system

Abstract

We provide the system. [Solution] Means of collecting information from multiple sources on the internet, A means for performing initial filtering of the collected information using natural language processing technology, A means for performing a detailed sentiment analysis on the initially filtered information, Methods for profiling the analyzed information and cross-checking it with past databases, Means for reporting information and senders deemed problematic to the relevant authorities, A system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, and includes steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a 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

Summary of the Invention

Problems to be Solved by the Invention

[0004] In recent years, the spread of slander, crime warnings, and false information on the Internet has increased, becoming a serious problem for society. Such information exists across multiple information sources, and it is difficult to manually check each piece of information, so an efficient countermeasure is required. Also, it is necessary to identify problem information at an early stage and report it to relevant agencies to improve social safety. Furthermore, the deletion of false information and the identification of problematic senders are also important, and technologies for automatically performing these are required.

Means for Solving the Problems

[0005] This invention provides a system for efficiently monitoring information on the internet and automatically identifying, analyzing, and reporting inappropriate information. Specifically, it collects data from multiple sources and performs initial filtering using natural language processing technology. Subsequently, it performs detailed sentiment analysis and profiling of the collected information and compares it with past databases. Information deemed problematic is immediately reported to the relevant authorities, including specific evidence, enabling efficient response. Through these means, malicious information on the internet can be quickly identified, improving social safety.

[0006] "Information sources" refer to places such as websites, social media platforms, and forums that provide or distribute information on the internet.

[0007] "Natural language processing technology" refers to the technology used by computers to understand, analyze, and generate natural language that humans use on a daily basis.

[0008] "Initial filtering" refers to the process of selecting inappropriate information from collected data based on basic keyword checks and simple rules.

[0009] "Sentiment analysis" refers to the technique of identifying opinions and emotions within a text and analyzing whether they are positive, negative, or neutral.

[0010] "Profiling" is a method of predicting the characteristics and behavior of individuals or groups based on specific behavioral patterns and attributes.

[0011] "Cross-checking" is a method of verifying the accuracy and consistency of collected data by comparing it with multiple sources and databases.

[0012] "Reporting" refers to the act of detecting problematic information and communicating that information to a designated agency or relevant party.

[0013] "Evidence" refers to specific data or records that support a particular claim or fact. [Brief explanation of the drawing]

[0014] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, which incorporates an emotion engine. [Figure 14] This is a sequence diagram showing the processing flow of the data processing system in Application Example 2, which combines an emotion engine.

Embodiments for Carrying Out the Invention

[0015] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0016] First, the terms used in the following description will be explained.

[0017] In the following embodiments, a numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0018] In the following embodiments, a numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0019] In the following embodiments, a numbered storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, and the like.

[0020] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0022] [First Embodiment]

[0023] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0024] As shown in Figure 1, the 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.

[0025] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0026] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0027] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and 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.

[0028] 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 perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0030] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] As shown in Figure 2, in the data processing device 12, a specific processing 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" related to the technology of this 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 according to the specific processing program 56 executed on the RAM 30.

[0032] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0033] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0034] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0035] This invention is a system that automatically collects data from various sources on the internet and analyzes it to identify problematic content such as defamation, crime threats, and misinformation, and to respond to it quickly. The system is server-centric and capable of efficiently processing large amounts of data.

[0036] The server first collects information from multiple websites and social media platforms. This data collection may be done directly via APIs or through web scraping techniques. The collected information is then initially filtered within the server using natural language processing. This process identifies potentially problematic content based on a pre-configured list of inappropriate keywords.

[0037] Next, the information that passes the initial filtering undergoes further sentiment and contextual analysis by the server. At this stage, machine learning models are used to determine whether the sentiment of a post is positive, negative, or neutral. Furthermore, contextual information is analyzed to understand whether the content of the post is threatening or related to criminal activity.

[0038] Subsequently, the server uses the analysis results to profile the characteristics and past behavior of the inappropriate sender, and cross-checks them with existing databases. This process checks for similar behavioral patterns in the past and infers the likelihood of recidivism.

[0039] When the system identifies inappropriate information or a source, the server notifies the relevant authorities. This notification includes detailed information about the problematic post and its source, as well as supporting data. This allows the authorities to take swift action.

[0040] As a concrete example, consider a case where a user posts a threatening message to a specific individual on social media. This post is automatically collected by the server, and threatening keywords are detected during the initial filtering stage. Subsequently, sentiment analysis confirms the negativity of the post, and during the profiling stage, it is discovered that the user may have engaged in similar behavior in the past. The information is immediately reported to the relevant authorities, and measures are taken against the user.

[0041] This system allows for the early detection of inappropriate information on the internet, enabling swift action.

[0042] The following describes the processing flow.

[0043] Step 1:

[0044] The server collects data from multiple sources. This may involve using a dedicated API to retrieve data, or directly analyzing web content using web scraping techniques. It regularly checks for updates and retrieves any new data immediately.

[0045] Step 2:

[0046] The server performs initial filtering. It uses natural language processing on the collected information to identify potentially problematic content based on keyword lists. Following basic rules, it excludes harmless information and extracts content that requires further analysis.

[0047] Step 3:

[0048] The server performs sentiment analysis on the extracted information. Using machine learning models, it classifies the sentiment of each piece of content as positive, negative, or neutral. It also performs contextual analysis to determine whether the information is threatening or potentially related to criminal activity.

[0049] Step 4:

[0050] The server profiles the senders of information that has passed the analysis and cross-checks them against the database. By comparing them with past behavioral patterns and known profiles, it assesses the likelihood and risk of recidivism. At this stage, it checks for any similar behavioral history.

[0051] Step 5:

[0052] The server reports any information deemed problematic to the relevant authorities. This report includes details of the identified information content and its origin, as well as specific evidence. This action prepares the police and other relevant agencies for a swift response.

[0053] Step 6:

[0054] The server will send warning messages to users as needed. Users who make inappropriate posts will be notified of violations of laws and regulations and encouraged to correct their behavior. In addition, the server will consider temporarily restricting access from the platform for malicious users.

[0055] (Example 1)

[0056] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0057] There is a need to quickly and accurately identify problematic information such as defamation, threats of crime, and misinformation from the vast amount of data on data communication networks, so that relevant agencies can respond immediately. This requires technological means to carry out the entire process, from data collection to analysis and notification, efficiently and effectively.

[0058] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0059] In this invention, the server includes means for acquiring information from multiple information sources on a data communication network, means for initial filtering of the acquired information using natural language processing technology, and means for performing rapid data analysis using a generative AI model. This makes it possible to accurately identify problematic information and efficiently notify relevant organizations.

[0060] A "data communication network" is a network infrastructure used to send and receive information between computers and digital devices.

[0061] An "information source" refers to the origin of specific data or information, including websites and social media platforms.

[0062] "Natural language processing technology" is a field of technology that uses computers to understand, interpret, and generate human language.

[0063] "Initial filtering" is the first stage of processing in which collected information is screened out based on inappropriate keyword lists or rules.

[0064] A "generative AI model" is an artificial intelligence algorithm that learns from large datasets and possesses generational and estimation capabilities.

[0065] "Sentiment analysis" is an analytical technique that classifies the emotions embedded in text data into positive, negative, neutral, and so on.

[0066] "Characteristic analysis" is the process of clarifying individual characteristics and behavioral patterns based on data and information.

[0067] "Contextual analysis" is a technique that analyzes text or conversations by considering the surrounding sentences and context in order to understand their meaning.

[0068] "Notification" is the act of informing other devices or organizations of specific information or results.

[0069] The server plays a central role in acquiring and processing vast amounts of information via the data communication network. Specifically, it utilizes high-performance processor-equipped server machines and cloud computing as hardware, and frameworks and libraries for implementing natural language processing technology (e.g., Python's NLTK and SpaCy) as software. Furthermore, advanced data analysis capabilities are realized by employing generative AI models such as TENSORFLOW® and PyTorch.

[0070] The server first automatically retrieves the latest information from websites and social media using APIs and web scraping techniques. Next, the retrieved information is initially filtered using natural language processing techniques, and problematic information is selected based on inappropriate keywords. By utilizing generative AI models, the speed and accuracy of this data analysis are improved.

[0071] Furthermore, sentiment analysis is performed using machine learning models, categorizing emotions into positive, negative, and neutral. In addition, contextual analysis is conducted to identify potential risk factors. Subsequently, the sender's behavioral patterns are established through trait analysis, and the likelihood of recidivism is evaluated by comparing them with past databases.

[0072] If problematic information or its source is identified, the server will promptly notify the relevant authorities and provide data input to resolve the issue. The notification process will include specific evidence data to enable the relevant authorities to take appropriate action immediately.

[0073] For example, if a user posts a threatening message containing specific keywords on social media, this message is automatically collected and analyzed by the server. As a result, the user's post is identified as problematic, and a notification is sent to the relevant authorities.

[0074] An example of a prompt message would be, "We want to analyze posts on social media that contain specific keywords and assess the likelihood of negative emotions and threats."

[0075] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0076] Step 1:

[0077] The server automatically retrieves information from multiple sources on the data network using APIs and web scraping techniques. The input is publicly available information from websites and social media, and the output is a structured initial dataset, which is stored for subsequent processing.

[0078] Step 2:

[0079] The server applies natural language processing techniques to the initial dataset and performs initial filtering. The input is a structured initial dataset, and the data is scanned using a list of inappropriate keywords. The output is a filtered dataset containing content that is likely to be problematic. This process removes obviously inappropriate data.

[0080] Step 3:

[0081] The server utilizes a generative AI model to perform sentiment analysis on a filtered dataset. The input is the filtered dataset, and the sentiment of the text is analyzed using a specific algorithm. The output is a dataset with sentiment labels, where each data point is assigned either positive, negative, or neutral. These labels play a crucial role in the analysis in the next step.

[0082] Step 4:

[0083] The server performs contextual analysis to understand the meaning and intent of each data point in the filtered dataset. The input is a dataset with sentiment labels, and the server analyzes the context and surrounding text to reveal its intent. The output is the analyzed contextual information, which provides material for potentially determining the threatening or criminal nature of the data.

[0084] Step 5:

[0085] The server performs characteristic analysis and matches the behavioral patterns of each data sender. The input consists of contextually analyzed data and sender information, which is then cross-referenced with the database to verify past behavioral history. The output is characteristic data of the behavioral patterns, which is used to assess the likelihood of recidivism and specific risks.

[0086] Step 6:

[0087] The server promptly notifies relevant organizations of information and senders identified as problematic. Input consists of characteristic data and analysis results, which generate a notification message. Output is a notification containing specific informational evidence sent to the relevant organizations. This notification enables a rapid response.

[0088] (Application Example 1)

[0089] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0090] The internet contains a large amount of inappropriate content, including defamation, threats of violence, and misinformation, and it is crucial to address this. However, there is a lack of effective methods to monitor information in real time and quickly warn users, which creates a risk of users accessing inappropriate content. An efficient system is needed to solve this problem.

[0091] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0092] In this invention, the server includes means for collecting information from multiple information sources on the Internet, means for initial filtering the collected information using natural language processing technology, means for performing detailed sentiment analysis on the initially filtered information, means for profiling the analyzed information and cross-checking it with past databases, and means for monitoring information accessed by users in real time and issuing warnings for inappropriate content. This reduces the risk of accessing inappropriate content and enables the provision of a safe online environment.

[0093] "Methods for collecting information from multiple sources on the internet" refers to the process of obtaining information from websites, social media, etc., and the technology of centrally collecting this information to use as basic data for analysis.

[0094] "Methods for initial filtering using natural language processing technology" refer to techniques that analyze collected text data and identify inappropriate information using specific keywords or patterns.

[0095] "Methods for conducting detailed sentiment analysis" refer to techniques that analyze the content of collected texts and determine and classify the emotional tone within that context.

[0096] "Profiling and cross-checking with past databases" refers to a technology that compares information or senders deemed inappropriate with past records to evaluate their relevance and the likelihood of recurrence.

[0097] "A means of monitoring user-accessed information in real time and issuing warnings for inappropriate content" refers to a technology that instantly analyzes information viewed or interacted with by users and notifies them of potential risks by issuing warnings if it is deemed inappropriate.

[0098] The system that implements this invention primarily operates on a server. The server retrieves data from multiple sources on the internet to collect and analyze information. APIs and web scraping techniques are used for data retrieval. The collected information is initially filtered using programming languages ​​such as Python and natural language processing libraries (e.g., NLTK and spaCy).

[0099] The server then uses sentiment analysis tools such as TextBlob and SentimentIntensityAnalyzer to perform sentiment analysis on the acquired information. At this stage, it evaluates whether the information is positive, negative, or neutral. The server also analyzes the context of the information and performs profiling and cross-checks with historical databases to determine potential threats.

[0100] Furthermore, applications installed on the user's device monitor website content in real time using JavaScript® and Python scripts. If inappropriate content is detected through this monitoring, the user is immediately notified with a warning.

