, and the output is the notification message sent.

[0382] Step 7:

[0383] The server displays detected threat information and sentiment information on a dashboard. The dashboard is implemented as a web application and updates information in real time. The input is threat detection results and sentiment information, and the output is a visual data display that is displayed on the administrator interface. Maps and graphs allow users to intuitively grasp the situation.

[0384] In this way, each step involves data collection, pre-processing, analysis, emotion recognition, threat detection, notification, and visualization.

[0385] 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.

[0386] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0387] 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.

[0388] [Second embodiment]

[0389] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0390] 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.

[0391] 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).

[0392] 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.

[0393] 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.

[0394] 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).

[0395] 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.

[0396] 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.

[0397] 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.

[0398] 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.

[0399] 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.

[0400] 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."

[0401] The present invention is a system that collects information in real time from SNS (social networking services) and analyzes it to detect and quickly respond to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[0402] System Overview

[0403] This system mainly consists of the following components:

[0404] 1. Data collection module: Collects tweets from social media in real time.

[0405] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[0406] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[0407] 4. Fault detection module: Detects faults based on the analysis results and generates alerts.

[0408] 5. Visualization module: Displays detected faults on a dashboard.

[0409] 6. Notification module: Sends notifications to relevant parties when a fault is detected.

[0410] Program processing procedure

[0411] Data collection and preprocessing

[0412] The server collects tweets in real time from social networking sites such as Twitter. Specifically, it uses the Twitter API to retrieve tweets related to specific keywords (e.g., "cell phone signal interference") and stores them in a temporary database.

[0413] The server preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[0414] Data analysis

[0415] The server analyzes the preprocessed tweet text using natural language processing technology. Specifically, it uses a morphological analysis tool to break it down into words and calculates the frequency of specific keywords. It then determines whether the tweet contains many keywords related to outages (e.g., "can't connect," "signal," "outage," etc.).

[0416] Fault detection and notification

[0417] The server determines if a keyword frequency exceeds a certain threshold, taking into account location and other relevant information. If an outage is detected, the server generates an alert and sends a notification to relevant parties via email, SMS, or a dedicated alert management system.

[0418] Information Visualization

[0419] The server displays detected fault information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[0420] Specific examples

[0421] For example, if mobile phone signals suddenly become unstable in a certain area, multiple users will post about the problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0422] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords from the tweets, such as "Shibuya," "signal," and "can't connect." If the frequency of these keywords exceeds a certain threshold, the server determines that a signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding. On-site investigations and remote diagnostics are conducted to restore service.

[0423] In this way, this system can improve the quality of mobile phone services more effectively than conventional methods by utilizing real-time information from SNS to detect failures early and enable rapid response.

[0424] The processing flow will be explained below.

[0425] Step 1:

[0426] The server connects to the API of the social media platform and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference") and uses the streaming API to receive data each time a new tweet is posted.

[0427] Step 2:

[0428] The server stores the received tweets in a temporary database that is kept for a short period of time for subsequent processing.

[0429] Step 3:

[0430] The server retrieves tweet data collected over a fixed period of time (e.g., 1 minute) for batch processing. The data is filtered as needed, for example, to remove irrelevant tweets such as advertisements and spam.

[0431] Step 4:

[0432] The server then breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the structure of the tweet is analyzed and meaningful words and phrases are extracted.

[0433] Step 5:

[0434] The server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[0435] Step 6:

[0436] If the frequency of a keyword exceeds a pre-defined threshold, the server determines that a particular region is experiencing an outage, based on location data from tweets and user profile information.

[0437] Step 7:

[0438] The server generates an alert when a failure is detected, which includes detailed information about the failure (e.g., the location where it occurred, the time of occurrence, and related keywords).

[0439] Step 8:

[0440] The server notifies designated parties (e.g., incident response teams) of generated alerts via email, SMS, or a dedicated alert management system.

[0441] Step 9:

[0442] The device (the fault response team) receives the notification and immediately begins to act. The notification includes detailed information about the fault and the specific area that needs to be addressed.

[0443] Step 10:

[0444] The device will use on-site inspections and remote diagnostics to identify the cause of the fault and take necessary measures, such as resetting the base station or updating the software.

[0445] Step 11:

[0446] The server displays the progress of the problem and the results of the response in real time on a dashboard, allowing administrators to intuitively grasp the progress of the problem response.

[0447] Step 12:

[0448] Once the issue is resolved, the device updates the dashboard to indicate that the problem has been resolved, so that other stakeholders are aware of the situation.

[0449] Through the above steps, the system of the present invention can utilize real-time information obtained from SNS to quickly and efficiently detect and respond to mobile phone service outages.

[0450] Example 1

[0451] 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."

[0452] Conventional device monitoring systems have had difficulty in early detection of failures in mobile phone services. Furthermore, while rapid response is required after failure detection, current systems lack sufficient real-time information collection and analysis, resulting in delayed response. Therefore, the present invention aims to provide a system that collects information from SNS in real time and quickly detects and responds to failures.

[0453] 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.

[0454] In this invention, the server includes means for collecting data from SNS in real time, means for filtering and preprocessing the collected data, means for analyzing the preprocessed data using natural language processing technology, means for detecting faults based on the analysis results and generating and sending alerts, means for visualizing the detected faults on a dashboard, and means for analyzing the collected data and determining the presence or absence of faults using a generative AI model. This makes it possible to analyze the data collected from SNS in real time and detect faults quickly and accurately.

[0455] "SNS" is an abbreviation for social networking service, an online platform where users share information and opinions.

[0456] "Data collection" is the process of obtaining tweets and posts in real time from social media sites and other sources.

[0457] "Filtering" is the process of removing noise and unnecessary information from collected data and converting it into a form suitable for analysis.

[0458] "Preprocessing" refers to the process of formatting data before analyzing it, making it analyzable. Examples of preprocessing include converting uppercase letters to lowercase letters and removing advertisements.

[0459] "Natural language processing technology" is a technology that enables computers to understand, interpret, and generate human language, and includes techniques such as morphological analysis and sentiment analysis.

[0460] "Analysis" is the process of examining pre-processed data using natural language processing techniques to find specific trends and patterns.

[0461] "Fault detection" is the process of determining whether a system or service has some kind of malfunction or problem based on the analysis results.

[0462] "Alert generation" is the process of creating a warning message to communicate real-time information when a fault is detected.

[0463] "Sending" is the process of notifying relevant parties and administrators of the generated alert.

[0464] A "dashboard" is an interface for visually displaying detected fault information, and is designed to make it easy for administrators to understand the situation.

[0465] A "generative AI model" is a model that uses artificial intelligence technology to analyze data and detect specific patterns or anomalies.

[0466] "Preprocessed data" refers to data that has undergone preprocessing such as filtering and normalization and has been converted into a form suitable for analysis.

[0467] "Frequency" refers to the number of times a particular event or keyword occurs, and this number serves as the basis for determining whether or not there is a problem.

[0468] A "threshold" is a boundary value for satisfying a specific condition, and when exceeded, an alert is generated.

[0469] The present invention is a system that collects information from SNS in real time, analyzes it, and thereby detects and responds quickly to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[0470] System Configuration

[0471] This system mainly consists of the following components:

[0472] 1. Data collection module: Collects tweets from social media in real time.

[0473] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[0474] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[0475] 4. Fault detection module: Detects faults based on the analysis results and generates alerts.

[0476] 5. Visualization module: Displays detected faults on a dashboard.

[0477] 6. Notification module: Sends notifications to relevant parties when a fault is detected.

[0478] 7. Generative AI model: Analyzes collected data and determines whether a fault exists.

[0479] Hardware and software used

[0480] The server uses Twitter API and other SNS APIs to collect data from SNS in real time. The collected data is temporarily stored in a database. For data preprocessing, Python libraries (e.g., Pandas, Numpy) are used to remove noise and normalize text.

[0481] For data analysis, we use natural language processing techniques such as morphological analysis tools (e.g., MeCab) and generative AI models (e.g., TensorFlow and PyTorch). Based on the analysis results, we detect failures and use the SendGrid API and other email sending services to generate alerts and send notifications.

[0482] Dashboard visualization uses a web application framework (e.g., Flask or Django) and a front-end visualization library (e.g., D3.js or Chart.js).

[0483] Specific examples

[0484] For example, if mobile phone signals suddenly become unstable in a certain area, multiple users will post about the problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0485] The server collects these tweets via the SNS API and stores them in a database. The server then extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets. If the frequency of these keywords exceeds a certain threshold, the server determines that a signal interference is occurring in the Shibuya area and generates an alert.

[0486] The generated alert is notified to the fault response team using the SendGrid API, and the device (trouble response team) immediately begins responding, conducting on-site investigations and remote diagnosis to restore service.

[0487] Examples of prompt statements

[0488] The following is a specific example of a prompt sentence to be input to the generative AI model:

[0489] "Analyze tweets containing keywords such as 'signal wave,' 'interference,' 'cannot connect,' and 'unstable' in the specified area and time period, and determine whether the frequency of the relevant keywords exceeds a certain threshold. Example: Tweets in Shibuya from 12:00 to 14:00 on October 12, 2023."

[0490] By using this prompt, data collected from social media can be analyzed by a generative AI model to quickly and accurately determine whether or not a disability exists.

[0491] As described above, this system analyzes data collected from SNS in real time, enabling rapid fault detection and response, thereby improving the quality of mobile phone services more effectively than conventional methods.

[0492] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0493] Step 1:

[0494] Data collection

[0495] The server collects tweets in real time from social networking sites such as Twitter. Specifically, it uses the Twitter API to retrieve tweets related to specific keywords (e.g., "cell phone signal interference"). The Twitter API key and search query are given as input, and the raw data of the retrieved tweets is stored in a database as output. Specifically, it periodically sends API requests to retrieve new tweets.

[0496] Step 2:

[0497] Data Preprocessing

[0498] The server pre-processes the collected data. As input, it takes raw data and as output, it gets pre-processed, clean data. The pre-processing includes the following specific operations:

[0499] Noise reduction: Utilize blacklists to remove advertising and spam tweets.

[0500] Text normalization: convert uppercase to lowercase and remove punctuation marks.

[0501] Language filtering: Filter out tweets in languages ​​other than Japanese.

[0502] Step 3:

[0503] Data analysis

[0504] The server analyzes the preprocessed data using natural language processing technology. The preprocessed clean data is given as input, and the analyzed data is obtained as output. Specifically, it breaks down the data into words using a morphological analysis tool (e.g., MeCab), extracts relevant keywords (e.g., "can't connect," "signal," "interference"), and calculates their frequency.

[0505] Step 4:

[0506] Fault Detection

[0507] The server detects a failure based on the analysis results. The analyzed data and a set threshold are given as input, and a flag indicating whether a failure has occurred and alert information are obtained as output. Specifically, if the frequency of a specific keyword exceeds the threshold, it determines that a failure has occurred in that area and generates an alert. Location information is also analyzed to identify the affected area.

[0508] Step 5:

[0509] Alert Generation

[0510] The server generates an alert when a failure is detected. The input is a failure detection flag and its detailed information (e.g., the area where the failure occurred, the time of occurrence), and the output is an alert message. Specifically, the server creates an alert message containing detailed information such as the type of failure, the area where the failure occurred, and the time of occurrence, and saves it in the database.

[0511] Step 6:

[0512] Send notifications

[0513] The server notifies the relevant parties of the generated alert. The alert message is given as input, and the sending result is obtained as output. Specifically, the server sends the alert using an API for sending email or SMS (e.g., SendGrid API), and sends the information to a dedicated alert management system.

[0514] Step 7:

[0515] Information Visualization

[0516] The server displays the detected outage information on a dashboard. The outage information is given as input, and the updated dashboard is obtained as output. Specifically, the outage information is displayed in real time using a web application framework (e.g., Flask or Django) and visualized using maps and graphs.

[0517] (Application example 1)

[0518] 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."

[0519] Conventional fault detection and ad delivery systems have difficulty collecting and analyzing information in real time, making it difficult to deliver ads effectively or quickly resolve faults. Furthermore, it is difficult to select optimal ads based on consumer opinions and emotions depending on the region and time of day, limiting the effectiveness of advertising.

[0520] 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.

[0521] In this invention, the server includes means for collecting information from SNS in real time, means for filtering and preprocessing the collected information, means for analyzing the preprocessed information using natural language processing technology, means for calculating region and emotion scores based on the analysis results and selecting advertisements, and means for sending the selected advertisements to users by push notification. This makes it possible to analyze consumer opinions and emotions in real time and quickly deliver optimal advertisements based on them.

[0522] "SNS" refers to social networking services, which are online platforms that allow users to interact with each other over the Internet.

[0523] "Real time" refers to processing that occurs simultaneously with actual time.

[0524] "Data filtering" is the process of selecting only necessary information from collected information.

[0525] "Preprocessing" refers to a series of steps that prepare information for analysis, including noise removal and text normalization.

[0526] "Natural language processing technology" refers to the technology that enables computers to understand and analyze human language, and is capable of performing a variety of language processing.

[0527] An "emotion score" is a numerical representation of the intensity of emotions contained in text data, classified as positive, negative, or neutral.

[0528] "Ad selection" refers to the process of selecting the most suitable advertisement based on the analysis results.

[0529] "Push notifications" are the ability for an application to send messages to a user's device in real time.

[0530] "User" refers to the general consumer who uses this system.

[0531] MODE FOR CARRYING OUT THE INVENTION

[0532] This invention is a system that collects and analyzes real-time information from social media and sends optimal advertisements to users via push notifications based on consumer opinions and emotions. The specific configuration and operation of this system are described below.

[0533] System Configuration

[0534] The system includes the following components:

[0535] 1. Data collection module: Collects information from social media in real time.

[0536] 2. Data preprocessing module: Filters and preprocesses the collected information.

[0537] 3. Analysis module: Analyzes the preprocessed information using natural language processing techniques and calculates the sentiment score.

[0538] 4. Advertisement selection module: Selects the most suitable advertisement based on the analysis results.

[0539] 5. Notification module: Send selected ads to users via push notifications.