[0101] For example, if a user is browsing a shopping site and encounters fraudulent reviews or false information, a warning is sent to the user based on the server's analysis results. This allows users to continue their online activities with peace of mind.

[0102] An example of a prompt statement to input into a generative AI model is: "Design an app that automatically extracts specific content from social media and issues a warning when it is detected as inappropriate content. Required functions are keyword filtering and sentiment analysis." This prompt statement clearly instructs the AI ​​model on the purpose and operation of the invention.

[0103] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0104] Step 1:

[0105] The server begins the process of collecting data from information sources on the internet. It uses a list of APIs and URLs as input and employs web scraping techniques to retrieve text data from each source. The output is in the form of raw text files or stream data.

[0106] Step 2:

[0107] The server performs initial filtering of the collected text data using natural language processing techniques. The input is the text data collected in step 1. It searches for inappropriate keywords within the data and selects the relevant content. The output is the filtered text data, which may be a problematic dataset.

[0108] Step 3:

[0109] The server performs sentiment analysis on filtered data. The input is filtered data. Using TextBlob and SentimentIntensityAnalyzer, it analyzes the emotional characteristics of each text and classifies them as positive, negative, or neutral. The output is data with the resulting sentiment evaluations attached.

[0110] Step 4:

[0111] The server profiles the sentiment analysis results and cross-references them with past database data. The input is the data after sentiment analysis. Based on the database, it checks the likelihood of recidivism and past patterns to create a profile. The output is the profiled analysis data.

[0112] Step 5:

[0113] The device monitors user access information in real time and issues a warning when inappropriate content is detected. Input is access information from the user's browser or applications. Real-time analysis is performed, and if a problem is found, a warning message is sent to the user. Output is the warning message or notification status.

[0114] Step 6:

[0115] Users ensure safe online behavior based on the warning messages they receive. The input is warning information from the device. Upon receiving a warning about inappropriate content, users either refrain from accessing the content or conduct further investigation. The output represents safe usage.

[0116] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0117] This invention is a system that combines an emotion engine with information gathering and analysis from the internet, enabling the identification and rapid response to inappropriate information, including defamation and threats of crime. This system operates primarily server-centric and achieves its objectives through a multi-layered processing flow.

[0118] The server first automatically collects data from multiple sources on the internet. During this process, it periodically retrieves new information using APIs and web scraping, incorporating the latest data into the system. The collected information undergoes initial filtering within the server using natural language processing techniques. Here, potentially inappropriate content is selected based on a pre-configured keyword list.

[0119] Next, the server performs a detailed sentiment analysis. At this stage, the sentiment engine is used to identify the user's emotions from the text data. The sentiment engine evaluates emotions in a multidimensional manner, such as positive, negative, and neutral, and analyzes the reliability and risk of the information. Furthermore, it profiles the sender's emotional tendencies and assesses the sender's risk by matching similarities with past behavioral patterns.

[0120] For example, if a user posts offensive content on a message board, the server immediately detects the post. Using natural language processing technology and a sentiment engine, the post is determined to be negative and threatening. If profiling reveals that the post matches a sender with a history of similar problematic behavior, the information is immediately reported to the relevant authorities. This report includes specific evidence and detailed information about the sender, helping the authorities take effective action.

[0121] In this way, the system can efficiently detect inappropriate information online and promote social safety. The server improves its detection accuracy and operates more effectively by receiving feedback from the emotion engine.

[0122] The following describes the processing flow.

[0123] Step 1:

[0124] The server collects data from internet sources. It uses APIs and web scraping techniques to automatically collect data regularly from message boards, social media, and news sites. New posts and updates are stored in the database whenever they are detected.

[0125] Step 2:

[0126] The server performs initial filtering of the collected information using natural language processing techniques. Keyword matching algorithms are used to identify posts containing potentially inappropriate words or phrases. Harmless information is excluded, and only content requiring further analysis is sent to the next process.

[0127] Step 3:

[0128] The server uses a sentiment engine on information that has passed the initial filtering. It recognizes positive, negative, or neutral sentiment from the text data and calculates a risk level for posts with particularly strong negative sentiment.

[0129] Step 4:

[0130] The server uses profiling techniques to analyze the sender's emotional tendencies and past behavior. It compares this with past posting history and criminal profiles in the database to perform a risk assessment. Particular attention is paid when behavioral patterns match.

[0131] Step 5:

[0132] The server reports information deemed to contain strong negative emotions to the relevant authorities, along with supporting data. This report includes the content of the post, the poster's profile, and other relevant evidence. This reporting allows the relevant authorities to take prompt action.

[0133] Step 6:

[0134] The server will send warning messages to users as needed. It will automatically issue warnings and corrective instructions to users who post inappropriate content. In particularly serious cases, measures such as temporarily freezing the user's account will also be taken.

[0135] (Example 2)

[0136] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0137] In modern society, inappropriate information such as defamation and threats of crime is increasing online, and there is a need to quickly and accurately detect and deal with it. However, conventional methods require a great deal of effort and time to collect and analyze information, and there is also a risk of false positives, making it difficult to implement effective countermeasures.

[0138] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0139] In this invention, the server includes a device for collecting data from multiple information sources on the Internet, a device for filtering the collected data in an initial stage using natural language processing technology, and a device for performing a detailed sentiment evaluation of the initially filtered data using a sentiment engine. This makes it possible to quickly and accurately identify inappropriate information and report it to the appropriate authorities.

[0140] "On the Internet" refers to the space where information accessed via a global network exists.

[0141] "Information source" refers to the origin of information, such as a specific website or online platform, that provides or stores the information.

[0142] "Data" refers to information in various forms, such as text, images, and audio, obtained from an information source.

[0143] "Device" refers to a component of hardware or software designed to perform a specific function.

[0144] "Natural language processing technology" refers to computer-based technologies for understanding and analyzing human language.

[0145] "Filtering" refers to the process of selecting information based on relevance and reliability, and extracting important information.

[0146] An "emotion engine" refers to an algorithm or program that analyzes emotions from text data and evaluates emotional tendencies such as positive or negative.

[0147] "Evaluation" refers to the process of determining certain properties or values ​​based on data.

[0148] "Profiling" refers to the technique of analyzing the characteristics of individual users and understanding their behavioral patterns and tendencies.

[0149] "Cross-referencing" refers to the process of comparing different datasets to find commonalities and patterns.

[0150] "Reporting" refers to the act of informing relevant organizations of specific information.

[0151] This invention provides a system for quickly detecting and responding to inappropriate information on the internet. This system primarily operates server-centric and has the function of collecting and analyzing data from multiple sources on the internet and notifying relevant organizations as needed.

[0152] First, the server collects data from multiple sources such as social media, online forums, and blogs using APIs and web scraping techniques. This data is stored on the server in text format and undergoes initial filtering using natural language processing techniques. Specific software such as text analysis libraries (e.g., NLTK or SpaCy) may be used.

[0153] Next, the server performs a sentiment evaluation of the collected data through a sentiment engine. This sentiment engine determines positive, negative, and neutral sentiment scores, and data with particularly high negative scores is analyzed in detail. It is common to use sentiment analysis APIs (e.g., IBM Watson® or Google® Cloud Natural Language) during this process.

[0154] The server also profiles users and cross-references this with their past behavioral history. This makes it possible to assess the risk associated with senders who repeatedly engage in specific behaviors. A similarity analysis algorithm is applied using a database that stores sender behavioral patterns.

[0155] Inappropriate information identified in this manner is immediately reported by the server to the appropriate authorities. The report includes specific evidence and detailed information about the sender, providing the authorities with the basis for effective action.

[0156] For example, if a user posts a threatening comment on social media, this system analyzes the post using a sentiment engine, compares it to the poster's past behavior based on a negative score, and immediately initiates a reporting process.

[0157] An example of a prompt for a generative AI model would be: "Explain how to identify defamatory comments posted on a message board and gather the necessary information to report them."

[0158] The introduction of this system will enable the prevention of online threats and promote social safety.

[0159] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0160] Step 1:

[0161] The server collects data from multiple sources on the internet. The input is a list of URLs for sources accessed via APIs or web scraping. The server executes queries and retrieves text data from the sources based on specific keywords. The output is the collected raw dataset. This process involves periodic updates, and the collected data is stored in a database.

[0162] Step 2:

[0163] The server performs initial filtering of the collected data using natural language processing techniques. The input here is the raw dataset obtained in step 1. Specifically, a text analysis library is used to analyze each post against a specified keyword list and determine whether negative keywords are included. The output of this operation is the initially filtered dataset.

[0164] Step 3:

[0165] The server performs a detailed sentiment assessment of the initially filtered data using the sentiment engine. The input is the dataset filtered in step 2. The sentiment analysis API is used to calculate sentiment scores for the text data, determining positive, negative, and neutral scores. The output of this step is a dataset with the sentiment scores added.

[0166] Step 4:

[0167] The server performs user profiling and cross-references past behavioral history based on data with added sentiment scores. The input for this step is the profiling information for each user included in the output data from step 3. The server applies a similarity analysis algorithm to compare the sender's past behavior with current data. The output is a dataset with risk scores evaluated.

[0168] Step 5:

[0169] The server notifies the appropriate authorities of information where the risk score exceeds a certain threshold. The high-risk information identified in the output of Step 4 is used as input. The server uses a generative AI model to document the notification and sends it to the appropriate authorities via email, along with supporting data. The output of this process is a record of the notification.

[0170] (Application Example 2)

[0171] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0172] In modern society, inappropriate information and offensive content on the internet and social media pose a serious threat to individual psychological safety and overall social harmony. Addressing this problem requires real-time monitoring and analysis of relevant information, followed by a rapid response. However, current systems lack sufficient neural networking and reporting capabilities, making it impossible to completely eliminate the risk of inappropriate information spreading.

[0173] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0174] In this invention, the server includes means for collecting information from multiple information sources on the Internet, means for initial filtering of the collected information using natural language processing technology, and means for monitoring the content of the information providers being followed. This makes it possible to quickly detect and respond to inappropriate information.

[0175] The "Internet" is an information network that connects computers around the world to share and communicate data.

[0176] "Information source" refers to the place or medium from which information is collected, and includes websites and social media platforms.

[0177] "Natural language processing technology" is a technology that uses computers to process and understand human language.

[0178] "Filtering" is the process of selecting data from collected information based on specific criteria.

[0179] "Sentiment analysis" is a technique that analyzes emotional tendencies from text data and evaluates them as positive, negative, neutral, etc.

[0180] "Profiling" is the process of organizing and analyzing the characteristics and patterns of a subject based on specific data and behaviors.

[0181] A "database" is an organized collection of data used to efficiently store and retrieve information.

[0182] "Reporting" is the act of formally informing an external organization or person in charge of a matter.

[0183] "Information providers" refer to entities that disseminate information on the internet, and include individuals and companies.

[0184] "To monitor" means to constantly observe a specific subject carefully and keep track of its movements.

[0185] "Notification" refers to a means of transmitting specific information to a recipient, and includes the function of providing information in real time.

[0186] The system that implements this application is server-centric. The server is developed using the Python programming language to perform a series of processes including information gathering, analysis, and notification. The main software components include the BeautifulSoup library for web scraping, TensorFlow for sentiment analysis, and Twython, which supports the Twitter API for acquiring social media data.

[0187] First, the server automatically collects information from multiple sources on the internet. The data obtained here consists of social media posts and comments via APIs, and the collected data is organized into the required format using BeautifulSoup.

[0188] Next, the collected information is initially filtered using natural language processing techniques. This process filters out posts containing unusual keywords or specific phrases. The filtered information is then subjected to sentiment analysis using TensorFlow, classifying it as positive, negative, neutral, etc.

[0189] Furthermore, the server monitors posts from followed information providers (individuals and organizations) and, if inappropriate content is found, prepares to report it to the relevant authorities using profiling results. This report is accompanied by concrete evidence, ensuring the reliability of information sharing. In addition, users receive push notifications immediately upon detection of inappropriate content, prompting them to take necessary action.

[0190] For example, parents who want to ensure their children's online safety at home may use this system. It allows them to monitor their children's social media accounts and respond quickly if offensive content is posted.

[0191] An example of a prompt might be: "List the key steps to develop a system that performs real-time sentiment analysis on posts from a social media account and issues warning notifications when offensive content is found."

[0192] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0193] Step 1:

[0194] The server retrieves data via social media APIs. The input is account information of followed information providers, and the output is the latest posts collected from these accounts. This data is retrieved in JSON format.

[0195] Step 2:

[0196] The server parses the JSON data obtained using BeautifulSoup and extracts the necessary text information. The input is the JSON-formatted post data obtained in the previous step, and the output is the text extracted from the posts. This text is then formatted for later analysis.

[0197] Step 3:

[0198] The server uses natural language processing techniques to perform initial filtering of the extracted text. The input is the formatted post text, and the output is a list of posts containing specific keywords or phrases. This list is used to carefully analyze any unusual content.

[0199] Step 4:

[0200] The server uses TensorFlow to perform sentiment analysis on initially filtered post text. The input is the filtered post text, and the output is a sentiment score (positive, negative, neutral, etc.) for each post. The sentiment tendency of the posts is evaluated based on this score.

[0201] Step 5:

[0202] The server uses profiling functionality to cross-check the originator of sentiment analysis posts against a historical database. Inputs include sentiment scores and information about the poster, while output is data on the poster's behavioral tendencies. This data is used to assess the risk associated with the poster.

[0203] Step 6:

[0204] The server prepares to report posts deemed to be risky to the relevant authorities. The input is profiled risk information, and the output is a notification containing the evidence data necessary for reporting.

[0205] Step 7:

[0206] The server sends push notifications to users, warning them about detected inappropriate content. The input is the notification content ready for reporting, and the output is a real-time warning notification to the user. This allows users to immediately understand the situation and take appropriate action.

[0207] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating 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.

[0208] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0209] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0210] [Second Embodiment]

[0211] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0212] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0213] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0214] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0215] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0216] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0217] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0218] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0219] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0220] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0221] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0222] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. 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".

[0223] This invention is a system that automatically collects data from various sources on the internet and analyzes it to identify problematic content such as defamation, crime threats, and misinformation, and to respond to it quickly. The system is server-centric and capable of efficiently processing large amounts of data.

[0224] The server first collects information from multiple websites and social media platforms. This data collection may be done directly via APIs or through web scraping techniques. The collected information is then initially filtered within the server using natural language processing. This process identifies potentially problematic content based on a pre-configured list of inappropriate keywords.

[0225] Next, the information that passes the initial filtering undergoes further sentiment and contextual analysis by the server. At this stage, machine learning models are used to determine whether the sentiment of a post is positive, negative, or neutral. Furthermore, contextual information is analyzed to understand whether the content of the post is threatening or related to criminal activity.

[0226] Subsequently, the server uses the analysis results to profile the characteristics and past behavior of the inappropriate sender, and cross-checks them with existing databases. This process checks for similar behavioral patterns in the past and infers the likelihood of recidivism.

[0227] When the system identifies inappropriate information or a source, the server notifies the relevant authorities. This notification includes detailed information about the problematic post and its source, as well as supporting data. This allows the authorities to take swift action.

[0228] As a concrete example, consider a case where a user posts a threatening message to a specific individual on social media. This post is automatically collected by the server, and threatening keywords are detected during the initial filtering stage. Subsequently, sentiment analysis confirms the negativity of the post, and during the profiling stage, it is discovered that the user may have engaged in similar behavior in the past. The information is immediately reported to the relevant authorities, and measures are taken against the user.

[0229] This system allows for the early detection of inappropriate information on the internet, enabling swift action.

[0230] The following describes the processing flow.

[0231] Step 1:

[0232] The server collects data from multiple sources. This may involve using a dedicated API to retrieve data, or directly analyzing web content using web scraping techniques. It regularly checks for updates and retrieves any new data immediately.

[0233] Step 2:

[0234] The server performs initial filtering. It uses natural language processing on the collected information to identify potentially problematic content based on keyword lists. Following basic rules, it excludes harmless information and extracts content that requires further analysis.

[0235] Step 3:

[0236] The server performs sentiment analysis on the extracted information. Using machine learning models, it classifies the sentiment of each piece of content as positive, negative, or neutral. It also performs contextual analysis to determine whether the information is threatening or potentially related to criminal activity.

[0237] Step 4:

[0238] The server profiles the senders of information that has passed the analysis and cross-checks them against the database. By comparing them with past behavioral patterns and known profiles, it assesses the likelihood and risk of recidivism. At this stage, it checks for any similar behavioral history.

[0239] Step 5:

[0240] The server reports any information deemed problematic to the relevant authorities. This report includes details of the identified information content and its origin, as well as specific evidence. This action prepares the police and other relevant agencies for a swift response.

[0241] Step 6:

[0242] The server will send warning messages to users as needed. Users who make inappropriate posts will be notified of violations of laws and regulations and encouraged to correct their behavior. In addition, the server will consider temporarily restricting access from the platform for malicious users.

[0243] (Example 1)

[0244] Next, we will describe Example 1. 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."

[0245] There is a need to quickly and accurately identify problematic information such as defamation, threats of crime, and misinformation from the vast amount of data on data communication networks, so that relevant agencies can respond immediately. This requires technological means to carry out the entire process, from data collection to analysis and notification, efficiently and effectively.

[0246] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0247] In this invention, the server includes means for acquiring information from multiple information sources on a data communication network, means for initial filtering of the acquired information using natural language processing technology, and means for performing rapid data analysis using a generative AI model. This makes it possible to accurately identify problematic information and efficiently notify relevant organizations.

[0248] A "data communication network" is a network infrastructure used to send and receive information between computers and digital devices.

[0249] An "information source" refers to the origin of specific data or information, including websites and social media platforms.

[0250] "Natural language processing technology" is a field of technology that uses computers to understand, interpret, and generate human language.

[0251] "Initial filtering" is the first stage of processing in which collected information is screened out based on inappropriate keyword lists or rules.

[0252] A "generative AI model" is an artificial intelligence algorithm that learns from large datasets and possesses generational and estimation capabilities.

[0253] "Sentiment analysis" is an analytical technique that classifies the emotions embedded in text data into positive, negative, neutral, and so on.

[0254] "Characteristic analysis" is the process of clarifying individual characteristics and behavioral patterns based on data and information.

[0255] "Contextual analysis" is a technique that analyzes text or conversations by considering the surrounding sentences and context in order to understand their meaning.

[0256] "Notification" is the act of informing other devices or organizations of specific information or results.

[0257] The server plays a central role in acquiring and processing vast amounts of information via the data communication network. Specifically, it utilizes high-performance processor-equipped server machines and cloud computing as hardware, and frameworks and libraries for implementing natural language processing technologies (e.g., Python's NLTK and SpaCy) as software. Furthermore, advanced data analysis capabilities are realized by employing generative AI models such as TensorFlow and PyTorch.

[0258] The server first automatically retrieves the latest information from websites and social media using APIs and web scraping techniques. Next, the retrieved information is initially filtered using natural language processing techniques, and problematic information is selected based on inappropriate keywords. By utilizing generative AI models, the speed and accuracy of this data analysis are improved.

[0259] Furthermore, sentiment analysis is performed using machine learning models, categorizing emotions into positive, negative, and neutral. In addition, contextual analysis is conducted to identify potential risk factors. Subsequently, the sender's behavioral patterns are established through trait analysis, and the likelihood of recidivism is evaluated by comparing them with past databases.

[0260] If problematic information or its source is identified, the server will promptly notify the relevant authorities and provide data input to resolve the issue. The notification process will include specific evidence data to enable the relevant authorities to take appropriate action immediately.

[0261] For example, if a user posts a threatening message containing specific keywords on social media, this message is automatically collected and analyzed by the server. As a result, the user's post is identified as problematic, and a notification is sent to the relevant authorities.

[0262] An example of a prompt message would be, "We want to analyze posts on social media that contain specific keywords and assess the likelihood of negative emotions and threats."

[0263] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0264] Step 1:

[0265] The server automatically retrieves information from multiple sources on the data network using APIs and web scraping techniques. The input is publicly available information from websites and social media, and the output is a structured initial dataset, which is stored for subsequent processing.

[0266] Step 2:

[0267] The server applies natural language processing techniques to the initial dataset and performs initial filtering. The input is a structured initial dataset, and the data is scanned using a list of inappropriate keywords. The output is a filtered dataset containing content that is likely to be problematic. This process removes obviously inappropriate data.

[0268] Step 3:

[0269] The server utilizes a generative AI model to perform sentiment analysis on a filtered dataset. The input is the filtered dataset, and the sentiment of the text is analyzed using a specific algorithm. The output is a dataset with sentiment labels, where each data point is assigned either positive, negative, or neutral. These labels play a crucial role in the analysis in the next step.

[0270] Step 4:

[0271] The server performs contextual analysis to understand the meaning and intent of each data point in the filtered dataset. The input is a dataset with sentiment labels, and the server analyzes the context and surrounding text to reveal its intent. The output is the analyzed contextual information, which provides material for potentially determining the threatening or criminal nature of the data.

[0272] Step 5:

[0273] The server performs characteristic analysis and matches the behavioral patterns of each data sender. The input consists of contextually analyzed data and sender information, which is then cross-referenced with the database to verify past behavioral history. The output is characteristic data of the behavioral patterns, which is used to assess the likelihood of recidivism and specific risks.

[0274] Step 6:

[0275] The server promptly notifies relevant organizations of information and senders identified as problematic. Input consists of characteristic data and analysis results, which generate a notification message. Output is a notification containing specific informational evidence sent to the relevant organizations. This notification enables a rapid response.

[0276] (Application Example 1)

[0277] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0278] The internet contains a large amount of inappropriate content, including defamation, threats of violence, and misinformation, and it is crucial to address this. However, there is a lack of effective methods to monitor information in real time and quickly warn users, which creates a risk of users accessing inappropriate content. An efficient system is needed to solve this problem.

[0279] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0280] In this invention, the server includes means for collecting information from a plurality of information sources on the Internet, means for initially filtering the collected information by natural language processing technology, means for performing a detailed sentiment analysis on the initially filtered information, means for profiling the analyzed information and cross-checking it with a past database, and means for monitoring in real time the information accessed by a user and issuing a warning for inappropriate content. Thereby, the risk of accessing inappropriate content can be reduced, and a safe online environment can be provided.

[0281] The means for collecting information from a plurality of information sources on the Internet refers to the process of obtaining information from websites, social media, etc., and is a technology for collecting this information centrally to serve as basic data for analysis.

[0282] The means for initially filtering by natural language processing technology is a technique for analyzing the collected text data and identifying inappropriate information using specific keywords or patterns.

[0283] The means for performing a detailed sentiment analysis is a technology for analyzing the content of the collected text, judging and classifying the emotional tone in its context.

[0284] The means for profiling and cross-checking with a past database is a technology for collating the information or sender judged to be inappropriate with past records and evaluating the relevance and likelihood of recurrence.

[0285] The means for monitoring in real time the information accessed by a user and issuing a warning for inappropriate content is a technology for immediately analyzing the information browsed or operated by a user and notifying the user of potential risks by issuing a warning when it is judged to be inappropriate.

[0286] The system that realizes this invention mainly operates on a server. The server acquires data from multiple information sources on the Internet in order to collect and analyze information. For data acquisition, APIs and web scraping technologies are used. The collected information is initially filtered using programming languages such as Python and natural language processing libraries (e.g., NLTK and spaCy).

[0287] The server then performs sentiment analysis on the acquired information using sentiment analysis tools such as TextBlob and SentimentIntensityAnalyzer. At this stage, it evaluates whether the information corresponds to positive, negative, or neutral. Also, the server analyzes the context of the information and conducts profiling and cross-checking with the past database to determine potential threats.

[0288] Furthermore, the application installed on the user's terminal monitors the content of the website in real time using JavaScript and Python scripts. If inappropriate content is detected through this monitoring, a warning is immediately notified to the user.

[0289] As a specific example, when a user is browsing a shopping site and fraudulent reviews or false information are displayed, a warning is sent to the user based on the analysis results on the server. This enables the user to continue their online activities with confidence.

[0290] An example of a prompt sentence for input into the generative AI model is "Please design an application that automatically extracts specific content from SNS and issues a warning when it is detected as inappropriate content. The necessary functions are keyword filtering and sentiment analysis." This prompt sentence can clearly instruct the AI model about the purpose and operation of the invention.

[0291] The flow of the specific process in Application Example 1 will be described using FIG. 12.

[0292] Step 1:

[0293] The server begins the process of collecting data from information sources on the internet. It uses a list of APIs and URLs as input and employs web scraping techniques to retrieve text data from each source. The output is in the form of raw text files or stream data.

[0294] Step 2:

[0295] The server performs initial filtering of the collected text data using natural language processing techniques. The input is the text data collected in step 1. It searches for inappropriate keywords within the data and selects the relevant content. The output is the filtered text data, which may be a problematic dataset.

[0296] Step 3:

[0297] The server performs sentiment analysis on filtered data. The input is filtered data. Using TextBlob and SentimentIntensityAnalyzer, it analyzes the emotional characteristics of each text and classifies them as positive, negative, or neutral. The output is data with the resulting sentiment evaluations attached.

[0298] Step 4:

[0299] The server profiles the sentiment analysis results and cross-references them with past database data. The input is the data after sentiment analysis. Based on the database, it checks the likelihood of recidivism and past patterns to create a profile. The output is the profiled analysis data.

[0300] Step 5:

[0301] The device monitors user access information in real time and issues a warning when inappropriate content is detected. Input is access information from the user's browser or applications. Real-time analysis is performed, and if a problem is found, a warning message is sent to the user. Output is the warning message or notification status.

[0302] Step 6:

[0303] Users ensure safe online behavior based on the warning messages they receive. The input is warning information from the device. Upon receiving a warning about inappropriate content, users either refrain from accessing the content or conduct further investigation. The output represents safe usage.