[0540] Hardware and software used

[0541] Hardware: Servers, smartphones

[0542] software:

[0543] Twitter API, other SNS APIs

[0544] Database (SQLite)

[0545] Natural language processing libraries (MeCab, NLTK, BERT)

[0546] Push notification service (Firebase Cloud Messaging)

[0547] Operation process

[0548] 1. Data Collection and Preprocessing

[0549] The server collects tweets related to specific keywords (e.g., "new product," "sale," "recommended," etc.) from social media (e.g., Twitter) in real time. This is done using social media APIs such as the Twitter API. The collected tweet data is stored in a temporary database on the server.

[0550] The server then preprocesses the collected tweet data, which includes noise removal (e.g., removing spam messages) and text normalization (e.g., converting uppercase to lowercase).

[0551] 2. Data Analysis

[0552] The preprocessed tweet text is analyzed on the server using natural language processing techniques. Specifically, a morphological analysis tool (e.g., MeCab) is used to break it down into words and calculate the frequency of specific keywords. A sentiment analysis model (e.g., NLTK or BERT) is also used to calculate positive, negative, and neutral sentiment scores.

[0553] 3. Ad selection and delivery

[0554] Based on the analysis results, the ad selection module selects the most suitable ad, based on the sentiment score and location information, and sends it to the user via smartphone push notification.

[0555] Specific examples

[0556] For example, suppose that users in Shibuya post an increasing number of tweets about a new smartphone on social media, many of which are positive. The collected tweet data contains keywords such as "Shibuya," "smartphone," and "great." If the analysis confirms that the tweets have a high positive sentiment score, a promotional ad for the new smartphone will be sent to the user's smartphone during that time period.

[0557] Prompt Sentence Examples

[0558] "Collect the latest positive tweets about new smartphones in Shibuya, calculate the sentiment score, and select ads."

[0559] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0560] Step 1:

[0561] The server collects tweets related to specified keywords from social media (e.g., Twitter) in real time. Here, the Twitter API is used to obtain tweets that match keywords such as "new product" or "sale." These collected tweets are the input data, and the output is that they are stored in a temporary database on the server.

[0562] Step 2:

[0563] The server preprocesses the collected tweet data. This preprocessing includes removing advertisements and spam messages and normalizing the text (e.g., converting uppercase to lowercase). This process outputs clean text data suitable for analysis. This process filters out noise data, resulting in a clear database.

[0564] Step 3:

[0565] The server analyzes the preprocessed tweet text using natural language processing techniques (e.g., MeCab, NLTK, BERT). It performs morphological analysis, breaks it down to the word level, and calculates the frequency of keywords. At the same time, it uses a sentiment analysis model to calculate a positive, negative, or neutral sentiment score for each tweet. The output of this analysis process is a list of keyword frequencies and a list of sentiment scores.

[0566] Step 4:

[0567] The server selects an advertisement based on the analysis results from step 3. Using the sentiment score and keyword frequency obtained from the analysis results, it selects the most suitable advertisement (e.g., new product promotion or sale information). This selected advertisement is the output data.

[0568] Step 5:

[0569] The server sends the selected advertisement to the smartphone user via push notification. This process uses a push notification service such as Firebase Cloud Messaging to deliver the advertisement content to the user's smartphone. The user's receiving device acts as the push notification receiver, and the advertisement message is output.

[0570] Step 6:

[0571] The user can then view the push notification and respond to the ad if necessary. This final step involves the user's actions becoming the final form of input and output, as well as the collection of data to measure the effectiveness of the ad.

[0572] 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.

[0573] The present invention is a system that collects information from SNS (social networking services) in real time, analyzes the information, and recognizes emotions, thereby detecting and quickly responding to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[0574] System Overview

[0575] This system mainly consists of the following components:

[0576] 1. Data collection module: Collects tweets from social media in real time.

[0577] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[0578] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[0579] 4. Emotion Engine: Recognize user emotions from preprocessed tweet text.

[0580] 5. Fault detection module: Detects faults based on the analysis results and information from the emotion engine and generates alerts.

[0581] 6. Visualization module: Displays detected faults on a dashboard.

[0582] 7. Notification module: Sends notifications to relevant parties when a fault is detected.

[0583] Program processing procedure

[0584] Data collection and preprocessing

[0585] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"). Using the streaming API, it receives data as new tweets are posted. The received tweets are stored in a temporary database.

[0586] The server preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[0587] Data analysis and emotion recognition

[0588] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, it analyzes the structure of the tweet and extracts meaningful words and phrases.

[0589] Next, the server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[0590] Furthermore, the server uses an emotion engine to recognize the user's emotions (e.g., anger, frustration, sadness, etc.) from the tweet text, which provides data to evaluate the severity of the problem and the user's dissatisfaction level.

[0591] Fault detection and notification

[0592] The server determines that an outage is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold, taking into account location and other relevant information. If an outage is detected, the server generates an alert and sends a notification to relevant parties. This notification is sent via email, SMS, or a dedicated alert management system.

[0593] Information Visualization

[0594] The server displays detected fault information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[0595] Specific examples

[0596] For example, if mobile phone signals suddenly become unstable in a certain area, many users will post about this problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0597] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and finds that many of them contain feelings of dissatisfaction and anger.

[0598] Based on this information, the server determines that a serious signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding by conducting an on-site investigation and remote diagnosis to restore service.

[0599] The server displays the progress of the incident and the results of the response in real time on a dashboard, and also visualizes user emotional information, allowing administrators to understand the severity of the incident in more detail.

[0600] In this way, this system can utilize real-time information obtained from social media and emotion recognition technology to quickly and effectively detect and respond to mobile phone service outages.

[0601] The processing flow will be explained below.

[0602] Step 1:

[0603] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"), and uses the streaming API to receive data each time a new tweet is posted.

[0604] Step 2:

[0605] The server stores the received tweets in a temporary database that is kept for a short period of time for subsequent processing.

[0606] Step 3:

[0607] The server retrieves tweet data collected over a fixed time period (e.g., 1 minute) for batch processing. If necessary, the data is gallery-filtered to remove irrelevant tweets such as advertisements and spam.

[0608] Step 4:

[0609] The server then breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the structure of the tweet is analyzed and meaningful words and phrases are extracted.

[0610] Step 5:

[0611] The server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[0612] Step 6:

[0613] The server recognizes the user's emotions from the preprocessed tweet text using an emotion engine, which analyzes the text using natural language processing techniques to identify emotions such as anger, frustration, and sadness.

[0614] Step 7:

[0615] The server determines that an outage is occurring in a region when the frequency of keywords and user sentiment exceed a certain threshold. The region information is determined based on tweet location data and user profile information.

[0616] Step 8:

[0617] The server generates an alert when a failure is detected, which includes detailed information about the failure (e.g., the area where the failure occurred, the time of occurrence, and related keywords) and the user's emotional state.

[0618] Step 9:

[0619] The server notifies designated parties (e.g., incident response teams) of generated alerts via email, SMS, or a dedicated alert management system.

[0620] Step 10:

[0621] The device (the fault response team) receives the notification and immediately begins to act. The notification includes detailed information about the fault and the specific area that needs to be addressed.

[0622] Step 11:

[0623] The device will use on-site inspections and remote diagnostics to identify the cause of the fault and take necessary measures, such as resetting the base station or updating the software.

[0624] Step 12:

[0625] The server displays the progress and results of the problem in real time on a dashboard, which also visualizes user sentiment, allowing administrators to better understand the severity of the problem.

[0626] Step 13:

[0627] Once the issue is resolved, the device updates the dashboard to indicate that the problem has been resolved, so that other stakeholders are aware of the situation.

[0628] Through the above steps, the system of the present invention can utilize real-time information obtained from SNS and emotion recognition technology to quickly and effectively detect and respond to mobile phone service outages.

[0629] Example 2

[0630] 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."

[0631] Conventional fault detection systems for mobile phone services have had problems with detecting faults in real time and being unable to assess the severity of faults based on user emotional information. As a result, early detection and rapid response to faults have not been fully realized, leading to growing user dissatisfaction.

[0632] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for collecting data from SNS in real time, means for filtering and preprocessing the collected data, means for analyzing the preprocessed data using natural language processing technology, means for calculating the frequency of appearance of keywords from the analyzed data, means for recognizing emotions from the analyzed data, means for detecting faults based on the frequency of keywords and emotional information, and means for generating and transmitting alerts, and means for visualizing the detected faults on a dashboard. This enables real-time fault detection and assessment of the severity of faults based on user emotions, thereby realizing a prompt and effective response.

[0633] "SNS" refers to social networking services, which are online platforms where users can post information and share it with other users in real time.

[0634] "Real-time" refers to a state in which data and information are processed and provided immediately without delay.

[0635] "Data collection" refers to the process of gathering specific information or data according to specified criteria.

[0636] "Filtering" refers to the process of removing unnecessary data and noise and selecting data that is suitable for the purpose.

[0637] "Preprocessing" refers to a series of steps that transform data into a form suitable for analysis and processing.

[0638] "Natural language processing technology" refers to the technology that allows computers to understand and analyze human language.

[0639] "Analysis" refers to the process of breaking down given data or information into detail and understanding its meaning and structure.

[0640] "Keyword frequency" refers to the number of times a particular word or phrase appears in text data.

[0641] "Emotion recognition" refers to determining a user's emotional state from data such as text and voice.

[0642] "Fault detection" refers to the process of identifying and identifying anomalies or problems in a system or service.

[0643] An "alert" is a notification sent to warn you when an abnormality or emergency occurs.

[0644] A "dashboard" is an interface that displays information and data visually and intuitively, allowing administrators to understand the situation.

[0645] The present invention is a system that collects information from SNS (social networking services) in real time, analyzes the information, and recognizes emotions in the information, thereby detecting and quickly responding to problems in mobile phone services. The following describes in detail an embodiment of the present invention.

[0646] Data collection

[0647] The server uses a data collection API of a social networking platform, such as the Twitter API, to collect real-time tweets filtered by specific keywords (e.g., "cell phone signal interference"). This real-time data collection is achieved by a program that implements Twitter's streaming API. The data is stored in a temporary database on the server.

[0648] Data Preprocessing

[0649] The server preprocesses the collected tweet data, including noise removal (removing advertisements and spam) and text normalization (converting uppercase letters to lowercase), converting the data into a form suitable for analysis.

[0650] Data analysis

[0651] The server breaks down the preprocessed text data into words using a morphological analysis tool (e.g., MeCab). Morphological analysis analyzes the structure of tweets and extracts meaningful words and phrases. The server then analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.).

[0652] emotion recognition

[0653] The server uses an emotion engine to recognize the user's emotion (e.g., anger, frustration, sadness, etc.) from the tweet text. The emotion engine uses a pre-trained model (e.g., BERT). This analysis determines the intensity and type of the user's emotion, and evaluates the severity of the problem and the user's level of dissatisfaction.

[0654] Fault Detection and Notification

[0655] The server determines that an outage is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold. Location and other relevant information are also taken into account. When an outage is detected, the server generates an alert and sends a notification to relevant parties. Notifications are sent via email, SMS, or a dedicated alert management system.

[0656] Information Visualization

[0657] The server displays detected fault information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[0658] Specific examples

[0659] For example, if mobile phone signals suddenly become unstable in a certain area, many users will post about the problem on social media. Specifically, multiple tweets will be posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0660] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and finds that many of them contain feelings of dissatisfaction and anger.

[0661] Based on this information, the server determines that a serious signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding by conducting an on-site investigation and remote diagnosis to restore service.

[0662] The server displays the progress of the problem and the results of the response in real time on a dashboard, along with visualizing user sentiment. Administrators can use the dashboard to gain a more detailed understanding of the severity of the problem and take prompt and appropriate action.

[0663] Example prompts for generative AI models

[0664] By inputting the following prompts into the generative AI model, we can obtain a concrete explanation of the system's behavior:

[0665] Based on the following passage, please explain in detail what happens when cell phone signals become unreliable in an area:

[0666] "The server connects to the social media platform's API and sets up a real-time tweet stream using specific keywords (e.g., "cell signal interference")..."

[0667] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0668] Step 1:

[0669] The server connects to the API of the social networking platform and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"). The input is the social networking platform's API key and the monitored keyword, and the output is the collection of real-time tweet data containing the specific keyword. These tweets are stored in a temporary database on the server.

[0670] Step 2:

[0671] The server preprocesses the collected tweet data. The input is the collected tweet data, and the output is filtered, clean data. This preprocessing step includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase to lowercase). Specific operations include data cleaning using regular expressions and text formatting.

[0672] Step 3:

[0673] The server analyzes the preprocessed data using natural language processing technology. The input is preprocessed clean text data, and the output is analyzed text data. Specific operations include breaking down the text into words using a morphological analysis tool (e.g., MeCab) and analyzing its structure. This analysis extracts meaningful words and phrases contained in the tweet content.

[0674] Step 4:

[0675] The server calculates the frequency of occurrence of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) from the analyzed data. The input is the analyzed data broken down into words, and the output is the frequency of occurrence of each keyword. The specific operation in this step is to count the frequency of keywords based on the extracted word list.

[0676] Step 5:

[0677] The server uses an emotion engine to recognize user emotions from tweet text. The input is the analyzed text data, and the output is emotion information for each tweet (e.g., anger, frustration, sadness). The emotion engine uses a pre-trained model (e.g., BERT) and includes a process to determine the intensity and type of user emotion.

[0678] Step 6:

[0679] If the keyword frequency and emotion information exceed a certain threshold, the server determines that a fault has occurred in that area and generates and sends an alert. The input is keyword frequency data and emotion recognition results, and the output is generated alerts and notifications. Specific operations include the process of sending notifications to relevant parties via email or SMS when data exceeding the set threshold is detected.

[0680] Step 7:

[0681] The server visualizes detected fault information and emotion information on a dashboard in real time. The input is alert information and emotion data, and the output is visualized information displayed on the dashboard screen. Specifically, it provides information visually using maps and graphs, allowing administrators to intuitively understand the situation.

[0682] (Application example 2)

[0683] 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."