[0304] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0305] This invention is a system that combines an emotion engine with information gathering and analysis from the internet, enabling the identification and rapid response to inappropriate information, including defamation and threats of crime. This system operates primarily server-centric and achieves its objectives through a multi-layered processing flow.

[0306] The server first automatically collects data from multiple sources on the internet. During this process, it periodically retrieves new information using APIs and web scraping, incorporating the latest data into the system. The collected information undergoes initial filtering within the server using natural language processing techniques. Here, potentially inappropriate content is selected based on a pre-configured keyword list.

[0307] Subsequently, the server performs a detailed sentiment analysis. At this stage, the sentiment engine is used to identify the user's sentiment from the text data. The sentiment engine evaluates sentiments such as positive, negative, and neutral in a multi-dimensional manner, and analyzes the reliability and risk of the information. Furthermore, by profiling the emotional tendency of the sender and comparing the similarity with past behavior patterns, the risk of the sender is evaluated.

[0308] As a specific example, when a user posts an offensive message on a bulletin board, the server immediately detects the post. Using natural language processing technology and the sentiment engine, it is determined that the post is negative and threatening. If profiling reveals a match with a sender who has had similar problematic behavior in the past, the information is immediately reported to the relevant authorities. This report includes specific evidentiary data and the sender's detailed information, which helps the relevant authorities take effective measures.

[0309] In this way, the system can efficiently detect inappropriate information on the network and promote social safety. The server improves detection accuracy while receiving feedback from the sentiment engine and is operated more effectively.

[0310] The following explains the processing flow.

[0311] Step 1:

[0312] The server collects data from information sources on the Internet. Using APIs and web scraping technology, data is automatically collected regularly from bulletin boards, social media, and news sites. It is stored in the database every time new posts or updated information are confirmed.

[0313] Step 2:

[0314] The server performs initial filtering of the collected information using natural language processing techniques. Keyword matching algorithms are used to identify posts containing potentially inappropriate words or phrases. Harmless information is excluded, and only content requiring further analysis is sent to the next process.

[0315] Step 3:

[0316] The server uses a sentiment engine on information that has passed the initial filtering. It recognizes positive, negative, or neutral sentiment from the text data and calculates a risk level for posts with particularly strong negative sentiment.

[0317] Step 4:

[0318] The server uses profiling techniques to analyze the sender's emotional tendencies and past behavior. It compares this with past posting history and criminal profiles in the database to perform a risk assessment. Particular attention is paid when behavioral patterns match.

[0319] Step 5:

[0320] The server reports information deemed to contain strong negative emotions to the relevant authorities, along with supporting data. This report includes the content of the post, the poster's profile, and other relevant evidence. This reporting allows the relevant authorities to take prompt action.

[0321] Step 6:

[0322] The server will send warning messages to users as needed. It will automatically issue warnings and corrective instructions to users who post inappropriate content. In particularly serious cases, measures such as temporarily freezing the user's account will also be taken.

[0323] (Example 2)

[0324] Next, we will describe Example 2. 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".

[0325] In modern society, inappropriate information such as defamation and threats of crime is increasing online, and there is a need to quickly and accurately detect and deal with it. However, conventional methods require a great deal of effort and time to collect and analyze information, and there is also a risk of false positives, making it difficult to implement effective countermeasures.

[0326] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0327] In this invention, the server includes a device for collecting data from multiple information sources on the Internet, a device for filtering the collected data in an initial stage using natural language processing technology, and a device for performing a detailed sentiment evaluation of the initially filtered data using a sentiment engine. This makes it possible to quickly and accurately identify inappropriate information and report it to the appropriate authorities.

[0328] "On the Internet" refers to the space where information accessed via a global network exists.

[0329] "Information source" refers to the origin of information, such as a specific website or online platform, that provides or stores the information.

[0330] "Data" refers to information in various forms, such as text, images, and audio, obtained from an information source.

[0331] "Device" refers to a component of hardware or software designed to perform a specific function.

[0332] "Natural language processing technology" refers to computer-based technologies for understanding and analyzing human language.

[0333] "Filtering" refers to the process of selecting information based on relevance and reliability, and extracting important information.

[0334] An "emotion engine" refers to an algorithm or program that analyzes emotions from text data and evaluates emotional tendencies such as positive or negative.

[0335] "Evaluation" refers to the process of determining certain properties or values ​​based on data.

[0336] "Profiling" refers to the technique of analyzing the characteristics of individual users and understanding their behavioral patterns and tendencies.

[0337] "Cross-referencing" refers to the process of comparing different datasets to find commonalities and patterns.

[0338] "Reporting" refers to the act of informing relevant organizations of specific information.

[0339] This invention provides a system for quickly detecting and responding to inappropriate information on the internet. This system primarily operates server-centric and has the function of collecting and analyzing data from multiple sources on the internet and notifying relevant organizations as needed.

[0340] First, the server collects data from multiple sources such as social media, online forums, and blogs using APIs and web scraping techniques. This data is stored on the server in text format and undergoes initial filtering using natural language processing techniques. Specific software such as text analysis libraries (e.g., NLTK or SpaCy) may be used.

[0341] Next, the server performs a sentiment evaluation of the collected data through a sentiment engine. This sentiment engine determines positive, negative, and neutral sentiment scores, and data with particularly high negative scores are analyzed in detail. It is common to use sentiment analysis APIs (e.g., IBM Watson or Google Cloud Natural Language) during this process.

[0342] The server also profiles users and cross-references this with their past behavioral history. This makes it possible to assess the risk associated with senders who repeatedly engage in specific behaviors. A similarity analysis algorithm is applied using a database that stores sender behavioral patterns.

[0343] Inappropriate information identified in this manner is immediately reported by the server to the appropriate authorities. The report includes specific evidence and detailed information about the sender, providing the authorities with the basis for effective action.

[0344] For example, if a user posts a threatening comment on social media, this system analyzes the post using a sentiment engine, compares it to the poster's past behavior based on a negative score, and immediately initiates a reporting process.

[0345] An example of a prompt for a generative AI model would be: "Explain how to identify defamatory comments posted on a message board and gather the necessary information to report them."

[0346] The introduction of this system will enable the prevention of online threats and promote social safety.

[0347] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0348] Step 1:

[0349] The server collects data from multiple sources on the internet. The input is a list of URLs for sources accessed via APIs or web scraping. The server executes queries and retrieves text data from the sources based on specific keywords. The output is the collected raw dataset. This process involves periodic updates, and the collected data is stored in a database.

[0350] Step 2:

[0351] The server performs initial filtering of the collected data using natural language processing techniques. The input here is the raw dataset obtained in step 1. Specifically, a text analysis library is used to analyze each post against a specified keyword list and determine whether negative keywords are included. The output of this operation is the initially filtered dataset.

[0352] Step 3:

[0353] The server performs a detailed sentiment assessment of the initially filtered data using the sentiment engine. The input is the dataset filtered in step 2. The sentiment analysis API is used to calculate sentiment scores for the text data, determining positive, negative, and neutral scores. The output of this step is a dataset with the sentiment scores added.

[0354] Step 4:

[0355] The server performs user profiling and cross-references past behavioral history based on data with added sentiment scores. The input for this step is the profiling information for each user included in the output data from step 3. The server applies a similarity analysis algorithm to compare the sender's past behavior with current data. The output is a dataset with risk scores evaluated.

[0356] Step 5:

[0357] The server notifies the appropriate authorities of information where the risk score exceeds a certain threshold. The high-risk information identified in the output of Step 4 is used as input. The server uses a generative AI model to document the notification and sends it to the appropriate authorities via email, along with supporting data. The output of this process is a record of the notification.

[0358] (Application Example 2)

[0359] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0360] In modern society, inappropriate information and offensive content on the internet and social media pose a serious threat to individual psychological safety and overall social harmony. Addressing this problem requires real-time monitoring and analysis of relevant information, followed by a rapid response. However, current systems lack sufficient neural networking and reporting capabilities, making it impossible to completely eliminate the risk of inappropriate information spreading.

[0361] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0362] In this invention, the server includes means for collecting information from multiple information sources on the Internet, means for initial filtering of the collected information using natural language processing technology, and means for monitoring the content of the information providers being followed. This makes it possible to quickly detect and respond to inappropriate information.

[0363] The "Internet" is an information network that connects computers around the world to share and communicate data.

[0364] "Information source" refers to the place or medium from which information is collected, and includes websites and social media platforms.

[0365] "Natural language processing technology" is a technology that uses computers to process and understand human language.

[0366] "Filtering" is the process of selecting data from collected information based on specific criteria.

[0367] "Sentiment analysis" is a technique that analyzes emotional tendencies from text data and evaluates them as positive, negative, neutral, etc.

[0368] "Profiling" is the process of organizing and analyzing the characteristics and patterns of a subject based on specific data and behaviors.

[0369] A "database" is an organized collection of data used to efficiently store and retrieve information.

[0370] "Reporting" is the act of formally informing an external organization or person in charge of a matter.

[0371] "Information providers" refer to entities that disseminate information on the internet, and include individuals and companies.

[0372] "To monitor" means to constantly observe a specific subject carefully and keep track of its movements.

[0373] "Notification" refers to a means of transmitting specific information to a recipient, and includes the function of providing information in real time.

[0374] The system that implements this application is server-centric. The server is developed using the Python programming language to perform a series of processes including information gathering, analysis, and notification. The main software components include the BeautifulSoup library for web scraping, TensorFlow for sentiment analysis, and Twython, which supports the Twitter API for acquiring social media data.

[0375] First, the server automatically collects information from multiple sources on the internet. The data obtained here consists of social media posts and comments via APIs, and the collected data is organized into the required format using BeautifulSoup.

[0376] Next, the collected information is initially filtered using natural language processing techniques. This process filters out posts containing unusual keywords or specific phrases. The filtered information is then subjected to sentiment analysis using TensorFlow, classifying it as positive, negative, neutral, etc.

[0377] Furthermore, the server monitors posts from followed information providers (individuals and organizations) and, if inappropriate content is found, prepares to report it to the relevant authorities using profiling results. This report is accompanied by concrete evidence, ensuring the reliability of information sharing. In addition, users receive push notifications immediately upon detection of inappropriate content, prompting them to take necessary action.

[0378] For example, parents who want to ensure their children's online safety at home may use this system. It allows them to monitor their children's social media accounts and respond quickly if offensive content is posted.

[0379] An example of a prompt might be: "List the key steps to develop a system that performs real-time sentiment analysis on posts from a social media account and issues warning notifications when offensive content is found."

[0380] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0381] Step 1:

[0382] The server retrieves data via social media APIs. The input is account information of followed information providers, and the output is the latest posts collected from these accounts. This data is retrieved in JSON format.

[0383] Step 2:

[0384] The server parses the JSON data obtained using BeautifulSoup and extracts the necessary text information. The input is the JSON-formatted post data obtained in the previous step, and the output is the text extracted from the posts. This text is then formatted for later analysis.

[0385] Step 3:

[0386] The server uses natural language processing techniques to perform initial filtering of the extracted text. The input is the formatted post text, and the output is a list of posts containing specific keywords or phrases. This list is used to carefully analyze any unusual content.

[0387] Step 4:

[0388] The server uses TensorFlow to perform sentiment analysis on initially filtered post text. The input is the filtered post text, and the output is a sentiment score (positive, negative, neutral, etc.) for each post. The sentiment tendency of the posts is evaluated based on this score.

[0389] Step 5:

[0390] The server uses profiling functionality to cross-check the originator of sentiment analysis posts against a historical database. Inputs include sentiment scores and information about the poster, while output is data on the poster's behavioral tendencies. This data is used to assess the risk associated with the poster.

[0391] Step 6:

[0392] The server prepares to report posts deemed to be risky to the relevant authorities. The input is profiled risk information, and the output is a notification containing the evidence data necessary for reporting.

[0393] Step 7:

[0394] The server sends push notifications to users, warning them about detected inappropriate content. The input is the notification content ready for reporting, and the output is a real-time warning notification to the user. This allows users to immediately understand the situation and take appropriate action.

[0395] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0396] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0397] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0398] [Third Embodiment]

[0399] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0400] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0401] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0402] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0403] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0404] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0405] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0406] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0407] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0408] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0409] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0410] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0411] This invention is a system that automatically collects data from various sources on the internet and analyzes it to identify problematic content such as defamation, crime threats, and misinformation, and to respond to it quickly. The system is server-centric and capable of efficiently processing large amounts of data.

[0412] The server first collects information from multiple websites and social media platforms. This data collection may be done directly via APIs or through web scraping techniques. The collected information is then initially filtered within the server using natural language processing. This process identifies potentially problematic content based on a pre-configured list of inappropriate keywords.

[0413] Next, the information that passes the initial filtering undergoes further sentiment and contextual analysis by the server. At this stage, machine learning models are used to determine whether the sentiment of a post is positive, negative, or neutral. Furthermore, contextual information is analyzed to understand whether the content of the post is threatening or related to criminal activity.

[0414] Subsequently, the server uses the analysis results to profile the characteristics and past behavior of the inappropriate sender, and cross-checks them with existing databases. This process checks for similar behavioral patterns in the past and infers the likelihood of recidivism.

[0415] When the system identifies inappropriate information or a source, the server notifies the relevant authorities. This notification includes detailed information about the problematic post and its source, as well as supporting data. This allows the authorities to take swift action.

[0416] As a concrete example, consider a case where a user posts a threatening message to a specific individual on social media. This post is automatically collected by the server, and threatening keywords are detected during the initial filtering stage. Subsequently, sentiment analysis confirms the negativity of the post, and during the profiling stage, it is discovered that the user may have engaged in similar behavior in the past. The information is immediately reported to the relevant authorities, and measures are taken against the user.

[0417] This system allows for the early detection of inappropriate information on the internet, enabling swift action.

[0418] The following describes the processing flow.

[0419] Step 1:

[0420] The server collects data from multiple sources. This may involve using a dedicated API to retrieve data, or directly analyzing web content using web scraping techniques. It regularly checks for updates and retrieves any new data immediately.

[0421] Step 2:

[0422] The server performs initial filtering. It uses natural language processing on the collected information to identify potentially problematic content based on keyword lists. Following basic rules, it excludes harmless information and extracts content that requires further analysis.

[0423] Step 3:

[0424] The server performs sentiment analysis on the extracted information. Using machine learning models, it classifies the sentiment of each piece of content as positive, negative, or neutral. It also performs contextual analysis to determine whether the information is threatening or potentially related to criminal activity.

[0425] Step 4:

[0426] The server profiles the senders of information that has passed the analysis and cross-checks them against the database. By comparing them with past behavioral patterns and known profiles, it assesses the likelihood and risk of recidivism. At this stage, it checks for any similar behavioral history.

[0427] Step 5:

[0428] The server reports any information deemed problematic to the relevant authorities. This report includes details of the identified information content and its origin, as well as specific evidence. This action prepares the police and other relevant agencies for a swift response.

[0429] Step 6:

[0430] The server will send warning messages to users as needed. Users who make inappropriate posts will be notified of violations of laws and regulations and encouraged to correct their behavior. In addition, the server will consider temporarily restricting access from the platform for malicious users.

[0431] (Example 1)

[0432] Next, we will describe Example 1. 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."

[0433] There is a need to quickly and accurately identify problematic information such as defamation, threats of crime, and misinformation from the vast amount of data on data communication networks, so that relevant agencies can respond immediately. This requires technological means to carry out the entire process, from data collection to analysis and notification, efficiently and effectively.

[0434] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0435] In this invention, the server includes means for acquiring information from multiple information sources on a data communication network, means for initial filtering of the acquired information using natural language processing technology, and means for performing rapid data analysis using a generative AI model. This makes it possible to accurately identify problematic information and efficiently notify relevant organizations.

[0436] A "data communication network" is a network infrastructure used to send and receive information between computers and digital devices.

[0437] An "information source" refers to the origin of specific data or information, including websites and social media platforms.

[0438] "Natural language processing technology" is a field of technology that uses computers to understand, interpret, and generate human language.

[0439] "Initial filtering" is the first stage of processing in which collected information is screened out based on inappropriate keyword lists or rules.

[0440] A "generative AI model" is an artificial intelligence algorithm that learns from large datasets and possesses generational and estimation capabilities.

[0441] "Sentiment analysis" is an analytical technique that classifies the emotions embedded in text data into positive, negative, neutral, and so on.

[0442] "Characteristic analysis" is the process of clarifying individual characteristics and behavioral patterns based on data and information.

[0443] "Contextual analysis" is a technique that analyzes text or conversations by considering the surrounding sentences and context in order to understand their meaning.

[0444] "Notification" is the act of informing other devices or organizations of specific information or results.

[0445] The server plays a central role in acquiring and processing vast amounts of information via the data communication network. Specifically, it utilizes high-performance processor-equipped server machines and cloud computing as hardware, and frameworks and libraries for implementing natural language processing technologies (e.g., Python's NLTK and SpaCy) as software. Furthermore, advanced data analysis capabilities are realized by employing generative AI models such as TensorFlow and PyTorch.

[0446] The server first automatically retrieves the latest information from websites and social media using APIs and web scraping techniques. Next, the retrieved information is initially filtered using natural language processing techniques, and problematic information is selected based on inappropriate keywords. By utilizing generative AI models, the speed and accuracy of this data analysis are improved.

[0447] Furthermore, sentiment analysis is performed using machine learning models, categorizing emotions into positive, negative, and neutral. In addition, contextual analysis is conducted to identify potential risk factors. Subsequently, the sender's behavioral patterns are established through trait analysis, and the likelihood of recidivism is evaluated by comparing them with past databases.

[0448] If problematic information or its source is identified, the server will promptly notify the relevant authorities and provide data input to resolve the issue. The notification process will include specific evidence data to enable the relevant authorities to take appropriate action immediately.

[0449] For example, if a user posts a threatening message containing specific keywords on social media, this message is automatically collected and analyzed by the server. As a result, the user's post is identified as problematic, and a notification is sent to the relevant authorities.

[0450] An example of a prompt message would be, "We want to analyze posts on social media that contain specific keywords and assess the likelihood of negative emotions and threats."

[0451] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0452] Step 1:

[0453] The server automatically retrieves information from multiple sources on the data network using APIs and web scraping techniques. The input is publicly available information from websites and social media, and the output is a structured initial dataset, which is stored for subsequent processing.

[0454] Step 2:

[0455] The server applies natural language processing techniques to the initial dataset and performs initial filtering. The input is a structured initial dataset, and the data is scanned using a list of inappropriate keywords. The output is a filtered dataset containing content that is likely to be problematic. This process removes obviously inappropriate data.

[0456] Step 3:

[0457] The server utilizes a generative AI model to perform sentiment analysis on a filtered dataset. The input is the filtered dataset, and the sentiment of the text is analyzed using a specific algorithm. The output is a dataset with sentiment labels, where each data point is assigned either positive, negative, or neutral. These labels play a crucial role in the analysis in the next step.

[0458] Step 4:

[0459] The server performs contextual analysis to understand the meaning and intent of each data point in the filtered dataset. The input is a dataset with sentiment labels, and the server analyzes the context and surrounding text to reveal its intent. The output is the analyzed contextual information, which provides material for potentially determining the threatening or criminal nature of the data.

[0460] Step 5:

[0461] The server performs characteristic analysis and matches the behavioral patterns of each data sender. The input consists of contextually analyzed data and sender information, which is then cross-referenced with the database to verify past behavioral history. The output is characteristic data of the behavioral patterns, which is used to assess the likelihood of recidivism and specific risks.

[0462] Step 6:

[0463] The server promptly notifies relevant organizations of information and senders identified as problematic. Input consists of characteristic data and analysis results, which generate a notification message. Output is a notification containing specific informational evidence sent to the relevant organizations. This notification enables a rapid response.

[0464] (Application Example 1)

[0465] Next, we will explain Application Example 1. In the following explanation, 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."

[0466] The internet contains a large amount of inappropriate content, including defamation, threats of violence, and misinformation, and it is crucial to address this. However, there is a lack of effective methods to monitor information in real time and quickly warn users, which creates a risk of users accessing inappropriate content. An efficient system is needed to solve this problem.

[0467] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0468] In this invention, the server includes means for collecting information from multiple information sources on the Internet, means for initial filtering the collected information using natural language processing technology, means for performing detailed sentiment analysis on the initially filtered information, means for profiling the analyzed information and cross-checking it with past databases, and means for monitoring information accessed by users in real time and issuing warnings for inappropriate content. This reduces the risk of accessing inappropriate content and enables the provision of a safe online environment.

[0469] "Methods for collecting information from multiple sources on the internet" refers to the process of obtaining information from websites, social media, etc., and the technology of centrally collecting this information to use as basic data for analysis.

[0470] "Methods for initial filtering using natural language processing technology" refer to techniques that analyze collected text data and identify inappropriate information using specific keywords or patterns.

[0471] "Methods for conducting detailed sentiment analysis" refer to techniques that analyze the content of collected texts and determine and classify the emotional tone within that context.

[0472] "Profiling and cross-checking with past databases" refers to a technology that compares information or senders deemed inappropriate with past records to evaluate their relevance and the likelihood of recurrence.

[0473] "A means of monitoring user-accessed information in real time and issuing warnings for inappropriate content" refers to a technology that instantly analyzes information viewed or interacted with by users and notifies them of potential risks by issuing warnings if it is deemed inappropriate.

[0474] The system that implements this invention primarily operates on a server. The server retrieves data from multiple sources on the internet to collect and analyze information. APIs and web scraping techniques are used for data retrieval. The collected information is initially filtered using programming languages ​​such as Python and natural language processing libraries (e.g., NLTK and spaCy).

[0475] The server then uses sentiment analysis tools such as TextBlob and SentimentIntensityAnalyzer to perform sentiment analysis on the acquired information. At this stage, it evaluates whether the information is positive, negative, or neutral. The server also analyzes the context of the information and performs profiling and cross-checks with historical databases to determine potential threats.

[0476] Furthermore, applications installed on the user's device monitor website content in real time using JavaScript and Python scripts. If inappropriate content is detected through this monitoring, the user is immediately notified with a warning.

[0477] For example, if a user is browsing a shopping site and encounters fraudulent reviews or false information, a warning is sent to the user based on the server's analysis results. This allows users to continue their online activities with peace of mind.

[0478] An example of a prompt statement to input into a generative AI model is: "Design an app that automatically extracts specific content from social media and issues a warning when it is detected as inappropriate content. Required functions are keyword filtering and sentiment analysis." This prompt statement clearly instructs the AI ​​model on the purpose and operation of the invention.

[0479] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0480] Step 1:

[0481] The server begins the process of collecting data from information sources on the internet. It uses a list of APIs and URLs as input and employs web scraping techniques to retrieve text data from each source. The output is in the form of raw text files or stream data.

[0482] Step 2:

[0483] The server performs initial filtering of the collected text data using natural language processing techniques. The input is the text data collected in step 1. It searches for inappropriate keywords within the data and selects the relevant content. The output is the filtered text data, which may be a problematic dataset.

[0484] Step 3:

[0485] The server performs sentiment analysis on filtered data. The input is filtered data. Using TextBlob and SentimentIntensityAnalyzer, it analyzes the emotional characteristics of each text and classifies them as positive, negative, or neutral. The output is data with the resulting sentiment evaluations attached.

[0486] Step 4:

[0487] The server profiles the sentiment analysis results and cross-references them with past database data. The input is the data after sentiment analysis. Based on the database, it checks the likelihood of recidivism and past patterns to create a profile. The output is the profiled analysis data.

[0488] Step 5:

[0489] The device monitors user access information in real time and issues a warning when inappropriate content is detected. Input is access information from the user's browser or applications. Real-time analysis is performed, and if a problem is found, a warning message is sent to the user. Output is the warning message or notification status.

[0490] Step 6:

[0491] Users ensure safe online behavior based on the warning messages they receive. The input is warning information from the device. Upon receiving a warning about inappropriate content, users either refrain from accessing the content or conduct further investigation. The output represents safe usage.

[0492] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0493] This invention is a system that combines an emotion engine with information gathering and analysis from the internet, enabling the identification and rapid response to inappropriate information, including defamation and threats of crime. This system operates primarily server-centric and achieves its objectives through a multi-layered processing flow.

[0494] The server first automatically collects data from multiple sources on the internet. During this process, it periodically retrieves new information using APIs and web scraping, incorporating the latest data into the system. The collected information undergoes initial filtering within the server using natural language processing techniques. Here, potentially inappropriate content is selected based on a pre-configured keyword list.

[0495] Next, the server performs a detailed sentiment analysis. At this stage, the sentiment engine is used to identify the user's emotions from the text data. The sentiment engine evaluates emotions in a multidimensional manner, such as positive, negative, and neutral, and analyzes the reliability and risk of the information. Furthermore, it profiles the sender's emotional tendencies and assesses the sender's risk by matching similarities with past behavioral patterns.

[0496] For example, if a user posts offensive content on a message board, the server immediately detects the post. Using natural language processing technology and a sentiment engine, the post is determined to be negative and threatening. If profiling reveals that the post matches a sender with a history of similar problematic behavior, the information is immediately reported to the relevant authorities. This report includes specific evidence and detailed information about the sender, helping the authorities take effective action.

[0497] In this way, the system can efficiently detect inappropriate information online and promote social safety. The server improves its detection accuracy and operates more effectively by receiving feedback from the emotion engine.

[0498] The following describes the processing flow.

[0499] Step 1:

[0500] The server collects data from internet sources. It uses APIs and web scraping techniques to automatically collect data regularly from message boards, social media, and news sites. New posts and updates are stored in the database whenever they are detected.

[0501] Step 2:

[0502] The server performs initial filtering of the collected information using natural language processing techniques. Keyword matching algorithms are used to identify posts containing potentially inappropriate words or phrases. Harmless information is excluded, and only content requiring further analysis is sent to the next process.