[0684] In recent years, with the spread of the Internet and social networking sites (SNS), cyber threats and attacks against businesses and individuals have been increasing. However, existing security systems have difficulty detecting and responding to threats in real time. Furthermore, there is a need to quickly identify and respond to truly significant threats from the vast amount of information available on SNS. The present invention aims to provide a system that uses real-time information obtained from SNS to quickly and effectively detect and respond to cyber threats.

[0685] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting information from SNS in real time, means for filtering and pre-processing the collected information, means for analyzing the pre-processed information using natural language processing technology, means for detecting threats based on the analysis results and generating and sending alerts, means for visualizing the detected threats on a dashboard, means for recognizing emotions from the collected information and evaluating the severity of the threats based on the emotion information, and means for notifying relevant parties of information regarding the detected threats. This makes it possible to quickly and effectively detect and respond to cyber threats by utilizing real-time information on SNS.

[0686] "SNS" means social networking service, an online service that allows users to share information and interact with each other via the Internet.

[0687] "Real-time" refers to a situation in which the latest information is collected and processed without momentary delay.

[0688] "Information collection means" refers to a mechanism or method for obtaining specific data or information, and in this invention refers to a means for extracting necessary data from SNS in real time.

[0689] "Filtering" refers to the process of removing unnecessary information from collected data and preparing it in a form suitable for analysis.

[0690] "Preprocessing" refers to the initial processing steps taken to prepare data in a certain format or condition, and includes processes such as text normalization and noise removal.

[0691] "Natural language processing technology" refers to technology for processing human language using computers, and includes text analysis and emotion recognition.

[0692] "Analysis tools" refers to methods and tools used to analyze collected and preprocessed data and extract meaning and patterns.

[0693] "Threat detection measures" refer to the means of identifying and responding to cyber attacks and dangerous behavior based on the results of analysis.

[0694] An "alert" is a warning message sent to notify relevant parties when an abnormality or threat is detected.

[0695] "Notification means" refers to mechanisms and methods for quickly communicating information about detected threats or anomalies to relevant parties.

[0696] "Dashboard" refers to an interface that visually displays an overview of the system status and data, and in the present invention refers to an interface that displays detected threat information.

[0697] "Emotion recognition technology" refers to technology for inferring a user's emotions from text, voice, etc.

[0698] The "importance assessment means" refers to a means for assessing the importance of a threat based on the recognized emotional information and determining the priority of the response.

[0699] The present invention provides a system for detecting and responding to cyber threats by collecting information from social networking sites in real time, filtering and preprocessing the information, and analyzing it using natural language processing technology. The following describes in detail an embodiment of the present invention.

[0700] System Configuration

[0701] This system mainly consists of the following components:

[0702] 1. Data collection module: A means of collecting information from social media in real time.

[0703] 2. Data Preprocessing Module: A means of filtering and preprocessing the collected information.

[0704] 3. Analysis module: A means of analyzing the preprocessed information using natural language processing techniques.

[0705] 4. Emotion Engine: A means of recognizing emotions from pre-processed information and assessing the severity of threats.

[0706] 5. Threat Detection Module: A means of detecting threats based on the analysis results and generating and sending alerts.

[0707] 6. Visualization module: A means of visualizing detected threats on a dashboard.

[0708] 7. Notification module: A means of notifying relevant parties about detected threats.

[0709] Program processing

[0710] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using specific keywords (e.g., "company name cyber attack"). Using the streaming API, it receives data each time a new tweet is posted. The received tweets are stored in a temporary database.

[0711] The server then preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[0712] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). This process analyzes the structure of the tweet and extracts meaningful words and phrases. The extracted word list is then analyzed to calculate the frequency of specific keywords (e.g., "attack," "threat," "danger," etc.). This determines whether the tweet is related to a threat.

[0713] Furthermore, the server uses an emotion engine to recognize the user's emotions (e.g., anger, frustration, sadness, etc.) from the tweet text, which provides data to evaluate the severity of the threat and the user's level of dissatisfaction.

[0714] The server determines that a threat is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold, taking into account location and other relevant information. If a threat is detected, the server generates an alert and sends a notification to relevant parties. This notification can be sent via email, SMS, or a dedicated alert management system.

[0715] Finally, the server displays the detected threat information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides visual information using maps and graphs, allowing administrators to intuitively grasp the situation.

[0716] Specific examples

[0717] For example, if a company faces an increasing threat of cyberattacks, many users will post about this issue on social media. Specifically, suppose multiple tweets are posted along the lines of, "It looks like an important company is under cyberattack right now. I wonder if they're okay?"

[0718] The server collects these tweets via the SNS API and stores them in a database. The server then extracts related keywords such as "company," "attack," and "danger" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and determines that they contain a large number of emotions such as fear and anxiety.

[0719] Based on this information, the server determines that a serious cyber threat has occurred to the company and generates an alert. A notification is sent to the threat response team, which then promptly begins responding by conducting on-site inspections and remote diagnostics to confirm the safety of the service.

[0720] The server displays the progress of threats and response results in real time on a dashboard, and also visualizes user sentiment, allowing administrators to gain a more detailed understanding of the severity of threats.

[0721] In this way, this system utilizes real-time information obtained from social media and emotion recognition technology to quickly and effectively detect and respond to cyber threats.

[0722] Examples of prompt statements

[0723] Please analyze the following text for any threats to your business:

[0724] 1. "It seems like an important company is under cyber attack. Is everything okay?"

[0725] 2. "The credibility of companies has been at risk recently."

[0726] 3. "The company network was attacked and I can't work."

[0727] As a result, please tell us which posts present a significant threat.

[0728] The above is an embodiment of the present invention.

[0729] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0730] Step 1:

[0731] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using specific keywords (e.g., "corporate cyberattack"). Using the streaming API, it receives data as new tweets are posted. The input here is the raw data provided by the API, and the output is the tweet information temporarily stored in a database.

[0732] Step 2:

[0733] The server preprocesses the collected tweet data. Preprocessing includes normalization such as deleting URLs and usernames, removing spaces and unnecessary characters, and converting uppercase to lowercase. The input is the tweet data collected in step 1, and the output is the preprocessed clean text data.

[0734] Step 3:

[0735] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the input is the preprocessed clean text data, and the output is a word list or tokenized text. The frequency of occurrence of specific keywords (e.g., "attack" and "threat") is calculated to determine threat relevance.

[0736] Step 4:

[0737] The server uses an emotion engine to recognize user emotions (e.g., anger, frustration, fear, etc.) from tweet text. The input is a word list or tokenized text, and the output is an emotion score. Emotion recognition can also use generative AI models to provide more accurate emotion information.

[0738] Step 5:

[0739] The server determines that a threat is occurring in a region if the keyword frequency and sentiment score exceed a certain threshold. The input is the keyword frequency analysis result and sentiment score, and the output is the threat detection result. Geographical information and other related information are also taken into account in the evaluation.

[0740] Step 6:

[0741] If a threat is detected, the server generates an alert and sends a notification to the relevant parties via email, SMS, a dedicated alert management system, etc. The input is the threat detection result, and the output is the notification message sent.

[0742] Step 7:

[0743] The server displays detected threat information and sentiment information on a dashboard. The dashboard is implemented as a web application and updates information in real time. The input is threat detection results and sentiment information, and the output is a visual data display that is displayed on the administrator interface. Maps and graphs allow users to intuitively grasp the situation.

[0744] In this way, each step involves data collection, pre-processing, analysis, emotion recognition, threat detection, notification, and visualization.

[0745] 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.

[0746] 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.

[0747] 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.

[0748] [Third embodiment]

[0749] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0750] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0751] 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).

[0752] 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.

[0753] 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.

[0754] 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).

[0755] 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.

[0756] 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.

[0757] 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.

[0758] 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.

[0759] 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.

[0760] 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."

[0761] The present invention is a system that collects information in real time from SNS (social networking services) and analyzes it to detect and quickly respond to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[0762] System Overview

[0763] This system mainly consists of the following components:

[0764] 1. Data collection module: Collects tweets from social media in real time.

[0765] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[0766] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[0767] 4. Fault detection module: Detects faults based on the analysis results and generates alerts.

[0768] 5. Visualization module: Displays detected faults on a dashboard.

[0769] 6. Notification module: Sends notifications to relevant parties when a fault is detected.

[0770] Program processing procedure

[0771] Data collection and preprocessing

[0772] The server collects tweets in real time from social networking sites such as Twitter. Specifically, it uses the Twitter API to retrieve tweets related to specific keywords (e.g., "cell phone signal interference") and stores them in a temporary database.

[0773] The server preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[0774] Data analysis

[0775] The server analyzes the preprocessed tweet text using natural language processing technology. Specifically, it uses a morphological analysis tool to break it down into words and calculates the frequency of specific keywords. It then determines whether the tweet contains many keywords related to outages (e.g., "can't connect," "signal," "outage," etc.).

[0776] Fault detection and notification

[0777] The server determines if a keyword frequency exceeds a certain threshold, taking into account location and other relevant information. If an outage is detected, the server generates an alert and sends a notification to relevant parties via email, SMS, or a dedicated alert management system.

[0778] Information Visualization

[0779] The server displays detected fault information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[0780] Specific examples

[0781] For example, if mobile phone signals suddenly become unstable in a certain area, multiple users will post about the problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0782] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords from the tweets, such as "Shibuya," "signal," and "can't connect." If the frequency of these keywords exceeds a certain threshold, the server determines that a signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding. On-site investigations and remote diagnostics are conducted to restore service.

[0783] In this way, this system can improve the quality of mobile phone services more effectively than conventional methods by utilizing real-time information from SNS to detect failures early and enable rapid response.

[0784] The processing flow will be explained below.

[0785] Step 1:

[0786] The server connects to the API of the social media platform and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference") and uses the streaming API to receive data each time a new tweet is posted.

[0787] Step 2:

[0788] The server stores the received tweets in a temporary database that is kept for a short period of time for subsequent processing.

[0789] Step 3:

[0790] The server retrieves tweet data collected over a fixed period of time (e.g., 1 minute) for batch processing. The data is filtered as needed, for example, to remove irrelevant tweets such as advertisements and spam.

[0791] Step 4:

[0792] The server then breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the structure of the tweet is analyzed and meaningful words and phrases are extracted.

[0793] Step 5:

[0794] The server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[0795] Step 6:

[0796] If the frequency of a keyword exceeds a pre-defined threshold, the server determines that a particular region is experiencing an outage, based on location data from tweets and user profile information.

[0797] Step 7:

[0798] The server generates an alert when a failure is detected, which includes detailed information about the failure (e.g., the location where it occurred, the time of occurrence, and related keywords).

[0799] Step 8:

[0800] The server notifies designated parties (e.g., incident response teams) of generated alerts via email, SMS, or a dedicated alert management system.

[0801] Step 9:

[0802] The device (the fault response team) receives the notification and immediately begins to act. The notification includes detailed information about the fault and the specific area that needs to be addressed.

[0803] Step 10:

[0804] The device will use on-site inspections and remote diagnostics to identify the cause of the fault and take necessary measures, such as resetting the base station or updating the software.

[0805] Step 11:

[0806] The server displays the progress of the problem and the results of the response in real time on a dashboard, allowing administrators to intuitively grasp the progress of the problem response.

[0807] Step 12:

[0808] Once the issue is resolved, the device updates the dashboard to indicate that the problem has been resolved, so that other stakeholders are aware of the situation.

[0809] Through the above steps, the system of the present invention can utilize real-time information obtained from SNS to quickly and efficiently detect and respond to mobile phone service outages.

[0810] Example 1

[0811] 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."

[0812] Conventional device monitoring systems have had difficulty in early detection of failures in mobile phone services. Furthermore, while rapid response is required after failure detection, current systems lack sufficient real-time information collection and analysis, resulting in delayed response. Therefore, the present invention aims to provide a system that collects information from SNS in real time and quickly detects and responds to failures.

[0813] 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.

[0814] In this invention, the server includes means for collecting data from SNS in real time, means for filtering and preprocessing the collected data, means for analyzing the preprocessed data using natural language processing technology, means for detecting faults based on the analysis results and generating and sending alerts, means for visualizing the detected faults on a dashboard, and means for analyzing the collected data and determining the presence or absence of faults using a generative AI model. This makes it possible to analyze the data collected from SNS in real time and detect faults quickly and accurately.

[0815] "SNS" is an abbreviation for social networking service, an online platform where users share information and opinions.

[0816] "Data collection" is the process of obtaining tweets and posts in real time from social media sites and other sources.

[0817] "Filtering" is the process of removing noise and unnecessary information from collected data and converting it into a form suitable for analysis.

[0818] "Preprocessing" refers to the process of formatting data before analyzing it, making it analyzable. Examples of preprocessing include converting uppercase letters to lowercase letters and removing advertisements.

[0819] "Natural language processing technology" is a technology that enables computers to understand, interpret, and generate human language, and includes techniques such as morphological analysis and sentiment analysis.

[0820] "Analysis" is the process of examining pre-processed data using natural language processing techniques to find specific trends and patterns.

[0821] "Fault detection" is the process of determining whether a system or service has some kind of malfunction or problem based on the analysis results.

[0822] "Alert generation" is the process of creating a warning message to communicate real-time information when a fault is detected.

[0823] "Sending" is the process of notifying relevant parties and administrators of the generated alert.

[0824] A "dashboard" is an interface for visually displaying detected fault information, and is designed to make it easy for administrators to understand the situation.

[0825] A "generative AI model" is a model that uses artificial intelligence technology to analyze data and detect specific patterns or anomalies.

[0826] "Preprocessed data" refers to data that has undergone preprocessing such as filtering and normalization and has been converted into a form suitable for analysis.

[0827] "Frequency" refers to the number of times a particular event or keyword occurs, and this number serves as the basis for determining whether or not there is a problem.

[0828] A "threshold" is a boundary value for satisfying a specific condition, and when exceeded, an alert is generated.

[0829] The present invention is a system that collects information from SNS in real time, analyzes it, and thereby detects and responds quickly to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[0830] System Configuration

[0831] This system mainly consists of the following components:

[0832] 1. Data collection module: Collects tweets from social media in real time.

[0833] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[0834] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[0835] 4. Fault detection module: Detects faults based on the analysis results and generates alerts.

[0836] 5. Visualization module: Displays detected faults on a dashboard.

[0837] 6. Notification module: Sends notifications to relevant parties when a fault is detected.

[0838] 7. Generative AI model: Analyzes collected data and determines whether a fault exists.

[0839] Hardware and software used

[0840] The server uses Twitter API and other SNS APIs to collect data from SNS in real time. The collected data is temporarily stored in a database. For data preprocessing, Python libraries (e.g., Pandas, Numpy) are used to remove noise and normalize text.

[0841] For data analysis, we use natural language processing techniques such as morphological analysis tools (e.g., MeCab) and generative AI models (e.g., TensorFlow and PyTorch). Based on the analysis results, we detect failures and use the SendGrid API and other email sending services to generate alerts and send notifications.

[0842] Dashboard visualization uses a web application framework (e.g., Flask or Django) and a front-end visualization library (e.g., D3.js or Chart.js).

[0843] Specific examples

[0844] For example, if mobile phone signals suddenly become unstable in a certain area, multiple users will post about the problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0845] The server collects these tweets via the SNS API and stores them in a database. The server then extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets. If the frequency of these keywords exceeds a certain threshold, the server determines that a signal interference is occurring in the Shibuya area and generates an alert.

[0846] The generated alert is notified to the fault response team using the SendGrid API, and the device (trouble response team) immediately begins responding, conducting on-site investigations and remote diagnosis to restore service.

[0847] Examples of prompt statements

[0848] The following is a specific example of a prompt sentence to be input to the generative AI model:

[0849] "Analyze tweets containing keywords such as 'signal wave,' 'interference,' 'cannot connect,' and 'unstable' in the specified area and time period, and determine whether the frequency of the relevant keywords exceeds a certain threshold. Example: Tweets in Shibuya from 12:00 to 14:00 on October 12, 2023."

[0850] By using this prompt, data collected from social media can be analyzed by a generative AI model to quickly and accurately determine whether or not a disability exists.

[0851] As described above, this system analyzes data collected from SNS in real time, enabling rapid fault detection and response, thereby improving the quality of mobile phone services more effectively than conventional methods.

[0852] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0853] Step 1:

[0854] Data collection

[0855] The server collects tweets in real time from social networking sites such as Twitter. Specifically, it uses the Twitter API to retrieve tweets related to specific keywords (e.g., "cell phone signal interference"). The Twitter API key and search query are given as input, and the raw data of the retrieved tweets is stored in a database as output. Specifically, it periodically sends API requests to retrieve new tweets.

[0856] Step 2:

[0857] Data Preprocessing

[0858] The server pre-processes the collected data. As input, it takes raw data and as output, it gets pre-processed, clean data. The pre-processing includes the following specific operations:

[0859] Noise reduction: Utilize blacklists to remove advertising and spam tweets.

[0860] Text normalization: convert uppercase to lowercase and remove punctuation marks.

[0861] Language filtering: Filter out tweets in languages ​​other than Japanese.

[0862] Step 3:

[0863] Data analysis

[0864] The server analyzes the preprocessed data using natural language processing technology. The preprocessed clean data is given as input, and the analyzed data is obtained as output. Specifically, it breaks down the data into words using a morphological analysis tool (e.g., MeCab), extracts relevant keywords (e.g., "can't connect," "signal," "interference"), and calculates their frequency.

[0865] Step 4:

[0866] Fault Detection

[0867] The server detects a failure based on the analysis results. The analyzed data and a set threshold are given as input, and a flag indicating whether a failure has occurred and alert information are obtained as output. Specifically, if the frequency of a specific keyword exceeds the threshold, it determines that a failure has occurred in that area and generates an alert. Location information is also analyzed to identify the affected area.

[0868] Step 5:

[0869] Alert Generation

[0870] The server generates an alert when a failure is detected. The input is a failure detection flag and its detailed information (e.g., the area where the failure occurred, the time of occurrence), and the output is an alert message. Specifically, the server creates an alert message containing detailed information such as the type of failure, the area where the failure occurred, and the time of occurrence, and saves it in the database.

[0871] Step 6:

[0872] Send notifications

[0873] The server notifies the relevant parties of the generated alert. The alert message is given as input, and the sending result is obtained as output. Specifically, the server sends the alert using an API for sending email or SMS (e.g., SendGrid API), and sends the information to a dedicated alert management system.

[0874] Step 7:

[0875] Information Visualization

[0876] The server displays the detected outage information on a dashboard. The outage information is given as input, and the updated dashboard is obtained as output. Specifically, the outage information is displayed in real time using a web application framework (e.g., Flask or Django) and visualized using maps and graphs.

[0877] (Application example 1)

[0878] 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."

[0879] Conventional fault detection and ad delivery systems have difficulty collecting and analyzing information in real time, making it difficult to deliver ads effectively or quickly resolve faults. Furthermore, it is difficult to select optimal ads based on consumer opinions and emotions depending on the region and time of day, limiting the effectiveness of advertising.

[0880] 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.

[0881] In this invention, the server includes means for collecting information from SNS in real time, means for filtering and preprocessing the collected information, means for analyzing the preprocessed information using natural language processing technology, means for calculating region and emotion scores based on the analysis results and selecting advertisements, and means for sending the selected advertisements to users by push notification. This makes it possible to analyze consumer opinions and emotions in real time and quickly deliver optimal advertisements based on them.

[0882] "SNS" refers to social networking services, which are online platforms that allow users to interact with each other over the Internet.

[0883] "Real time" refers to processing that occurs simultaneously with actual time.

[0884] "Data filtering" is the process of selecting only necessary information from collected information.

[0885] "Preprocessing" refers to a series of steps that prepare information for analysis, including noise removal and text normalization.

[0886] "Natural language processing technology" refers to the technology that enables computers to understand and analyze human language, and is capable of performing a variety of language processing.

[0887] An "emotion score" is a numerical representation of the intensity of emotions contained in text data, classified as positive, negative, or neutral.

[0888] "Ad selection" refers to the process of selecting the most suitable advertisement based on the analysis results.

[0889] "Push notifications" are the ability for an application to send messages to a user's device in real time.

[0890] "User" refers to the general consumer who uses this system.

[0891] MODE FOR CARRYING OUT THE INVENTION

[0892] This invention is a system that collects and analyzes real-time information from social media and sends optimal advertisements to users via push notifications based on consumer opinions and emotions. The specific configuration and operation of this system are described below.

[0893] System Configuration

[0894] The system includes the following components:

[0895] 1. Data collection module: Collects information from social media in real time.

[0896] 2. Data preprocessing module: Filters and preprocesses the collected information.

[0897] 3. Analysis module: Analyzes the preprocessed information using natural language processing techniques and calculates the sentiment score.

[0898] 4. Advertisement selection module: Selects the most suitable advertisement based on the analysis results.

[0899] 5. Notification module: Send selected ads to users via push notifications.

[0900] Hardware and software used

[0901] Hardware: Servers, smartphones

[0902] software:

[0903] Twitter API, other SNS APIs

[0904] Database (SQLite)

[0905] Natural language processing libraries (MeCab, NLTK, BERT)

[0906] Push notification service (Firebase Cloud Messaging)

[0907] Operation process

[0908] 1. Data Collection and Preprocessing

[0909] The server collects tweets related to specific keywords (e.g., "new product," "sale," "recommended," etc.) from social media (e.g., Twitter) in real time. This is done using social media APIs such as the Twitter API. The collected tweet data is stored in a temporary database on the server.

[0910] The server then preprocesses the collected tweet data, which includes noise removal (e.g., removing spam messages) and text normalization (e.g., converting uppercase to lowercase).

[0911] 2. Data Analysis

[0912] The preprocessed tweet text is analyzed on the server using natural language processing techniques. Specifically, a morphological analysis tool (e.g., MeCab) is used to break it down into words and calculate the frequency of specific keywords. A sentiment analysis model (e.g., NLTK or BERT) is also used to calculate positive, negative, and neutral sentiment scores.

[0913] 3. Ad selection and delivery

[0914] Based on the analysis results, the ad selection module selects the most suitable ad, based on the sentiment score and location information, and sends it to the user via smartphone push notification.

[0915] Specific examples

[0916] For example, suppose that users in Shibuya post an increasing number of tweets about a new smartphone on social media, many of which are positive. The collected tweet data contains keywords such as "Shibuya," "smartphone," and "great." If the analysis confirms that the tweets have a high positive sentiment score, a promotional ad for the new smartphone will be sent to the user's smartphone during that time period.

[0917] Prompt Sentence Examples

[0918] "Collect the latest positive tweets about new smartphones in Shibuya, calculate the sentiment score, and select ads."

[0919] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0920] Step 1:

[0921] The server collects tweets related to specified keywords from social media (e.g., Twitter) in real time. Here, the Twitter API is used to obtain tweets that match keywords such as "new product" or "sale." These collected tweets are the input data, and the output is that they are stored in a temporary database on the server.

[0922] Step 2:

[0923] The server preprocesses the collected tweet data. This preprocessing includes removing advertisements and spam messages and normalizing the text (e.g., converting uppercase to lowercase). This process outputs clean text data suitable for analysis. This process filters out noise data, resulting in a clear database.

[0924] Step 3:

[0925] The server analyzes the preprocessed tweet text using natural language processing techniques (e.g., MeCab, NLTK, BERT). It performs morphological analysis, breaks it down to the word level, and calculates the frequency of keywords. At the same time, it uses a sentiment analysis model to calculate a positive, negative, or neutral sentiment score for each tweet. The output of this analysis process is a list of keyword frequencies and a list of sentiment scores.

[0926] Step 4:

[0927] The server selects an advertisement based on the analysis results from step 3. Using the sentiment score and keyword frequency obtained from the analysis results, it selects the most suitable advertisement (e.g., new product promotion or sale information). This selected advertisement is the output data.

[0928] Step 5:

[0929] The server sends the selected advertisement to the smartphone user via push notification. This process uses a push notification service such as Firebase Cloud Messaging to deliver the advertisement content to the user's smartphone. The user's receiving device acts as the push notification receiver, and the advertisement message is output.

[0930] Step 6:

[0931] The user can then view the push notification and respond to the ad if necessary. This final step involves the user's actions becoming the final form of input and output, as well as the collection of data to measure the effectiveness of the ad.

[0932] 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.

[0933] The present invention is a system that collects information from SNS (social networking services) in real time, analyzes the information, and recognizes emotions, thereby detecting and quickly responding to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[0934] System Overview

[0935] This system mainly consists of the following components:

[0936] 1. Data collection module: Collects tweets from social media in real time.

[0937] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[0938] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[0939] 4. Emotion Engine: Recognize user emotions from preprocessed tweet text.

[0940] 5. Fault detection module: Detects faults based on the analysis results and information from the emotion engine and generates alerts.

[0941] 6. Visualization module: Displays detected faults on a dashboard.

[0942] 7. Notification module: Sends notifications to relevant parties when a fault is detected.

[0943] Program processing procedure

[0944] Data collection and preprocessing

[0945] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"). Using the streaming API, it receives data as new tweets are posted. The received tweets are stored in a temporary database.

[0946] The server preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[0947] Data analysis and emotion recognition

[0948] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, it analyzes the structure of the tweet and extracts meaningful words and phrases.

[0949] Next, the server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[0950] Furthermore, the server uses an emotion engine to recognize the user's emotions (e.g., anger, frustration, sadness, etc.) from the tweet text, which provides data to evaluate the severity of the problem and the user's dissatisfaction level.

[0951] Fault detection and notification

[0952] The server determines that an outage is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold, taking into account location and other relevant information. If an outage is detected, the server generates an alert and sends a notification to relevant parties. This notification is sent via email, SMS, or a dedicated alert management system.

[0953] Information Visualization

[0954] The server displays detected fault information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[0955] Specific examples

[0956] For example, if mobile phone signals suddenly become unstable in a certain area, many users will post about this problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[0957] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and finds that many of them contain feelings of dissatisfaction and anger.

[0958] Based on this information, the server determines that a serious signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding by conducting an on-site investigation and remote diagnosis to restore service.

[0959] The server displays the progress of the incident and the results of the response in real time on a dashboard, and also visualizes user emotional information, allowing administrators to understand the severity of the incident in more detail.

[0960] In this way, this system can utilize real-time information obtained from social media and emotion recognition technology to quickly and effectively detect and respond to mobile phone service outages.

[0961] The processing flow will be explained below.

[0962] Step 1:

[0963] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"), and uses the streaming API to receive data each time a new tweet is posted.

[0964] Step 2:

[0965] The server stores the received tweets in a temporary database that is kept for a short period of time for subsequent processing.

[0966] Step 3:

[0967] The server retrieves tweet data collected over a fixed time period (e.g., 1 minute) for batch processing. If necessary, the data is gallery-filtered to remove irrelevant tweets such as advertisements and spam.

[0968] Step 4:

[0969] The server then breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the structure of the tweet is analyzed and meaningful words and phrases are extracted.

[0970] Step 5:

[0971] The server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[0972] Step 6:

[0973] The server recognizes the user's emotions from the preprocessed tweet text using an emotion engine, which analyzes the text using natural language processing techniques to identify emotions such as anger, frustration, and sadness.

[0974] Step 7:

[0975] The server determines that an outage is occurring in a region when the frequency of keywords and user sentiment exceed a certain threshold. The region information is determined based on tweet location data and user profile information.

[0976] Step 8:

[0977] The server generates an alert when a failure is detected, which includes detailed information about the failure (e.g., the area where the failure occurred, the time of occurrence, and related keywords) and the user's emotional state.

[0978] Step 9:

[0979] The server notifies designated parties (e.g., incident response teams) of generated alerts via email, SMS, or a dedicated alert management system.

[0980] Step 10:

[0981] The device (the fault response team) receives the notification and immediately begins to act. The notification includes detailed information about the fault and the specific area that needs to be addressed.