[0503] Step 3:

[0504] The server uses a sentiment engine on information that has passed the initial filtering. It recognizes positive, negative, or neutral sentiment from the text data and calculates a risk level for posts with particularly strong negative sentiment.

[0505] Step 4:

[0506] The server uses profiling techniques to analyze the sender's emotional tendencies and past behavior. It compares this with past posting history and criminal profiles in the database to perform a risk assessment. Particular attention is paid when behavioral patterns match.

[0507] Step 5:

[0508] The server reports information deemed to contain strong negative emotions to the relevant authorities, along with supporting data. This report includes the content of the post, the poster's profile, and other relevant evidence. This reporting allows the relevant authorities to take prompt action.

[0509] Step 6:

[0510] The server will send warning messages to users as needed. It will automatically issue warnings and corrective instructions to users who post inappropriate content. In particularly serious cases, measures such as temporarily freezing the user's account will also be taken.

[0511] (Example 2)

[0512] Next, we will describe Example 2. 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."

[0513] In modern society, inappropriate information such as defamation and threats of crime is increasing online, and there is a need to quickly and accurately detect and deal with it. However, conventional methods require a great deal of effort and time to collect and analyze information, and there is also a risk of false positives, making it difficult to implement effective countermeasures.

[0514] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0515] In this invention, the server includes a device for collecting data from multiple information sources on the Internet, a device for filtering the collected data in an initial stage using natural language processing technology, and a device for performing a detailed sentiment evaluation of the initially filtered data using a sentiment engine. This makes it possible to quickly and accurately identify inappropriate information and report it to the appropriate authorities.

[0516] "On the Internet" refers to the space where information accessed via a global network exists.

[0517] "Information source" refers to the origin of information, such as a specific website or online platform, that provides or stores the information.

[0518] "Data" refers to information in various forms, such as text, images, and audio, obtained from an information source.

[0519] "Device" refers to a component of hardware or software designed to perform a specific function.

[0520] "Natural language processing technology" refers to computer-based technologies for understanding and analyzing human language.

[0521] "Filtering" refers to the process of selecting information based on relevance and reliability, and extracting important information.

[0522] An "emotion engine" refers to an algorithm or program that analyzes emotions from text data and evaluates emotional tendencies such as positive or negative.

[0523] "Evaluation" refers to the process of determining certain properties or values ​​based on data.

[0524] "Profiling" refers to the technique of analyzing the characteristics of individual users and understanding their behavioral patterns and tendencies.

[0525] "Cross-referencing" refers to the process of comparing different datasets to find commonalities and patterns.

[0526] "Reporting" refers to the act of informing relevant organizations of specific information.

[0527] This invention provides a system for quickly detecting and responding to inappropriate information on the internet. This system primarily operates server-centric and has the function of collecting and analyzing data from multiple sources on the internet and notifying relevant organizations as needed.

[0528] First, the server collects data from multiple sources such as social media, online forums, and blogs using APIs and web scraping techniques. This data is stored on the server in text format and undergoes initial filtering using natural language processing techniques. Specific software such as text analysis libraries (e.g., NLTK or SpaCy) may be used.

[0529] Next, the server performs a sentiment evaluation of the collected data through a sentiment engine. This sentiment engine determines positive, negative, and neutral sentiment scores, and data with particularly high negative scores are analyzed in detail. It is common to use sentiment analysis APIs (e.g., IBM Watson or Google Cloud Natural Language) during this process.

[0530] The server also profiles users and cross-references this with their past behavioral history. This makes it possible to assess the risk associated with senders who repeatedly engage in specific behaviors. A similarity analysis algorithm is applied using a database that stores sender behavioral patterns.

[0531] Inappropriate information identified in this manner is immediately reported by the server to the appropriate authorities. The report includes specific evidence and detailed information about the sender, providing the authorities with the basis for effective action.

[0532] For example, if a user posts a threatening comment on social media, this system analyzes the post using a sentiment engine, compares it to the poster's past behavior based on a negative score, and immediately initiates a reporting process.

[0533] An example of a prompt for a generative AI model would be: "Explain how to identify defamatory comments posted on a message board and gather the necessary information to report them."

[0534] The introduction of this system will enable the prevention of online threats and promote social safety.

[0535] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0536] Step 1:

[0537] The server collects data from multiple sources on the internet. The input is a list of URLs for sources accessed via APIs or web scraping. The server executes queries and retrieves text data from the sources based on specific keywords. The output is the collected raw dataset. This process involves periodic updates, and the collected data is stored in a database.

[0538] Step 2:

[0539] The server performs initial filtering of the collected data using natural language processing techniques. The input here is the raw dataset obtained in step 1. Specifically, a text analysis library is used to analyze each post against a specified keyword list and determine whether negative keywords are included. The output of this operation is the initially filtered dataset.

[0540] Step 3:

[0541] The server performs a detailed sentiment assessment of the initially filtered data using the sentiment engine. The input is the dataset filtered in step 2. The sentiment analysis API is used to calculate sentiment scores for the text data, determining positive, negative, and neutral scores. The output of this step is a dataset with the sentiment scores added.

[0542] Step 4:

[0543] The server performs user profiling and cross-references past behavioral history based on data with added sentiment scores. The input for this step is the profiling information for each user included in the output data from step 3. The server applies a similarity analysis algorithm to compare the sender's past behavior with current data. The output is a dataset with risk scores evaluated.

[0544] Step 5:

[0545] The server notifies the appropriate authorities of information where the risk score exceeds a certain threshold. The high-risk information identified in the output of Step 4 is used as input. The server uses a generative AI model to document the notification and sends it to the appropriate authorities via email, along with supporting data. The output of this process is a record of the notification.

[0546] (Application Example 2)

[0547] Next, we will explain application example 2. In the following explanation, 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."

[0548] In modern society, inappropriate information and offensive content on the internet and social media pose a serious threat to individual psychological safety and overall social harmony. Addressing this problem requires real-time monitoring and analysis of relevant information, followed by a rapid response. However, current systems lack sufficient neural networking and reporting capabilities, making it impossible to completely eliminate the risk of inappropriate information spreading.

[0549] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0550] In this invention, the server includes means for collecting information from multiple information sources on the Internet, means for initial filtering of the collected information using natural language processing technology, and means for monitoring the content of the information providers being followed. This makes it possible to quickly detect and respond to inappropriate information.

[0551] The "Internet" is an information network that connects computers around the world to share and communicate data.

[0552] "Information source" refers to the place or medium from which information is collected, and includes websites and social media platforms.

[0553] "Natural language processing technology" is a technology that uses computers to process and understand human language.

[0554] "Filtering" is the process of selecting data from collected information based on specific criteria.

[0555] "Sentiment analysis" is a technique that analyzes emotional tendencies from text data and evaluates them as positive, negative, neutral, etc.

[0556] "Profiling" is the process of organizing and analyzing the characteristics and patterns of a subject based on specific data and behaviors.

[0557] A "database" is an organized collection of data used to efficiently store and retrieve information.

[0558] "Reporting" is the act of formally informing an external organization or person in charge of a matter.

[0559] "Information providers" refer to entities that disseminate information on the internet, and include individuals and companies.

[0560] "To monitor" means to constantly observe a specific subject carefully and keep track of its movements.

[0561] "Notification" refers to a means of transmitting specific information to a recipient, and includes the function of providing information in real time.

[0562] The system that implements this application is server-centric. The server is developed using the Python programming language to perform a series of processes including information gathering, analysis, and notification. The main software components include the BeautifulSoup library for web scraping, TensorFlow for sentiment analysis, and Twython, which supports the Twitter API for acquiring social media data.

[0563] First, the server automatically collects information from multiple sources on the internet. The data obtained here consists of social media posts and comments via APIs, and the collected data is organized into the required format using BeautifulSoup.

[0564] Next, the collected information is initially filtered using natural language processing techniques. This process filters out posts containing unusual keywords or specific phrases. The filtered information is then subjected to sentiment analysis using TensorFlow, classifying it as positive, negative, neutral, etc.

[0565] Furthermore, the server monitors posts from followed information providers (individuals and organizations) and, if inappropriate content is found, prepares to report it to the relevant authorities using profiling results. This report is accompanied by concrete evidence, ensuring the reliability of information sharing. In addition, users receive push notifications immediately upon detection of inappropriate content, prompting them to take necessary action.

[0566] For example, parents who want to ensure their children's online safety at home may use this system. It allows them to monitor their children's social media accounts and respond quickly if offensive content is posted.

[0567] An example of a prompt might be: "List the key steps to develop a system that performs real-time sentiment analysis on posts from a social media account and issues warning notifications when offensive content is found."

[0568] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0569] Step 1:

[0570] The server retrieves data via social media APIs. The input is account information of followed information providers, and the output is the latest posts collected from these accounts. This data is retrieved in JSON format.

[0571] Step 2:

[0572] The server parses the JSON data obtained using BeautifulSoup and extracts the necessary text information. The input is the JSON-formatted post data obtained in the previous step, and the output is the text extracted from the posts. This text is then formatted for later analysis.

[0573] Step 3:

[0574] The server uses natural language processing techniques to perform initial filtering of the extracted text. The input is the formatted post text, and the output is a list of posts containing specific keywords or phrases. This list is used to carefully analyze any unusual content.

[0575] Step 4:

[0576] The server uses TensorFlow to perform sentiment analysis on initially filtered post text. The input is the filtered post text, and the output is a sentiment score (positive, negative, neutral, etc.) for each post. The sentiment tendency of the posts is evaluated based on this score.

[0577] Step 5:

[0578] The server uses profiling functionality to cross-check the originator of sentiment analysis posts against a historical database. Inputs include sentiment scores and information about the poster, while output is data on the poster's behavioral tendencies. This data is used to assess the risk associated with the poster.

[0579] Step 6:

[0580] The server prepares to report posts deemed to be risky to the relevant authorities. The input is profiled risk information, and the output is a notification containing the evidence data necessary for reporting.

[0581] Step 7:

[0582] The server sends push notifications to users, warning them about detected inappropriate content. The input is the notification content ready for reporting, and the output is a real-time warning notification to the user. This allows users to immediately understand the situation and take appropriate action.

[0583] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0584] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0585] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0586] [Fourth Embodiment]

[0587] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0588] As shown in Figure 7, the 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.

[0589] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0590] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0591] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0592] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0593] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0594] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive 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 robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0595] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0596] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0597] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0598] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0599] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0600] This invention is a system that automatically collects data from various sources on the internet and analyzes it to identify problematic content such as defamation, crime threats, and misinformation, and to respond to it quickly. The system is server-centric and capable of efficiently processing large amounts of data.

[0601] The server first collects information from multiple websites and social media platforms. This data collection may be done directly via APIs or through web scraping techniques. The collected information is then initially filtered within the server using natural language processing. This process identifies potentially problematic content based on a pre-configured list of inappropriate keywords.

[0602] Next, the information that passes the initial filtering undergoes further sentiment and contextual analysis by the server. At this stage, machine learning models are used to determine whether the sentiment of a post is positive, negative, or neutral. Furthermore, contextual information is analyzed to understand whether the content of the post is threatening or related to criminal activity.

[0603] Subsequently, the server uses the analysis results to profile the characteristics and past behavior of the inappropriate sender, and cross-checks them with existing databases. This process checks for similar behavioral patterns in the past and infers the likelihood of recidivism.

[0604] When the system identifies inappropriate information or a source, the server notifies the relevant authorities. This notification includes detailed information about the problematic post and its source, as well as supporting data. This allows the authorities to take swift action.

[0605] As a concrete example, consider a case where a user posts a threatening message to a specific individual on social media. This post is automatically collected by the server, and threatening keywords are detected during the initial filtering stage. Subsequently, sentiment analysis confirms the negativity of the post, and during the profiling stage, it is discovered that the user may have engaged in similar behavior in the past. The information is immediately reported to the relevant authorities, and measures are taken against the user.

[0606] This system allows for the early detection of inappropriate information on the internet, enabling swift action.

[0607] The following describes the processing flow.

[0608] Step 1:

[0609] The server collects data from multiple sources. This may involve using a dedicated API to retrieve data, or directly analyzing web content using web scraping techniques. It regularly checks for updates and retrieves any new data immediately.

[0610] Step 2:

[0611] The server performs initial filtering. It uses natural language processing on the collected information to identify potentially problematic content based on keyword lists. Following basic rules, it excludes harmless information and extracts content that requires further analysis.

[0612] Step 3:

[0613] The server performs sentiment analysis on the extracted information. Using machine learning models, it classifies the sentiment of each piece of content as positive, negative, or neutral. It also performs contextual analysis to determine whether the information is threatening or potentially related to criminal activity.

[0614] Step 4:

[0615] The server profiles the senders of information that has passed the analysis and cross-checks them against the database. By comparing them with past behavioral patterns and known profiles, it assesses the likelihood and risk of recidivism. At this stage, it checks for any similar behavioral history.

[0616] Step 5:

[0617] The server reports any information deemed problematic to the relevant authorities. This report includes details of the identified information content and its origin, as well as specific evidence. This action prepares the police and other relevant agencies for a swift response.

[0618] Step 6:

[0619] The server will send warning messages to users as needed. Users who make inappropriate posts will be notified of violations of laws and regulations and encouraged to correct their behavior. In addition, the server will consider temporarily restricting access from the platform for malicious users.

[0620] (Example 1)

[0621] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0622] There is a need to quickly and accurately identify problematic information such as defamation, threats of crime, and misinformation from the vast amount of data on data communication networks, so that relevant agencies can respond immediately. This requires technological means to carry out the entire process, from data collection to analysis and notification, efficiently and effectively.

[0623] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0624] In this invention, the server includes means for acquiring information from multiple information sources on a data communication network, means for initial filtering of the acquired information using natural language processing technology, and means for performing rapid data analysis using a generative AI model. This makes it possible to accurately identify problematic information and efficiently notify relevant organizations.

[0625] A "data communication network" is a network infrastructure used to send and receive information between computers and digital devices.

[0626] An "information source" refers to the origin of specific data or information, including websites and social media platforms.

[0627] "Natural language processing technology" is a field of technology that uses computers to understand, interpret, and generate human language.

[0628] "Initial filtering" is the first stage of processing in which collected information is screened out based on inappropriate keyword lists or rules.

[0629] A "generative AI model" is an artificial intelligence algorithm that learns from large datasets and possesses generational and estimation capabilities.

[0630] "Sentiment analysis" is an analytical technique that classifies the emotions embedded in text data into positive, negative, neutral, and so on.

[0631] "Characteristic analysis" is the process of clarifying individual characteristics and behavioral patterns based on data and information.

[0632] "Contextual analysis" is a technique that analyzes text or conversations by considering the surrounding sentences and context in order to understand their meaning.

[0633] "Notification" is the act of informing other devices or organizations of specific information or results.

[0634] The server plays a central role in acquiring and processing vast amounts of information via the data communication network. Specifically, it utilizes high-performance processor-equipped server machines and cloud computing as hardware, and frameworks and libraries for implementing natural language processing technologies (e.g., Python's NLTK and SpaCy) as software. Furthermore, advanced data analysis capabilities are realized by employing generative AI models such as TensorFlow and PyTorch.

[0635] The server first automatically retrieves the latest information from websites and social media using APIs and web scraping techniques. Next, the retrieved information is initially filtered using natural language processing techniques, and problematic information is selected based on inappropriate keywords. By utilizing generative AI models, the speed and accuracy of this data analysis are improved.

[0636] Furthermore, sentiment analysis is performed using machine learning models, categorizing emotions into positive, negative, and neutral. In addition, contextual analysis is conducted to identify potential risk factors. Subsequently, the sender's behavioral patterns are established through trait analysis, and the likelihood of recidivism is evaluated by comparing them with past databases.

[0637] If problematic information or its source is identified, the server will promptly notify the relevant authorities and provide data input to resolve the issue. The notification process will include specific evidence data to enable the relevant authorities to take appropriate action immediately.

[0638] For example, if a user posts a threatening message containing specific keywords on social media, this message is automatically collected and analyzed by the server. As a result, the user's post is identified as problematic, and a notification is sent to the relevant authorities.

[0639] An example of a prompt message would be, "We want to analyze posts on social media that contain specific keywords and assess the likelihood of negative emotions and threats."

[0640] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0641] Step 1:

[0642] The server automatically retrieves information from multiple sources on the data network using APIs and web scraping techniques. The input is publicly available information from websites and social media, and the output is a structured initial dataset, which is stored for subsequent processing.

[0643] Step 2:

[0644] The server applies natural language processing techniques to the initial dataset and performs initial filtering. The input is a structured initial dataset, and the data is scanned using a list of inappropriate keywords. The output is a filtered dataset containing content that is likely to be problematic. This process removes obviously inappropriate data.

[0645] Step 3:

[0646] The server utilizes a generative AI model to perform sentiment analysis on a filtered dataset. The input is the filtered dataset, and the sentiment of the text is analyzed using a specific algorithm. The output is a dataset with sentiment labels, where each data point is assigned either positive, negative, or neutral. These labels play a crucial role in the analysis in the next step.

[0647] Step 4:

[0648] The server performs contextual analysis to understand the meaning and intent of each data point in the filtered dataset. The input is a dataset with sentiment labels, and the server analyzes the context and surrounding text to reveal its intent. The output is the analyzed contextual information, which provides material for potentially determining the threatening or criminal nature of the data.

[0649] Step 5:

[0650] The server performs characteristic analysis and matches the behavioral patterns of each data sender. The input consists of contextually analyzed data and sender information, which is then cross-referenced with the database to verify past behavioral history. The output is characteristic data of the behavioral patterns, which is used to assess the likelihood of recidivism and specific risks.

[0651] Step 6:

[0652] The server promptly notifies relevant organizations of information and senders identified as problematic. Input consists of characteristic data and analysis results, which generate a notification message. Output is a notification containing specific informational evidence sent to the relevant organizations. This notification enables a rapid response.

[0653] (Application Example 1)

[0654] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0655] The internet contains a large amount of inappropriate content, including defamation, threats of violence, and misinformation, and it is crucial to address this. However, there is a lack of effective methods to monitor information in real time and quickly warn users, which creates a risk of users accessing inappropriate content. An efficient system is needed to solve this problem.

[0656] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0657] In this invention, the server includes means for collecting information from multiple information sources on the Internet, means for initial filtering the collected information using natural language processing technology, means for performing detailed sentiment analysis on the initially filtered information, means for profiling the analyzed information and cross-checking it with past databases, and means for monitoring information accessed by users in real time and issuing warnings for inappropriate content. This reduces the risk of accessing inappropriate content and enables the provision of a safe online environment.

[0658] "Methods for collecting information from multiple sources on the internet" refers to the process of obtaining information from websites, social media, etc., and the technology of centrally collecting this information to use as basic data for analysis.

[0659] "Methods for initial filtering using natural language processing technology" refer to techniques that analyze collected text data and identify inappropriate information using specific keywords or patterns.

[0660] "Methods for conducting detailed sentiment analysis" refer to techniques that analyze the content of collected texts and determine and classify the emotional tone within that context.

[0661] "Profiling and cross-checking with past databases" refers to a technology that compares information or senders deemed inappropriate with past records to evaluate their relevance and the likelihood of recurrence.

[0662] "A means of monitoring user-accessed information in real time and issuing warnings for inappropriate content" refers to a technology that instantly analyzes information viewed or interacted with by users and notifies them of potential risks by issuing warnings if it is deemed inappropriate.

[0663] The system that implements this invention primarily operates on a server. The server retrieves data from multiple sources on the internet to collect and analyze information. APIs and web scraping techniques are used for data retrieval. The collected information is initially filtered using programming languages ​​such as Python and natural language processing libraries (e.g., NLTK and spaCy).

[0664] The server then uses sentiment analysis tools such as TextBlob and SentimentIntensityAnalyzer to perform sentiment analysis on the acquired information. At this stage, it evaluates whether the information is positive, negative, or neutral. The server also analyzes the context of the information and performs profiling and cross-checks with historical databases to determine potential threats.

[0665] Furthermore, applications installed on the user's device monitor website content in real time using JavaScript and Python scripts. If inappropriate content is detected through this monitoring, the user is immediately notified with a warning.

[0666] For example, if a user is browsing a shopping site and encounters fraudulent reviews or false information, a warning is sent to the user based on the server's analysis results. This allows users to continue their online activities with peace of mind.

[0667] An example of a prompt statement to input into a generative AI model is: "Design an app that automatically extracts specific content from social media and issues a warning when it is detected as inappropriate content. Required functions are keyword filtering and sentiment analysis." This prompt statement clearly instructs the AI ​​model on the purpose and operation of the invention.

[0668] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0669] Step 1:

[0670] The server begins the process of collecting data from information sources on the internet. It uses a list of APIs and URLs as input and employs web scraping techniques to retrieve text data from each source. The output is in the form of raw text files or stream data.

[0671] Step 2:

[0672] The server performs initial filtering of the collected text data using natural language processing techniques. The input is the text data collected in step 1. It searches for inappropriate keywords within the data and selects the relevant content. The output is the filtered text data, which may be a problematic dataset.

[0673] Step 3:

[0674] The server performs sentiment analysis on filtered data. The input is filtered data. Using TextBlob and SentimentIntensityAnalyzer, it analyzes the emotional characteristics of each text and classifies them as positive, negative, or neutral. The output is data with the resulting sentiment evaluations attached.

[0675] Step 4:

[0676] The server profiles the sentiment analysis results and cross-references them with past database data. The input is the data after sentiment analysis. Based on the database, it checks the likelihood of recidivism and past patterns to create a profile. The output is the profiled analysis data.

[0677] Step 5:

[0678] The device monitors user access information in real time and issues a warning when inappropriate content is detected. Input is access information from the user's browser or applications. Real-time analysis is performed, and if a problem is found, a warning message is sent to the user. Output is the warning message or notification status.

[0679] Step 6:

[0680] Users ensure safe online behavior based on the warning messages they receive. The input is warning information from the device. Upon receiving a warning about inappropriate content, users either refrain from accessing the content or conduct further investigation. The output represents safe usage.

[0681] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0682] This invention is a system that combines an emotion engine with information gathering and analysis from the internet, enabling the identification and rapid response to inappropriate information, including defamation and threats of crime. This system operates primarily server-centric and achieves its objectives through a multi-layered processing flow.

[0683] The server first automatically collects data from multiple sources on the internet. During this process, it periodically retrieves new information using APIs and web scraping, incorporating the latest data into the system. The collected information undergoes initial filtering within the server using natural language processing techniques. Here, potentially inappropriate content is selected based on a pre-configured keyword list.

[0684] Next, the server performs a detailed sentiment analysis. At this stage, the sentiment engine is used to identify the user's emotions from the text data. The sentiment engine evaluates emotions in a multidimensional manner, such as positive, negative, and neutral, and analyzes the reliability and risk of the information. Furthermore, it profiles the sender's emotional tendencies and assesses the sender's risk by matching similarities with past behavioral patterns.

[0685] For example, if a user posts offensive content on a message board, the server immediately detects the post. Using natural language processing technology and a sentiment engine, the post is determined to be negative and threatening. If profiling reveals that the post matches a sender with a history of similar problematic behavior, the information is immediately reported to the relevant authorities. This report includes specific evidence and detailed information about the sender, helping the authorities take effective action.

[0686] In this way, the system can efficiently detect inappropriate information online and promote social safety. The server improves its detection accuracy and operates more effectively by receiving feedback from the emotion engine.

[0687] The following describes the processing flow.

[0688] Step 1:

[0689] The server collects data from internet sources. It uses APIs and web scraping techniques to automatically collect data regularly from message boards, social media, and news sites. New posts and updates are stored in the database whenever they are detected.

[0690] Step 2:

[0691] The server performs initial filtering of the collected information using natural language processing techniques. Keyword matching algorithms are used to identify posts containing potentially inappropriate words or phrases. Harmless information is excluded, and only content requiring further analysis is sent to the next process.

[0692] Step 3:

[0693] The server uses a sentiment engine on information that has passed the initial filtering. It recognizes positive, negative, or neutral sentiment from the text data and calculates a risk level for posts with particularly strong negative sentiment.

[0694] Step 4:

[0695] The server uses profiling techniques to analyze the sender's emotional tendencies and past behavior. It compares this with past posting history and criminal profiles in the database to perform a risk assessment. Particular attention is paid when behavioral patterns match.

[0696] Step 5:

[0697] The server reports information deemed to contain strong negative emotions to the relevant authorities, along with supporting data. This report includes the content of the post, the poster's profile, and other relevant evidence. This reporting allows the relevant authorities to take prompt action.

[0698] Step 6:

[0699] The server will send warning messages to users as needed. It will automatically issue warnings and corrective instructions to users who post inappropriate content. In particularly serious cases, measures such as temporarily freezing the user's account will also be taken.

[0700] (Example 2)

[0701] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0702] In modern society, inappropriate information such as defamation and threats of crime is increasing online, and there is a need to quickly and accurately detect and deal with it. However, conventional methods require a great deal of effort and time to collect and analyze information, and there is also a risk of false positives, making it difficult to implement effective countermeasures.

[0703] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0704] In this invention, the server includes a device for collecting data from multiple information sources on the Internet, a device for filtering the collected data in an initial stage using natural language processing technology, and a device for performing a detailed sentiment evaluation of the initially filtered data using a sentiment engine. This makes it possible to quickly and accurately identify inappropriate information and report it to the appropriate authorities.

[0705] "On the Internet" refers to the space where information accessed via a global network exists.

[0706] "Information source" refers to the origin of information, such as a specific website or online platform, that provides or stores the information.

[0707] "Data" refers to information in various forms, such as text, images, and audio, obtained from an information source.

[0708] "Device" refers to a component of hardware or software designed to perform a specific function.

[0709] "Natural language processing technology" refers to computer-based technologies for understanding and analyzing human language.