[0982] Step 11:

[0983] The device will use on-site inspections and remote diagnostics to identify the cause of the fault and take necessary measures, such as resetting the base station or updating the software.

[0984] Step 12:

[0985] The server displays the progress and results of the problem in real time on a dashboard, which also visualizes user sentiment, allowing administrators to better understand the severity of the problem.

[0986] Step 13:

[0987] Once the issue is resolved, the device updates the dashboard to indicate that the problem has been resolved, so that other stakeholders are aware of the situation.

[0988] Through the above steps, the system of the present invention can utilize real-time information obtained from SNS and emotion recognition technology to quickly and effectively detect and respond to mobile phone service outages.

[0989] Example 2

[0990] 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."

[0991] Conventional fault detection systems for mobile phone services have had problems with detecting faults in real time and being unable to assess the severity of faults based on user emotional information. As a result, early detection and rapid response to faults have not been fully realized, leading to growing user dissatisfaction.

[0992] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for collecting data from SNS in real time, means for filtering and preprocessing the collected data, means for analyzing the preprocessed data using natural language processing technology, means for calculating the frequency of appearance of keywords from the analyzed data, means for recognizing emotions from the analyzed data, means for detecting faults based on the frequency of keywords and emotional information, and means for generating and transmitting alerts, and means for visualizing the detected faults on a dashboard. This enables real-time fault detection and assessment of the severity of faults based on user emotions, thereby realizing a prompt and effective response.

[0993] "SNS" refers to social networking services, which are online platforms where users can post information and share it with other users in real time.

[0994] "Real-time" refers to a state in which data and information are processed and provided immediately without delay.

[0995] "Data collection" refers to the process of gathering specific information or data according to specified criteria.

[0996] "Filtering" refers to the process of removing unnecessary data and noise and selecting data that is suitable for the purpose.

[0997] "Preprocessing" refers to a series of steps that transform data into a form suitable for analysis and processing.

[0998] "Natural language processing technology" refers to the technology that allows computers to understand and analyze human language.

[0999] "Analysis" refers to the process of breaking down given data or information into detail and understanding its meaning and structure.

[1000] "Keyword frequency" refers to the number of times a particular word or phrase appears in text data.

[1001] "Emotion recognition" refers to determining a user's emotional state from data such as text and voice.

[1002] "Fault detection" refers to the process of identifying and identifying anomalies or problems in a system or service.

[1003] An "alert" is a notification sent to warn you when an abnormality or emergency occurs.

[1004] A "dashboard" is an interface that displays information and data visually and intuitively, allowing administrators to understand the situation.

[1005] The present invention is a system that collects information from SNS (social networking services) in real time, analyzes the information, and recognizes emotions in the information, thereby detecting and quickly responding to problems in mobile phone services. The following describes in detail an embodiment of the present invention.

[1006] Data collection

[1007] The server uses a data collection API of a social networking platform, such as the Twitter API, to collect real-time tweets filtered by specific keywords (e.g., "cell phone signal interference"). This real-time data collection is achieved by a program that implements Twitter's streaming API. The data is stored in a temporary database on the server.

[1008] Data Preprocessing

[1009] The server preprocesses the collected tweet data, including noise removal (removing advertisements and spam) and text normalization (converting uppercase letters to lowercase), converting the data into a form suitable for analysis.

[1010] Data analysis

[1011] The server breaks down the preprocessed text data into words using a morphological analysis tool (e.g., MeCab). Morphological analysis analyzes the structure of tweets and extracts meaningful words and phrases. The server then analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.).

[1012] emotion recognition

[1013] The server uses an emotion engine to recognize the user's emotion (e.g., anger, frustration, sadness, etc.) from the tweet text. The emotion engine uses a pre-trained model (e.g., BERT). This analysis determines the intensity and type of the user's emotion, and evaluates the severity of the problem and the user's level of dissatisfaction.

[1014] Fault Detection and Notification

[1015] The server determines that an outage is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold. Location and other relevant information are also taken into account. When an outage is detected, the server generates an alert and sends a notification to relevant parties. Notifications are sent via email, SMS, or a dedicated alert management system.

[1016] Information Visualization

[1017] The server displays detected fault information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[1018] Specific examples

[1019] For example, if mobile phone signals suddenly become unstable in a certain area, many users will post about the problem on social media. Specifically, multiple tweets will be posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[1020] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and finds that many of them contain feelings of dissatisfaction and anger.

[1021] Based on this information, the server determines that a serious signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding by conducting an on-site investigation and remote diagnosis to restore service.

[1022] The server displays the progress of the problem and the results of the response in real time on a dashboard, along with visualizing user sentiment. Administrators can use the dashboard to gain a more detailed understanding of the severity of the problem and take prompt and appropriate action.

[1023] Example prompts for generative AI models

[1024] By inputting the following prompts into the generative AI model, we can obtain a concrete explanation of the system's behavior:

[1025] Based on the following passage, please explain in detail what happens when cell phone signals become unreliable in an area:

[1026] "The server connects to the social media platform's API and sets up a real-time tweet stream using specific keywords (e.g., "cell signal interference")..."

[1027] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1028] Step 1:

[1029] The server connects to the API of the social networking platform and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"). The input is the social networking platform's API key and the monitored keyword, and the output is the collection of real-time tweet data containing the specific keyword. These tweets are stored in a temporary database on the server.

[1030] Step 2:

[1031] The server preprocesses the collected tweet data. The input is the collected tweet data, and the output is filtered, clean data. This preprocessing step includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase to lowercase). Specific operations include data cleaning using regular expressions and text formatting.

[1032] Step 3:

[1033] The server analyzes the preprocessed data using natural language processing technology. The input is preprocessed clean text data, and the output is analyzed text data. Specific operations include breaking down the text into words using a morphological analysis tool (e.g., MeCab) and analyzing its structure. This analysis extracts meaningful words and phrases contained in the tweet content.

[1034] Step 4:

[1035] The server calculates the frequency of occurrence of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) from the analyzed data. The input is the analyzed data broken down into words, and the output is the frequency of occurrence of each keyword. The specific operation in this step is to count the frequency of keywords based on the extracted word list.

[1036] Step 5:

[1037] The server uses an emotion engine to recognize user emotions from tweet text. The input is the analyzed text data, and the output is emotion information for each tweet (e.g., anger, frustration, sadness). The emotion engine uses a pre-trained model (e.g., BERT) and includes a process to determine the intensity and type of user emotion.

[1038] Step 6:

[1039] If the keyword frequency and emotion information exceed a certain threshold, the server determines that a fault has occurred in that area and generates and sends an alert. The input is keyword frequency data and emotion recognition results, and the output is generated alerts and notifications. Specific operations include the process of sending notifications to relevant parties via email or SMS when data exceeding the set threshold is detected.

[1040] Step 7:

[1041] The server visualizes detected fault information and emotion information on a dashboard in real time. The input is alert information and emotion data, and the output is visualized information displayed on the dashboard screen. Specifically, it provides information visually using maps and graphs, allowing administrators to intuitively understand the situation.

[1042] (Application example 2)

[1043] 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."

[1044] In recent years, with the spread of the Internet and social networking sites (SNS), cyber threats and attacks against businesses and individuals have been increasing. However, existing security systems have difficulty detecting and responding to threats in real time. Furthermore, there is a need to quickly identify and respond to truly significant threats from the vast amount of information available on SNS. The present invention aims to provide a system that uses real-time information obtained from SNS to quickly and effectively detect and respond to cyber threats.

[1045] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting information from SNS in real time, means for filtering and pre-processing the collected information, means for analyzing the pre-processed information using natural language processing technology, means for detecting threats based on the analysis results and generating and sending alerts, means for visualizing the detected threats on a dashboard, means for recognizing emotions from the collected information and evaluating the severity of the threats based on the emotion information, and means for notifying relevant parties of information regarding the detected threats. This makes it possible to quickly and effectively detect and respond to cyber threats by utilizing real-time information on SNS.

[1046] "SNS" means social networking service, an online service that allows users to share information and interact with each other via the Internet.

[1047] "Real-time" refers to a situation in which the latest information is collected and processed without momentary delay.

[1048] "Information collection means" refers to a mechanism or method for obtaining specific data or information, and in this invention refers to a means for extracting necessary data from SNS in real time.

[1049] "Filtering" refers to the process of removing unnecessary information from collected data and preparing it in a form suitable for analysis.

[1050] "Preprocessing" refers to the initial processing steps taken to prepare data in a certain format or condition, and includes processes such as text normalization and noise removal.

[1051] "Natural language processing technology" refers to technology for processing human language using computers, and includes text analysis and emotion recognition.

[1052] "Analysis tools" refers to methods and tools used to analyze collected and preprocessed data and extract meaning and patterns.

[1053] "Threat detection measures" refer to the means of identifying and responding to cyber attacks and dangerous behavior based on the results of analysis.

[1054] An "alert" is a warning message sent to notify relevant parties when an abnormality or threat is detected.

[1055] "Notification means" refers to mechanisms and methods for quickly communicating information about detected threats or anomalies to relevant parties.

[1056] "Dashboard" refers to an interface that visually displays an overview of the system status and data, and in the present invention refers to an interface that displays detected threat information.

[1057] "Emotion recognition technology" refers to technology for inferring a user's emotions from text, voice, etc.

[1058] The "importance assessment means" refers to a means for assessing the importance of a threat based on the recognized emotional information and determining the priority of the response.

[1059] The present invention provides a system for detecting and responding to cyber threats by collecting information from social networking sites in real time, filtering and preprocessing the information, and analyzing it using natural language processing technology. The following describes in detail an embodiment of the present invention.

[1060] System Configuration

[1061] This system mainly consists of the following components:

[1062] 1. Data collection module: A means of collecting information from social media in real time.

[1063] 2. Data Preprocessing Module: A means of filtering and preprocessing the collected information.

[1064] 3. Analysis module: A means of analyzing the preprocessed information using natural language processing techniques.

[1065] 4. Emotion Engine: A means of recognizing emotions from pre-processed information and assessing the severity of threats.

[1066] 5. Threat Detection Module: A means of detecting threats based on the analysis results and generating and sending alerts.

[1067] 6. Visualization module: A means of visualizing detected threats on a dashboard.

[1068] 7. Notification module: A means of notifying relevant parties about detected threats.

[1069] Program processing

[1070] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using specific keywords (e.g., "company name cyber attack"). Using the streaming API, it receives data each time a new tweet is posted. The received tweets are stored in a temporary database.

[1071] The server then preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[1072] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). This process analyzes the structure of the tweet and extracts meaningful words and phrases. The extracted word list is then analyzed to calculate the frequency of specific keywords (e.g., "attack," "threat," "danger," etc.). This determines whether the tweet is related to a threat.

[1073] Furthermore, the server uses an emotion engine to recognize the user's emotions (e.g., anger, frustration, sadness, etc.) from the tweet text, which provides data to evaluate the severity of the threat and the user's level of dissatisfaction.

[1074] The server determines that a threat is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold, taking into account location and other relevant information. If a threat is detected, the server generates an alert and sends a notification to relevant parties. This notification can be sent via email, SMS, or a dedicated alert management system.

[1075] Finally, the server displays the detected threat information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides visual information using maps and graphs, allowing administrators to intuitively grasp the situation.

[1076] Specific examples

[1077] For example, if a company faces an increasing threat of cyberattacks, many users will post about this issue on social media. Specifically, suppose multiple tweets are posted along the lines of, "It looks like an important company is under cyberattack right now. I wonder if they're okay?"

[1078] The server collects these tweets via the SNS API and stores them in a database. The server then extracts related keywords such as "company," "attack," and "danger" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and determines that they contain a large number of emotions such as fear and anxiety.

[1079] Based on this information, the server determines that a serious cyber threat has occurred to the company and generates an alert. A notification is sent to the threat response team, which then promptly begins responding by conducting on-site inspections and remote diagnostics to confirm the safety of the service.

[1080] The server displays the progress of threats and response results in real time on a dashboard, and also visualizes user sentiment, allowing administrators to gain a more detailed understanding of the severity of threats.

[1081] In this way, this system utilizes real-time information obtained from social media and emotion recognition technology to quickly and effectively detect and respond to cyber threats.

[1082] Examples of prompt statements

[1083] Please analyze the following text for any threats to your business:

[1084] 1. "It seems like an important company is under cyber attack. Is everything okay?"

[1085] 2. "The credibility of companies has been at risk recently."

[1086] 3. "The company network was attacked and I can't work."

[1087] As a result, please tell us which posts present a significant threat.

[1088] The above is an embodiment of the present invention.

[1089] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1090] Step 1:

[1091] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using specific keywords (e.g., "corporate cyberattack"). Using the streaming API, it receives data as new tweets are posted. The input here is the raw data provided by the API, and the output is the tweet information temporarily stored in a database.

[1092] Step 2:

[1093] The server preprocesses the collected tweet data. Preprocessing includes normalization such as deleting URLs and usernames, removing spaces and unnecessary characters, and converting uppercase to lowercase. The input is the tweet data collected in step 1, and the output is the preprocessed clean text data.

[1094] Step 3:

[1095] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the input is the preprocessed clean text data, and the output is a word list or tokenized text. The frequency of occurrence of specific keywords (e.g., "attack" and "threat") is calculated to determine threat relevance.

[1096] Step 4:

[1097] The server uses an emotion engine to recognize user emotions (e.g., anger, frustration, fear, etc.) from tweet text. The input is a word list or tokenized text, and the output is an emotion score. Emotion recognition can also use generative AI models to provide more accurate emotion information.

[1098] Step 5:

[1099] The server determines that a threat is occurring in a region if the keyword frequency and sentiment score exceed a certain threshold. The input is the keyword frequency analysis result and sentiment score, and the output is the threat detection result. Geographical information and other related information are also taken into account in the evaluation.

[1100] Step 6:

[1101] If a threat is detected, the server generates an alert and sends a notification to the relevant parties via email, SMS, a dedicated alert management system, etc. The input is the threat detection result, and the output is the notification message sent.