[0710] "Filtering" refers to the process of selecting information based on relevance and reliability, and extracting important information.

[0711] An "emotion engine" refers to an algorithm or program that analyzes emotions from text data and evaluates emotional tendencies such as positive or negative.

[0712] "Evaluation" refers to the process of determining certain properties or values ​​based on data.

[0713] "Profiling" refers to the technique of analyzing the characteristics of individual users and understanding their behavioral patterns and tendencies.

[0714] "Cross-referencing" refers to the process of comparing different datasets to find commonalities and patterns.

[0715] "Reporting" refers to the act of informing relevant organizations of specific information.

[0716] This invention provides a system for quickly detecting and responding to inappropriate information on the internet. This system primarily operates server-centric and has the function of collecting and analyzing data from multiple sources on the internet and notifying relevant organizations as needed.

[0717] First, the server collects data from multiple sources such as social media, online forums, and blogs using APIs and web scraping techniques. This data is stored on the server in text format and undergoes initial filtering using natural language processing techniques. Specific software such as text analysis libraries (e.g., NLTK or SpaCy) may be used.

[0718] Next, the server performs a sentiment evaluation of the collected data through a sentiment engine. This sentiment engine determines positive, negative, and neutral sentiment scores, and data with particularly high negative scores are analyzed in detail. It is common to use sentiment analysis APIs (e.g., IBM Watson or Google Cloud Natural Language) during this process.

[0719] The server also profiles users and cross-references this with their past behavioral history. This makes it possible to assess the risk associated with senders who repeatedly engage in specific behaviors. A similarity analysis algorithm is applied using a database that stores sender behavioral patterns.

[0720] Inappropriate information identified in this manner is immediately reported by the server to the appropriate authorities. The report includes specific evidence and detailed information about the sender, providing the authorities with the basis for effective action.

[0721] For example, if a user posts a threatening comment on social media, this system analyzes the post using a sentiment engine, compares it to the poster's past behavior based on a negative score, and immediately initiates a reporting process.

[0722] An example of a prompt for a generative AI model would be: "Explain how to identify defamatory comments posted on a message board and gather the necessary information to report them."

[0723] The introduction of this system will enable the prevention of online threats and promote social safety.

[0724] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0725] Step 1:

[0726] The server collects data from multiple sources on the internet. The input is a list of URLs for sources accessed via APIs or web scraping. The server executes queries and retrieves text data from the sources based on specific keywords. The output is the collected raw dataset. This process involves periodic updates, and the collected data is stored in a database.

[0727] Step 2:

[0728] The server performs initial filtering of the collected data using natural language processing techniques. The input here is the raw dataset obtained in step 1. Specifically, a text analysis library is used to analyze each post against a specified keyword list and determine whether negative keywords are included. The output of this operation is the initially filtered dataset.

[0729] Step 3:

[0730] The server performs a detailed sentiment assessment of the initially filtered data using the sentiment engine. The input is the dataset filtered in step 2. The sentiment analysis API is used to calculate sentiment scores for the text data, determining positive, negative, and neutral scores. The output of this step is a dataset with the sentiment scores added.

[0731] Step 4:

[0732] The server performs user profiling and cross-references past behavioral history based on data with added sentiment scores. The input for this step is the profiling information for each user included in the output data from step 3. The server applies a similarity analysis algorithm to compare the sender's past behavior with current data. The output is a dataset with risk scores evaluated.

[0733] Step 5:

[0734] The server notifies the appropriate authorities of information where the risk score exceeds a certain threshold. The high-risk information identified in the output of Step 4 is used as input. The server uses a generative AI model to document the notification and sends it to the appropriate authorities via email, along with supporting data. The output of this process is a record of the notification.

[0735] (Application Example 2)

[0736] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0737] In modern society, inappropriate information and offensive content on the internet and social media pose a serious threat to individual psychological safety and overall social harmony. Addressing this problem requires real-time monitoring and analysis of relevant information, followed by a rapid response. However, current systems lack sufficient neural networking and reporting capabilities, making it impossible to completely eliminate the risk of inappropriate information spreading.

[0738] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0739] In this invention, the server includes means for collecting information from multiple information sources on the Internet, means for initial filtering of the collected information using natural language processing technology, and means for monitoring the content of the information providers being followed. This makes it possible to quickly detect and respond to inappropriate information.

[0740] The "Internet" is an information network that connects computers around the world to share and communicate data.

[0741] "Information source" refers to the place or medium from which information is collected, and includes websites and social media platforms.

[0742] "Natural language processing technology" is a technology that uses computers to process and understand human language.

[0743] "Filtering" is the process of selecting data from collected information based on specific criteria.

[0744] "Sentiment analysis" is a technique that analyzes emotional tendencies from text data and evaluates them as positive, negative, neutral, etc.

[0745] "Profiling" is the process of organizing and analyzing the characteristics and patterns of a subject based on specific data and behaviors.

[0746] A "database" is an organized collection of data used to efficiently store and retrieve information.

[0747] "Reporting" is the act of formally informing an external organization or person in charge of a matter.

[0748] "Information providers" refer to entities that disseminate information on the internet, and include individuals and companies.

[0749] "To monitor" means to constantly observe a specific subject carefully and keep track of its movements.

[0750] "Notification" refers to a means of transmitting specific information to a recipient, and includes the function of providing information in real time.

[0751] The system that implements this application is server-centric. The server is developed using the Python programming language to perform a series of processes including information gathering, analysis, and notification. The main software components include the BeautifulSoup library for web scraping, TensorFlow for sentiment analysis, and Twython, which supports the Twitter API for acquiring social media data.

[0752] First, the server automatically collects information from multiple sources on the internet. The data obtained here consists of social media posts and comments via APIs, and the collected data is organized into the required format using BeautifulSoup.

[0753] Next, the collected information is initially filtered using natural language processing techniques. This process filters out posts containing unusual keywords or specific phrases. The filtered information is then subjected to sentiment analysis using TensorFlow, classifying it as positive, negative, neutral, etc.

[0754] Furthermore, the server monitors posts from followed information providers (individuals and organizations) and, if inappropriate content is found, prepares to report it to the relevant authorities using profiling results. This report is accompanied by concrete evidence, ensuring the reliability of information sharing. In addition, users receive push notifications immediately upon detection of inappropriate content, prompting them to take necessary action.

[0755] For example, parents who want to ensure their children's online safety at home may use this system. It allows them to monitor their children's social media accounts and respond quickly if offensive content is posted.

[0756] An example of a prompt might be: "List the key steps to develop a system that performs real-time sentiment analysis on posts from a social media account and issues warning notifications when offensive content is found."

[0757] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0758] Step 1:

[0759] The server retrieves data via social media APIs. The input is account information of followed information providers, and the output is the latest posts collected from these accounts. This data is retrieved in JSON format.

[0760] Step 2:

[0761] The server parses the JSON data obtained using BeautifulSoup and extracts the necessary text information. The input is the JSON-formatted post data obtained in the previous step, and the output is the text extracted from the posts. This text is then formatted for later analysis.

[0762] Step 3:

[0763] The server uses natural language processing techniques to perform initial filtering of the extracted text. The input is the formatted post text, and the output is a list of posts containing specific keywords or phrases. This list is used to carefully analyze any unusual content.

[0764] Step 4:

[0765] The server uses TensorFlow to perform sentiment analysis on initially filtered post text. The input is the filtered post text, and the output is a sentiment score (positive, negative, neutral, etc.) for each post. The sentiment tendency of the posts is evaluated based on this score.

[0766] Step 5:

[0767] The server uses profiling functionality to cross-check the originator of sentiment analysis posts against a historical database. Inputs include sentiment scores and information about the poster, while output is data on the poster's behavioral tendencies. This data is used to assess the risk associated with the poster.

[0768] Step 6:

[0769] The server prepares to report posts deemed to be risky to the relevant authorities. The input is profiled risk information, and the output is a notification containing the evidence data necessary for reporting.

[0770] Step 7:

[0771] The server sends push notifications to users, warning them about detected inappropriate content. The input is the notification content ready for reporting, and the output is a real-time warning notification to the user. This allows users to immediately understand the situation and take appropriate action.

[0772] 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 controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0773] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0774] 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 this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0775] Furthermore, the emotion identification model 59, acting 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 a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0776] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0777] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0778] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0779] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0780] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is 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 the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0781] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0782] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0783] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0784] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0785] 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.

[0786] Furthermore, it is not necessary to store the entirety 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 the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0787] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0788] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of 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). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0789] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0790] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0791] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0792] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[0793] The following is further disclosed regarding the embodiments described above.

[0794] (Claim 1)

[0795] Means of collecting information from multiple sources on the internet,

[0796] A means for performing initial filtering of the collected information using natural language processing technology,

[0797] A means for performing a detailed sentiment analysis on the initially filtered information,

[0798] Methods for profiling the analyzed information and cross-checking it with past databases,

[0799] Means for reporting information and senders deemed problematic to the relevant authorities,

[0800] A system that includes this.

[0801] (Claim 2)

[0802] The system according to claim 1, further comprising means for performing contextual analysis of collected information based on the aforementioned natural language processing technology.

[0803] (Claim 3)

[0804] The system according to claim 1, comprising means for including specific evidence in the notification to the aforementioned related agency.

[0805] "Example 1"

[0806] (Claim 1)

[0807] A device that acquires information from multiple information sources on a data communication network,

[0808] A device that performs initial filtering of the acquired information using natural language processing technology,

[0809] A device that performs high-precision sentiment analysis on the initially filtered information,

[0810] A device that characterizes the analyzed information and compares it with past data sets,

[0811] A device that notifies relevant organizations of information and senders that have been deemed problematic,

[0812] A device that performs rapid data analysis using generative AI models,

[0813] A system that includes this.

[0814] (Claim 2)

[0815] The system according to claim 1, comprising a device for performing contextual analysis of information acquired based on the aforementioned natural language processing technology.

[0816] (Claim 3)

[0817] The system according to claim 1, comprising a device for including specific informational evidence in the notification to the aforementioned related organizations.

[0818] "Application Example 1"

[0819] (Claim 1)

[0820] Means of collecting information from multiple sources on the internet,

[0821] A means for performing initial filtering of the collected information using natural language processing technology,

[0822] A means for performing a detailed sentiment analysis on the initially filtered information,

[0823] Methods for profiling the analyzed information and cross-checking it with past databases,

[0824] Means for reporting information and senders deemed problematic to the relevant authorities,

[0825] A means of monitoring the information users access in real time and issuing warnings for inappropriate content,

[0826] A system that includes this.

[0827] (Claim 2)

[0828] The system according to claim 1, further comprising means for performing contextual analysis of collected information based on the aforementioned natural language processing technology.

[0829] (Claim 3)

[0830] The system according to claim 1, comprising means for including specific evidence in the notification to the aforementioned related agency.

[0831] "Example 2 of combining an emotion engine"

[0832] (Claim 1)

[0833] A device that collects data from multiple sources on the internet,

[0834] A device that filters the collected data in an initial stage using natural language processing technology,

[0835] A device that performs a detailed emotional evaluation of the initially filtered data using an emotion engine,

[0836] A device that cross-references the analyzed data with the user's profiling and past behavioral history,

[0837] A device that notifies the appropriate organization of the information and its sender when the specified conditions are met,

[0838] A system that includes this.

[0839] (Claim 2)

[0840] The system according to claim 1, comprising a device that performs contextual analysis of collected data using the aforementioned natural language processing technology.

[0841] (Claim 3)

[0842] The system according to claim 1, comprising a device for reporting to the appropriate organization with supporting evidence data.

[0843] "Application example 2 when combining with an emotional engine"

[0844] (Claim 1)

[0845] Means of collecting information from multiple sources on the internet,

[0846] A means for performing initial filtering of the collected information using natural language processing technology,

[0847] A means for performing a detailed sentiment analysis on the initially filtered information,

[0848] Methods for profiling the analyzed information and cross-checking it with past databases,

[0849] Means for reporting information and senders deemed problematic to the relevant authorities,

[0850] Means to monitor the content of information providers you follow,

[0851] A means of issuing a notification when inappropriate content is identified,

[0852] A system that includes this.

[0853] (Claim 2)

[0854] The system according to claim 1, comprising means for performing contextual analysis of collected information based on the aforementioned natural language processing technology, and monitoring posts from followed information providers.

[0855] (Claim 3)

[0856] The system according to claim 1, which includes means for including specific evidence in reporting to the relevant authorities and notifies the user of a warning. [Explanation of Symbols]

[0857] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. Means of collecting information from multiple sources on the internet, A means for performing initial filtering of the collected information using natural language processing technology, A means for performing a detailed sentiment analysis on the initially filtered information, Methods for profiling the analyzed information and cross-checking it with past databases, Means for reporting information and senders deemed problematic to the relevant authorities, A system that includes this.

2. The system according to claim 1, further comprising means for performing contextual analysis of collected information based on the aforementioned natural language processing technology.

3. The system according to claim 1, comprising means for including specific evidence in the notification to the aforementioned related organizations.

Images