[1102] Step 7:

[1103] The server displays detected threat information and sentiment information on a dashboard. The dashboard is implemented as a web application and updates information in real time. The input is threat detection results and sentiment information, and the output is a visual data display that is displayed on the administrator interface. Maps and graphs allow users to intuitively grasp the situation.

[1104] In this way, each step involves data collection, pre-processing, analysis, emotion recognition, threat detection, notification, and visualization.

[1105] 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.

[1106] 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.

[1107] 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.

[1108] [Fourth embodiment]

[1109] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1110] 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.

[1111] 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).

[1112] 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.

[1113] 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.

[1114] 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).

[1115] 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.

[1116] 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.

[1117] 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.

[1118] 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.

[1119] 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.

[1120] 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.

[1121] 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."

[1122] The present invention is a system that collects information in real time from SNS (social networking services) and analyzes it to detect and quickly respond to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[1123] System Overview

[1124] This system mainly consists of the following components:

[1125] 1. Data collection module: Collects tweets from social media in real time.

[1126] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[1127] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[1128] 4. Fault detection module: Detects faults based on the analysis results and generates alerts.

[1129] 5. Visualization module: Displays detected faults on a dashboard.

[1130] 6. Notification module: Sends notifications to relevant parties when a fault is detected.

[1131] Program processing procedure

[1132] Data collection and preprocessing

[1133] The server collects tweets in real time from social networking sites such as Twitter. Specifically, it uses the Twitter API to retrieve tweets related to specific keywords (e.g., "cell phone signal interference") and stores them in a temporary database.

[1134] The server preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[1135] Data analysis

[1136] The server analyzes the preprocessed tweet text using natural language processing technology. Specifically, it uses a morphological analysis tool to break it down into words and calculates the frequency of specific keywords. It then determines whether the tweet contains many keywords related to outages (e.g., "can't connect," "signal," "outage," etc.).

[1137] Fault detection and notification

[1138] The server determines if a keyword frequency exceeds a certain threshold, taking into account location and other relevant information. If an outage is detected, the server generates an alert and sends a notification to relevant parties via email, SMS, or a dedicated alert management system.

[1139] Information Visualization

[1140] The server displays detected fault information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[1141] Specific examples

[1142] For example, if mobile phone signals suddenly become unstable in a certain area, multiple users will post about the problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[1143] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords from the tweets, such as "Shibuya," "signal," and "can't connect." If the frequency of these keywords exceeds a certain threshold, the server determines that a signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding. On-site investigations and remote diagnostics are conducted to restore service.

[1144] In this way, this system can improve the quality of mobile phone services more effectively than conventional methods by utilizing real-time information from SNS to detect failures early and enable rapid response.

[1145] The processing flow will be explained below.

[1146] Step 1:

[1147] The server connects to the API of the social media platform and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference") and uses the streaming API to receive data each time a new tweet is posted.

[1148] Step 2:

[1149] The server stores the received tweets in a temporary database that is kept for a short period of time for subsequent processing.

[1150] Step 3:

[1151] The server retrieves tweet data collected over a fixed period of time (e.g., 1 minute) for batch processing. The data is filtered as needed, for example, to remove irrelevant tweets such as advertisements and spam.

[1152] Step 4:

[1153] The server then breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the structure of the tweet is analyzed and meaningful words and phrases are extracted.

[1154] Step 5:

[1155] The server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[1156] Step 6:

[1157] If the frequency of a keyword exceeds a pre-defined threshold, the server determines that a particular region is experiencing an outage, based on location data from tweets and user profile information.

[1158] Step 7:

[1159] The server generates an alert when a failure is detected, which includes detailed information about the failure (e.g., the location where it occurred, the time of occurrence, and related keywords).

[1160] Step 8:

[1161] The server notifies designated parties (e.g., incident response teams) of generated alerts via email, SMS, or a dedicated alert management system.

[1162] Step 9:

[1163] The device (the fault response team) receives the notification and immediately begins to act. The notification includes detailed information about the fault and the specific area that needs to be addressed.

[1164] Step 10:

[1165] The device will use on-site inspections and remote diagnostics to identify the cause of the fault and take necessary measures, such as resetting the base station or updating the software.

[1166] Step 11:

[1167] The server displays the progress of the problem and the results of the response in real time on a dashboard, allowing administrators to intuitively grasp the progress of the problem response.

[1168] Step 12:

[1169] Once the issue is resolved, the device updates the dashboard to indicate that the problem has been resolved, so that other stakeholders are aware of the situation.

[1170] Through the above steps, the system of the present invention can utilize real-time information obtained from SNS to quickly and efficiently detect and respond to mobile phone service outages.

[1171] Example 1

[1172] 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."

[1173] Conventional device monitoring systems have had difficulty in early detection of failures in mobile phone services. Furthermore, while rapid response is required after failure detection, current systems lack sufficient real-time information collection and analysis, resulting in delayed response. Therefore, the present invention aims to provide a system that collects information from SNS in real time and quickly detects and responds to failures.

[1174] 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.

[1175] In this invention, the server includes means for collecting data from SNS in real time, means for filtering and preprocessing the collected data, means for analyzing the preprocessed data using natural language processing technology, means for detecting faults based on the analysis results and generating and sending alerts, means for visualizing the detected faults on a dashboard, and means for analyzing the collected data and determining the presence or absence of faults using a generative AI model. This makes it possible to analyze the data collected from SNS in real time and detect faults quickly and accurately.

[1176] "SNS" is an abbreviation for social networking service, an online platform where users share information and opinions.

[1177] "Data collection" is the process of obtaining tweets and posts in real time from social media sites and other sources.

[1178] "Filtering" is the process of removing noise and unnecessary information from collected data and converting it into a form suitable for analysis.

[1179] "Preprocessing" refers to the process of formatting data before analyzing it, making it analyzable. Examples of preprocessing include converting uppercase letters to lowercase letters and removing advertisements.

[1180] "Natural language processing technology" is a technology that enables computers to understand, interpret, and generate human language, and includes techniques such as morphological analysis and sentiment analysis.

[1181] "Analysis" is the process of examining pre-processed data using natural language processing techniques to find specific trends and patterns.

[1182] "Fault detection" is the process of determining whether a system or service has some kind of malfunction or problem based on the analysis results.

[1183] "Alert generation" is the process of creating a warning message to communicate real-time information when a fault is detected.

[1184] "Sending" is the process of notifying relevant parties and administrators of the generated alert.

[1185] A "dashboard" is an interface for visually displaying detected fault information, and is designed to make it easy for administrators to understand the situation.

[1186] A "generative AI model" is a model that uses artificial intelligence technology to analyze data and detect specific patterns or anomalies.

[1187] "Preprocessed data" refers to data that has undergone preprocessing such as filtering and normalization and has been converted into a form suitable for analysis.

[1188] "Frequency" refers to the number of times a particular event or keyword occurs, and this number serves as the basis for determining whether or not there is a problem.

[1189] A "threshold" is a boundary value for satisfying a specific condition, and when exceeded, an alert is generated.

[1190] The present invention is a system that collects information from SNS in real time, analyzes it, and thereby detects and responds quickly to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[1191] System Configuration

[1192] This system mainly consists of the following components:

[1193] 1. Data collection module: Collects tweets from social media in real time.

[1194] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[1195] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[1196] 4. Fault detection module: Detects faults based on the analysis results and generates alerts.

[1197] 5. Visualization module: Displays detected faults on a dashboard.

[1198] 6. Notification module: Sends notifications to relevant parties when a fault is detected.

[1199] 7. Generative AI model: Analyzes collected data and determines whether a fault exists.

[1200] Hardware and software used

[1201] The server uses Twitter API and other SNS APIs to collect data from SNS in real time. The collected data is temporarily stored in a database. For data preprocessing, Python libraries (e.g., Pandas, Numpy) are used to remove noise and normalize text.

[1202] For data analysis, we use natural language processing techniques such as morphological analysis tools (e.g., MeCab) and generative AI models (e.g., TensorFlow and PyTorch). Based on the analysis results, we detect failures and use the SendGrid API and other email sending services to generate alerts and send notifications.

[1203] Dashboard visualization uses a web application framework (e.g., Flask or Django) and a front-end visualization library (e.g., D3.js or Chart.js).

[1204] Specific examples

[1205] For example, if mobile phone signals suddenly become unstable in a certain area, multiple users will post about the problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[1206] The server collects these tweets via the SNS API and stores them in a database. The server then extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets. If the frequency of these keywords exceeds a certain threshold, the server determines that a signal interference is occurring in the Shibuya area and generates an alert.

[1207] The generated alert is notified to the fault response team using the SendGrid API, and the device (trouble response team) immediately begins responding, conducting on-site investigations and remote diagnosis to restore service.

[1208] Examples of prompt statements

[1209] The following is a specific example of a prompt sentence to be input to the generative AI model:

[1210] "Analyze tweets containing keywords such as 'signal wave,' 'interference,' 'cannot connect,' and 'unstable' in the specified area and time period, and determine whether the frequency of the relevant keywords exceeds a certain threshold. Example: Tweets in Shibuya from 12:00 to 14:00 on October 12, 2023."

[1211] By using this prompt, data collected from social media can be analyzed by a generative AI model to quickly and accurately determine whether or not a disability exists.

[1212] As described above, this system analyzes data collected from SNS in real time, enabling rapid fault detection and response, thereby improving the quality of mobile phone services more effectively than conventional methods.

[1213] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1214] Step 1:

[1215] Data collection

[1216] The server collects tweets in real time from social networking sites such as Twitter. Specifically, it uses the Twitter API to retrieve tweets related to specific keywords (e.g., "cell phone signal interference"). The Twitter API key and search query are given as input, and the raw data of the retrieved tweets is stored in a database as output. Specifically, it periodically sends API requests to retrieve new tweets.

[1217] Step 2:

[1218] Data Preprocessing

[1219] The server pre-processes the collected data. As input, it takes raw data and as output, it gets pre-processed, clean data. The pre-processing includes the following specific operations:

[1220] Noise reduction: Utilize blacklists to remove advertising and spam tweets.

[1221] Text normalization: convert uppercase to lowercase and remove punctuation marks.

[1222] Language filtering: Filter out tweets in languages ​​other than Japanese.

[1223] Step 3:

[1224] Data analysis

[1225] The server analyzes the preprocessed data using natural language processing technology. The preprocessed clean data is given as input, and the analyzed data is obtained as output. Specifically, it breaks down the data into words using a morphological analysis tool (e.g., MeCab), extracts relevant keywords (e.g., "can't connect," "signal," "interference"), and calculates their frequency.

[1226] Step 4:

[1227] Fault Detection

[1228] The server detects a failure based on the analysis results. The analyzed data and a set threshold are given as input, and a flag indicating whether a failure has occurred and alert information are obtained as output. Specifically, if the frequency of a specific keyword exceeds the threshold, it determines that a failure has occurred in that area and generates an alert. Location information is also analyzed to identify the affected area.

[1229] Step 5:

[1230] Alert Generation

[1231] The server generates an alert when a failure is detected. The input is a failure detection flag and its detailed information (e.g., the area where the failure occurred, the time of occurrence), and the output is an alert message. Specifically, the server creates an alert message containing detailed information such as the type of failure, the area where the failure occurred, and the time of occurrence, and saves it in the database.

[1232] Step 6:

[1233] Send notifications

[1234] The server notifies the relevant parties of the generated alert. The alert message is given as input, and the sending result is obtained as output. Specifically, the server sends the alert using an API for sending email or SMS (e.g., SendGrid API), and sends the information to a dedicated alert management system.

[1235] Step 7:

[1236] Information Visualization

[1237] The server displays the detected outage information on a dashboard. The outage information is given as input, and the updated dashboard is obtained as output. Specifically, the outage information is displayed in real time using a web application framework (e.g., Flask or Django) and visualized using maps and graphs.

[1238] (Application example 1)

[1239] 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."

[1240] Conventional fault detection and ad delivery systems have difficulty collecting and analyzing information in real time, making it difficult to deliver ads effectively or quickly resolve faults. Furthermore, it is difficult to select optimal ads based on consumer opinions and emotions depending on the region and time of day, limiting the effectiveness of advertising.

[1241] 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.

[1242] In this invention, the server includes means for collecting information from SNS in real time, means for filtering and preprocessing the collected information, means for analyzing the preprocessed information using natural language processing technology, means for calculating region and emotion scores based on the analysis results and selecting advertisements, and means for sending the selected advertisements to users by push notification. This makes it possible to analyze consumer opinions and emotions in real time and quickly deliver optimal advertisements based on them.

[1243] "SNS" refers to social networking services, which are online platforms that allow users to interact with each other over the Internet.

[1244] "Real time" refers to processing that occurs simultaneously with actual time.

[1245] "Data filtering" is the process of selecting only necessary information from collected information.

[1246] "Preprocessing" refers to a series of steps that prepare information for analysis, including noise removal and text normalization.

[1247] "Natural language processing technology" refers to the technology that enables computers to understand and analyze human language, and is capable of performing a variety of language processing.

[1248] An "emotion score" is a numerical representation of the intensity of emotions contained in text data, classified as positive, negative, or neutral.

[1249] "Ad selection" refers to the process of selecting the most suitable advertisement based on the analysis results.

[1250] "Push notifications" are the ability for an application to send messages to a user's device in real time.

[1251] "User" refers to the general consumer who uses this system.

[1252] MODE FOR CARRYING OUT THE INVENTION

[1253] This invention is a system that collects and analyzes real-time information from social media and sends optimal advertisements to users via push notifications based on consumer opinions and emotions. The specific configuration and operation of this system are described below.

[1254] System Configuration

[1255] The system includes the following components:

[1256] 1. Data collection module: Collects information from social media in real time.

[1257] 2. Data preprocessing module: Filters and preprocesses the collected information.

[1258] 3. Analysis module: Analyzes the preprocessed information using natural language processing techniques and calculates the sentiment score.

[1259] 4. Advertisement selection module: Selects the most suitable advertisement based on the analysis results.

[1260] 5. Notification module: Send selected ads to users via push notifications.

[1261] Hardware and software used

[1262] Hardware: Servers, smartphones

[1263] software:

[1264] Twitter API, other SNS APIs

[1265] Database (SQLite)

[1266] Natural language processing libraries (MeCab, NLTK, BERT)

[1267] Push notification service (Firebase Cloud Messaging)

[1268] Operation process

[1269] 1. Data Collection and Preprocessing

[1270] The server collects tweets related to specific keywords (e.g., "new product," "sale," "recommended," etc.) from social media (e.g., Twitter) in real time. This is done using social media APIs such as the Twitter API. The collected tweet data is stored in a temporary database on the server.

[1271] The server then preprocesses the collected tweet data, which includes noise removal (e.g., removing spam messages) and text normalization (e.g., converting uppercase to lowercase).

[1272] 2. Data Analysis

[1273] The preprocessed tweet text is analyzed on the server using natural language processing techniques. Specifically, a morphological analysis tool (e.g., MeCab) is used to break it down into words and calculate the frequency of specific keywords. A sentiment analysis model (e.g., NLTK or BERT) is also used to calculate positive, negative, and neutral sentiment scores.

[1274] 3. Ad selection and delivery

[1275] Based on the analysis results, the ad selection module selects the most suitable ad, based on the sentiment score and location information, and sends it to the user via smartphone push notification.

[1276] Specific examples

[1277] For example, suppose that users in Shibuya post an increasing number of tweets about a new smartphone on social media, many of which are positive. The collected tweet data contains keywords such as "Shibuya," "smartphone," and "great." If the analysis confirms that the tweets have a high positive sentiment score, a promotional ad for the new smartphone will be sent to the user's smartphone during that time period.

[1278] Prompt Sentence Examples

[1279] "Collect the latest positive tweets about new smartphones in Shibuya, calculate the sentiment score, and select ads."

[1280] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1281] Step 1:

[1282] The server collects tweets related to specified keywords from social media (e.g., Twitter) in real time. Here, the Twitter API is used to obtain tweets that match keywords such as "new product" or "sale." These collected tweets are the input data, and the output is that they are stored in a temporary database on the server.

[1283] Step 2:

[1284] The server preprocesses the collected tweet data. This preprocessing includes removing advertisements and spam messages and normalizing the text (e.g., converting uppercase to lowercase). This process outputs clean text data suitable for analysis. This process filters out noise data, resulting in a clear database.

[1285] Step 3:

[1286] The server analyzes the preprocessed tweet text using natural language processing techniques (e.g., MeCab, NLTK, BERT). It performs morphological analysis, breaks it down to the word level, and calculates the frequency of keywords. At the same time, it uses a sentiment analysis model to calculate a positive, negative, or neutral sentiment score for each tweet. The output of this analysis process is a list of keyword frequencies and a list of sentiment scores.

[1287] Step 4:

[1288] The server selects an advertisement based on the analysis results from step 3. Using the sentiment score and keyword frequency obtained from the analysis results, it selects the most suitable advertisement (e.g., new product promotion or sale information). This selected advertisement is the output data.

[1289] Step 5:

[1290] The server sends the selected advertisement to the smartphone user via push notification. This process uses a push notification service such as Firebase Cloud Messaging to deliver the advertisement content to the user's smartphone. The user's receiving device acts as the push notification receiver, and the advertisement message is output.

[1291] Step 6:

[1292] The user can then view the push notification and respond to the ad if necessary. This final step involves the user's actions becoming the final form of input and output, as well as the collection of data to measure the effectiveness of the ad.

[1293] 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.

[1294] The present invention is a system that collects information from SNS (social networking services) in real time, analyzes the information, and recognizes emotions, thereby detecting and quickly responding to mobile phone service failures that are difficult to detect using conventional device monitoring. The following describes in detail an embodiment of the present invention.

[1295] System Overview

[1296] This system mainly consists of the following components:

[1297] 1. Data collection module: Collects tweets from social media in real time.

[1298] 2. Data preprocessing module: The collected tweets are filtered and converted into a form suitable for analysis.

[1299] 3. Analysis module: Analyzes the preprocessed tweets using natural language processing techniques.

[1300] 4. Emotion Engine: Recognize user emotions from preprocessed tweet text.

[1301] 5. Fault detection module: Detects faults based on the analysis results and information from the emotion engine and generates alerts.

[1302] 6. Visualization module: Displays detected faults on a dashboard.

[1303] 7. Notification module: Sends notifications to relevant parties when a fault is detected.

[1304] Program processing procedure

[1305] Data collection and preprocessing

[1306] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"). Using the streaming API, it receives data as new tweets are posted. The received tweets are stored in a temporary database.

[1307] The server preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[1308] Data analysis and emotion recognition

[1309] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, it analyzes the structure of the tweet and extracts meaningful words and phrases.

[1310] Next, the server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[1311] Furthermore, the server uses an emotion engine to recognize the user's emotions (e.g., anger, frustration, sadness, etc.) from the tweet text, which provides data to evaluate the severity of the problem and the user's dissatisfaction level.

[1312] Fault detection and notification

[1313] The server determines that an outage is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold, taking into account location and other relevant information. If an outage is detected, the server generates an alert and sends a notification to relevant parties. This notification is sent via email, SMS, or a dedicated alert management system.

[1314] Information Visualization

[1315] The server displays detected fault information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[1316] Specific examples

[1317] For example, if mobile phone signals suddenly become unstable in a certain area, many users will post about this problem on social media. Specifically, suppose multiple tweets are posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[1318] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and finds that many of them contain feelings of dissatisfaction and anger.

[1319] Based on this information, the server determines that a serious signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding by conducting an on-site investigation and remote diagnosis to restore service.

[1320] The server displays the progress of the incident and the results of the response in real time on a dashboard, and also visualizes user emotional information, allowing administrators to understand the severity of the incident in more detail.

[1321] In this way, this system can utilize real-time information obtained from social media and emotion recognition technology to quickly and effectively detect and respond to mobile phone service outages.

[1322] The processing flow will be explained below.

[1323] Step 1:

[1324] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"), and uses the streaming API to receive data each time a new tweet is posted.

[1325] Step 2:

[1326] The server stores the received tweets in a temporary database that is kept for a short period of time for subsequent processing.

[1327] Step 3:

[1328] The server retrieves tweet data collected over a fixed time period (e.g., 1 minute) for batch processing. If necessary, the data is gallery-filtered to remove irrelevant tweets such as advertisements and spam.

[1329] Step 4:

[1330] The server then breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the structure of the tweet is analyzed and meaningful words and phrases are extracted.

[1331] Step 5:

[1332] The server analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) to determine whether a tweet is related to an outage.

[1333] Step 6:

[1334] The server recognizes the user's emotions from the preprocessed tweet text using an emotion engine, which analyzes the text using natural language processing techniques to identify emotions such as anger, frustration, and sadness.

[1335] Step 7:

[1336] The server determines that an outage is occurring in a region when the frequency of keywords and user sentiment exceed a certain threshold. The region information is determined based on tweet location data and user profile information.

[1337] Step 8:

[1338] The server generates an alert when a failure is detected, which includes detailed information about the failure (e.g., the area where the failure occurred, the time of occurrence, and related keywords) and the user's emotional state.

[1339] Step 9:

[1340] The server notifies designated parties (e.g., incident response teams) of generated alerts via email, SMS, or a dedicated alert management system.

[1341] Step 10:

[1342] The device (the fault response team) receives the notification and immediately begins to act. The notification includes detailed information about the fault and the specific area that needs to be addressed.

[1343] Step 11:

[1344] The device will use on-site inspections and remote diagnostics to identify the cause of the fault and take necessary measures, such as resetting the base station or updating the software.

[1345] Step 12:

[1346] The server displays the progress and results of the problem in real time on a dashboard, which also visualizes user sentiment, allowing administrators to better understand the severity of the problem.

[1347] Step 13:

[1348] Once the issue is resolved, the device updates the dashboard to indicate that the problem has been resolved, so that other stakeholders are aware of the situation.

[1349] Through the above steps, the system of the present invention can utilize real-time information obtained from SNS and emotion recognition technology to quickly and effectively detect and respond to mobile phone service outages.

[1350] Example 2

[1351] 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."

[1352] Conventional fault detection systems for mobile phone services have had problems with detecting faults in real time and being unable to assess the severity of faults based on user emotional information. As a result, early detection and rapid response to faults have not been fully realized, leading to growing user dissatisfaction.

[1353] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for collecting data from SNS in real time, means for filtering and preprocessing the collected data, means for analyzing the preprocessed data using natural language processing technology, means for calculating the frequency of appearance of keywords from the analyzed data, means for recognizing emotions from the analyzed data, means for detecting faults based on the frequency of keywords and emotional information, and means for generating and transmitting alerts, and means for visualizing the detected faults on a dashboard. This enables real-time fault detection and assessment of the severity of faults based on user emotions, thereby realizing a prompt and effective response.

[1354] "SNS" refers to social networking services, which are online platforms where users can post information and share it with other users in real time.

[1355] "Real-time" refers to a state in which data and information are processed and provided immediately without delay.

[1356] "Data collection" refers to the process of gathering specific information or data according to specified criteria.

[1357] "Filtering" refers to the process of removing unnecessary data and noise and selecting data that is suitable for the purpose.

[1358] "Preprocessing" refers to a series of steps that transform data into a form suitable for analysis and processing.

[1359] "Natural language processing technology" refers to the technology that allows computers to understand and analyze human language.

[1360] "Analysis" refers to the process of breaking down given data or information into detail and understanding its meaning and structure.

[1361] "Keyword frequency" refers to the number of times a particular word or phrase appears in text data.

[1362] "Emotion recognition" refers to determining a user's emotional state from data such as text and voice.

[1363] "Fault detection" refers to the process of identifying and identifying anomalies or problems in a system or service.

[1364] An "alert" is a notification sent to warn you when an abnormality or emergency occurs.

[1365] A "dashboard" is an interface that displays information and data visually and intuitively, allowing administrators to understand the situation.

[1366] The present invention is a system that collects information from SNS (social networking services) in real time, analyzes the information, and recognizes emotions in the information, thereby detecting and quickly responding to problems in mobile phone services. The following describes in detail an embodiment of the present invention.

[1367] Data collection

[1368] The server uses a data collection API of a social networking platform, such as the Twitter API, to collect real-time tweets filtered by specific keywords (e.g., "cell phone signal interference"). This real-time data collection is achieved by a program that implements Twitter's streaming API. The data is stored in a temporary database on the server.

[1369] Data Preprocessing

[1370] The server preprocesses the collected tweet data, including noise removal (removing advertisements and spam) and text normalization (converting uppercase letters to lowercase), converting the data into a form suitable for analysis.

[1371] Data analysis

[1372] The server breaks down the preprocessed text data into words using a morphological analysis tool (e.g., MeCab). Morphological analysis analyzes the structure of tweets and extracts meaningful words and phrases. The server then analyzes the extracted word list and calculates the frequency of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.).

[1373] emotion recognition

[1374] The server uses an emotion engine to recognize the user's emotion (e.g., anger, frustration, sadness, etc.) from the tweet text. The emotion engine uses a pre-trained model (e.g., BERT). This analysis determines the intensity and type of the user's emotion, and evaluates the severity of the problem and the user's level of dissatisfaction.

[1375] Fault Detection and Notification

[1376] The server determines that an outage is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold. Location and other relevant information are also taken into account. When an outage is detected, the server generates an alert and sends a notification to relevant parties. Notifications are sent via email, SMS, or a dedicated alert management system.

[1377] Information Visualization

[1378] The server displays detected fault information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides information visually using maps and graphs, allowing administrators to intuitively grasp the situation.

[1379] Specific examples

[1380] For example, if mobile phone signals suddenly become unstable in a certain area, many users will post about the problem on social media. Specifically, multiple tweets will be posted saying, "I can't get any mobile phone signals in Shibuya right now. What's going on?"

[1381] The server collects these tweets via the SNS API and stores them in a database. Next, the server extracts related keywords such as "Shibuya," "signal loss," and "can't connect" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and finds that many of them contain feelings of dissatisfaction and anger.

[1382] Based on this information, the server determines that a serious signal outage is occurring in the Shibuya area and generates an alert. A notification is sent to the troubleshooting team, which then promptly begins responding by conducting an on-site investigation and remote diagnosis to restore service.

[1383] The server displays the progress of the problem and the results of the response in real time on a dashboard, along with visualizing user sentiment. Administrators can use the dashboard to gain a more detailed understanding of the severity of the problem and take prompt and appropriate action.

[1384] Example prompts for generative AI models

[1385] By inputting the following prompts into the generative AI model, we can obtain a concrete explanation of the system's behavior:

[1386] Based on the following passage, please explain in detail what happens when cell phone signals become unreliable in an area:

[1387] "The server connects to the social media platform's API and sets up a real-time tweet stream using specific keywords (e.g., "cell signal interference")..."

[1388] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1389] Step 1:

[1390] The server connects to the API of the social networking platform and sets up a real-time tweet stream using a specific keyword (e.g., "cell signal interference"). The input is the social networking platform's API key and the monitored keyword, and the output is the collection of real-time tweet data containing the specific keyword. These tweets are stored in a temporary database on the server.

[1391] Step 2:

[1392] The server preprocesses the collected tweet data. The input is the collected tweet data, and the output is filtered, clean data. This preprocessing step includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase to lowercase). Specific operations include data cleaning using regular expressions and text formatting.

[1393] Step 3:

[1394] The server analyzes the preprocessed data using natural language processing technology. The input is preprocessed clean text data, and the output is analyzed text data. Specific operations include breaking down the text into words using a morphological analysis tool (e.g., MeCab) and analyzing its structure. This analysis extracts meaningful words and phrases contained in the tweet content.

[1395] Step 4:

[1396] The server calculates the frequency of occurrence of specific keywords (e.g., "signal wave," "interference," "cannot connect," etc.) from the analyzed data. The input is the analyzed data broken down into words, and the output is the frequency of occurrence of each keyword. The specific operation in this step is to count the frequency of keywords based on the extracted word list.

[1397] Step 5:

[1398] The server uses an emotion engine to recognize user emotions from tweet text. The input is the analyzed text data, and the output is emotion information for each tweet (e.g., anger, frustration, sadness). The emotion engine uses a pre-trained model (e.g., BERT) and includes a process to determine the intensity and type of user emotion.

[1399] Step 6:

[1400] If the keyword frequency and emotion information exceed a certain threshold, the server determines that a fault has occurred in that area and generates and sends an alert. The input is keyword frequency data and emotion recognition results, and the output is generated alerts and notifications. Specific operations include the process of sending notifications to relevant parties via email or SMS when data exceeding the set threshold is detected.

[1401] Step 7:

[1402] The server visualizes detected fault information and emotion information on a dashboard in real time. The input is alert information and emotion data, and the output is visualized information displayed on the dashboard screen. Specifically, it provides information visually using maps and graphs, allowing administrators to intuitively understand the situation.

[1403] (Application example 2)

[1404] 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."

[1405] In recent years, with the spread of the Internet and social networking sites (SNS), cyber threats and attacks against businesses and individuals have been increasing. However, existing security systems have difficulty detecting and responding to threats in real time. Furthermore, there is a need to quickly identify and respond to truly significant threats from the vast amount of information available on SNS. The present invention aims to provide a system that uses real-time information obtained from SNS to quickly and effectively detect and respond to cyber threats.

[1406] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting information from SNS in real time, means for filtering and pre-processing the collected information, means for analyzing the pre-processed information using natural language processing technology, means for detecting threats based on the analysis results and generating and sending alerts, means for visualizing the detected threats on a dashboard, means for recognizing emotions from the collected information and evaluating the severity of the threats based on the emotion information, and means for notifying relevant parties of information regarding the detected threats. This makes it possible to quickly and effectively detect and respond to cyber threats by utilizing real-time information on SNS.

[1407] "SNS" means social networking service, an online service that allows users to share information and interact with each other via the Internet.

[1408] "Real-time" refers to a situation in which the latest information is collected and processed without momentary delay.

[1409] "Information collection means" refers to a mechanism or method for obtaining specific data or information, and in this invention refers to a means for extracting necessary data from SNS in real time.

[1410] "Filtering" refers to the process of removing unnecessary information from collected data and preparing it in a form suitable for analysis.

[1411] "Preprocessing" refers to the initial processing steps taken to prepare data in a certain format or condition, and includes processes such as text normalization and noise removal.

[1412] "Natural language processing technology" refers to technology for processing human language using computers, and includes text analysis and emotion recognition.

[1413] "Analysis tools" refers to methods and tools used to analyze collected and preprocessed data and extract meaning and patterns.

[1414] "Threat detection measures" refer to the means of identifying and responding to cyber attacks and dangerous behavior based on the results of analysis.

[1415] An "alert" is a warning message sent to notify relevant parties when an abnormality or threat is detected.

[1416] "Notification means" refers to mechanisms and methods for quickly communicating information about detected threats or anomalies to relevant parties.

[1417] "Dashboard" refers to an interface that visually displays an overview of the system status and data, and in the present invention refers to an interface that displays detected threat information.

[1418] "Emotion recognition technology" refers to technology for inferring a user's emotions from text, voice, etc.

[1419] The "importance assessment means" refers to a means for assessing the importance of a threat based on the recognized emotional information and determining the priority of the response.

[1420] The present invention provides a system for detecting and responding to cyber threats by collecting information from social networking sites in real time, filtering and preprocessing the information, and analyzing it using natural language processing technology. The following describes in detail an embodiment of the present invention.

[1421] System Configuration

[1422] This system mainly consists of the following components:

[1423] 1. Data collection module: A means of collecting information from social media in real time.

[1424] 2. Data Preprocessing Module: A means of filtering and preprocessing the collected information.

[1425] 3. Analysis module: A means of analyzing the preprocessed information using natural language processing techniques.

[1426] 4. Emotion Engine: A means of recognizing emotions from pre-processed information and assessing the severity of threats.

[1427] 5. Threat Detection Module: A means of detecting threats based on the analysis results and generating and sending alerts.

[1428] 6. Visualization module: A means of visualizing detected threats on a dashboard.

[1429] 7. Notification module: A means of notifying relevant parties about detected threats.

[1430] Program processing

[1431] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using specific keywords (e.g., "company name cyber attack"). Using the streaming API, it receives data each time a new tweet is posted. The received tweets are stored in a temporary database.

[1432] The server then preprocesses the collected tweet data, which includes noise removal (e.g., removing advertisements and spam) and text normalization (e.g., converting uppercase letters to lowercase). The preprocessed data is then sent to the next analysis step.

[1433] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). This process analyzes the structure of the tweet and extracts meaningful words and phrases. The extracted word list is then analyzed to calculate the frequency of specific keywords (e.g., "attack," "threat," "danger," etc.). This determines whether the tweet is related to a threat.

[1434] Furthermore, the server uses an emotion engine to recognize the user's emotions (e.g., anger, frustration, sadness, etc.) from the tweet text, which provides data to evaluate the severity of the threat and the user's level of dissatisfaction.

[1435] The server determines that a threat is occurring in a region if the frequency of keywords and user sentiment exceed a certain threshold, taking into account location and other relevant information. If a threat is detected, the server generates an alert and sends a notification to relevant parties. This notification can be sent via email, SMS, or a dedicated alert management system.

[1436] Finally, the server displays the detected threat information and sentiment information on a dashboard. The dashboard is implemented as a web application and is updated in real time. It provides visual information using maps and graphs, allowing administrators to intuitively grasp the situation.

[1437] Specific examples

[1438] For example, if a company faces an increasing threat of cyberattacks, many users will post about this issue on social media. Specifically, suppose multiple tweets are posted along the lines of, "It looks like an important company is under cyberattack right now. I wonder if they're okay?"

[1439] The server collects these tweets via the SNS API and stores them in a database. The server then extracts related keywords such as "company," "attack," and "danger" from the tweets, and simultaneously recognizes emotions from the text posted by users. For example, the emotion engine analyzes these tweets and determines that they contain a large number of emotions such as fear and anxiety.

[1440] Based on this information, the server determines that a serious cyber threat has occurred to the company and generates an alert. A notification is sent to the threat response team, which then promptly begins responding by conducting on-site inspections and remote diagnostics to confirm the safety of the service.

[1441] The server displays the progress of threats and response results in real time on a dashboard, and also visualizes user sentiment, allowing administrators to gain a more detailed understanding of the severity of threats.

[1442] In this way, this system utilizes real-time information obtained from social media and emotion recognition technology to quickly and effectively detect and respond to cyber threats.

[1443] Examples of prompt statements

[1444] Please analyze the following text for any threats to your business:

[1445] 1. "It seems like an important company is under cyber attack. Is everything okay?"

[1446] 2. "The credibility of companies has been at risk recently."

[1447] 3. "The company network was attacked and I can't work."

[1448] As a result, please tell us which posts present a significant threat.

[1449] The above is an embodiment of the present invention.

[1450] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1451] Step 1:

[1452] The server connects to the API of a social media platform (e.g., Twitter) and sets up a real-time tweet stream using specific keywords (e.g., "corporate cyberattack"). Using the streaming API, it receives data as new tweets are posted. The input here is the raw data provided by the API, and the output is the tweet information temporarily stored in a database.

[1453] Step 2:

[1454] The server preprocesses the collected tweet data. Preprocessing includes normalization such as deleting URLs and usernames, removing spaces and unnecessary characters, and converting uppercase to lowercase. The input is the tweet data collected in step 1, and the output is the preprocessed clean text data.

[1455] Step 3:

[1456] The server breaks down the preprocessed tweet text into words using a morphological analyzer (e.g., MeCab). In this process, the input is the preprocessed clean text data, and the output is a word list or tokenized text. The frequency of occurrence of specific keywords (e.g., "attack" and "threat") is calculated to determine threat relevance.

[1457] Step 4:

[1458] The server uses an emotion engine to recognize user emotions (e.g., anger, frustration, fear, etc.) from tweet text. The input is a word list or tokenized text, and the output is an emotion score. Emotion recognition can also use generative AI models to provide more accurate emotion information.

[1459] Step 5:

[1460] The server determines that a threat is occurring in a region if the keyword frequency and sentiment score exceed a certain threshold. The input is the keyword frequency analysis result and sentiment score, and the output is the threat detection result. Geographical information and other related information are also taken into account in the evaluation.

[1461] Step 6:

[1462] If a threat is detected, the server generates an alert and sends a notification to the relevant parties via email, SMS, a dedicated alert management system, etc. The input is the threat detection result, and the output is the notification message sent.

[1463] Step 7:

[1464] The server displays detected threat information and sentiment information on a dashboard. The dashboard is implemented as a web application and updates information in real time. The input is threat detection results and sentiment information, and the output is a visual data display that is displayed on the administrator interface. Maps and graphs allow users to intuitively grasp the situation.

[1465] In this way, each step involves data collection, pre-processing, analysis, emotion recognition, threat detection, notification, and visualization.

[1466] 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.

[1467] 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.

[1468] 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.

[1469] 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.

[1470] 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.

[1471] 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.

[1472] 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).

[1473] 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.

[1474] 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."

[1475] 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.

[1476] 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).

[1477] 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.

[1478] 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.

[1479] 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.

[1480] 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.

[1481] 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.

[1482] 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.

[1483] 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.

[1484] 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.

[1485] 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.

[1486] 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.

[1487] The following is further disclosed regarding the above embodiment.

[1488] (Claim 1)

[1489] A means of collecting information in real time from SNS,

[1490] means for filtering and pre-processing the collected information;

[1491] means for analyzing the preprocessed information using natural language processing technology;

[1492] a means for detecting faults and generating and sending alerts based on the analysis results;

[1493] The system includes a means for visualizing detected faults on a dashboard.

[1494] (Claim 2)

[1495] 10. The system of claim 1, further comprising means for sending a notification to a troubleshooting team based on the alert.

[1496] (Claim 3)

[1497] 10. The system of claim 1, further comprising means for performing fault detection based on a plurality of area information using the filtered and preprocessed information.

[1498] (Claim 4)

[1499] 2. The system according to claim 1, further comprising means for investigating and recovering from a failure based on the notified information.

[1500] (Claim 5)

[1501] 10. The system of claim 1, further comprising means for displaying the visualized information in the form of maps and graphs in the dashboard.

[1502] "Example 1"

[1503] (Claim 1)

[1504] A means of collecting data from SNS in real time,

[1505] means for filtering and pre-processing the collected data;

[1506] means for analyzing the preprocessed data using natural language processing techniques;

[1507] a means for detecting faults and generating and sending alerts based on the analysis results;

[1508] A means to visualize detected faults on a dashboard,

[1509] a means for analyzing the collected data using a generative AI model to determine whether a fault exists; and

[1510] A system including:

[1511] (Claim 2)

[1512] 10. The system of claim 1, further comprising means for sending a notification to a troubleshooting team based on the alert.

[1513] (Claim 3)

[1514] 10. The system of claim 1, further comprising means for performing fault detection based on the plurality of area data using the filtered and preprocessed data.

[1515] "Application Example 1"

[1516] (Claim 1)

[1517] A means of collecting information in real time from SNS,

[1518] means for filtering and pre-processing the collected information;

[1519] means for analyzing the preprocessed information using natural language processing technology;

[1520] A means for calculating region and sentiment scores based on the analysis results and selecting advertisements;

[1521] A system including means for sending selected advertisements to users via push notifications.

[1522] (Claim 2)

[1523] 10. The system of claim 1, further comprising means for selecting an optimal advertisement based on the calculated sentiment score and the region information and sending a notification.

[1524] (Claim 3)

[1525] 10. The system of claim 1, further comprising means for using the filtered and pre-processed information to make advertisement selections based on a plurality of geographical and time-of-day information.

[1526] "Example 2: Combining Emotion Engines"

[1527] (Claim 1)

[1528] A means of collecting data from SNS in real time,

[1529] means for filtering and pre-processing the collected data;

[1530] means for analyzing the preprocessed data using natural language processing techniques;

[1531] A means for calculating the frequency of occurrence of keywords from the analyzed data;

[1532] A means for recognizing emotions from the analyzed data;

[1533] means for detecting faults and generating and sending alerts based on keyword frequency and sentiment information;

[1534] The system includes a means for visualizing detected faults on a dashboard.

[1535] (Claim 2)

[1536] 10. The system of claim 1, further comprising means for sending a notification to a troubleshooting team based on the alert.

[1537] (Claim 3)

[1538] 10. The system of claim 1, further comprising means for performing fault detection based on multiple area information using the filtered and preprocessed data.

[1539] "Application example 2 when combining emotion engines"

[1540] (Claim 1)

[1541] A means of collecting information in real time from SNS,

[1542] means for filtering and pre-processing the collected information;

[1543] means for analyzing the preprocessed information using natural language processing technology;

[1544] A means of detecting threats and generating and sending alerts based on the analysis results;

[1545] A means to visualize detected threats on a dashboard, and

[1546] A means for recognizing emotions from collected information and assessing the severity of threats based on the emotional information;

[1547] A system that includes a means to notify relevant parties about detected threats.

[1548] (Claim 2)

[1549] 10. The system of claim 1, further comprising means for sending a notification to a threat response team based on the alert.

[1550] (Claim 3)

[1551] 10. The system of claim 1, further comprising means for using the filtered and preprocessed information to perform threat detection based on multiple region information. [Explanation of symbols]

[1552] 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 in real time from SNS, means for filtering and pre-processing the collected information; means for analyzing the preprocessed information using natural language processing technology; a means for detecting faults and generating and sending alerts based on the analysis results; The system includes a means for visualizing detected faults on a dashboard.

2. The system of claim 1 , further comprising means for sending a notification to a troubleshooting team based on the alert.

3. 10. The system of claim 1, further comprising means for performing fault detection based on a plurality of area information using the filtered and preprocessed information.

4. 2. The system according to claim 1, further comprising means for investigating and recovering from a failure based on the notified information.

5. The system of claim 1 , further comprising means for displaying the visualized information in the form of maps and graphs in the dashboard.

Images