system

Abstract

We provide the system. [Solution] A means for receiving text data, image data, and audio data extracted from a user, A means of analyzing received data using natural language processing and multimodal analysis techniques to evaluate aggression in communication, The analysis results in a means of providing a warning to the user when the level of aggression exceeds a predetermined threshold, A means of suggesting counseling services to users based on warnings, 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 persona chatbot control method performed by at least one processor, including 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 communication via the Internet, especially as problems brought about by anonymity and rapid information transmission, it is difficult to detect bullying behavior that occurs, and thus there is a problem that prompt and appropriate countermeasures cannot be taken. Also, in the midst of the intermingling of various forms of data (text, images, voice, etc.), it is required to accurately detect signs of bullying and take appropriate countermeasures based on it.

Means for Solving the Problems

[0005] This invention provides a means for detecting bullying with high accuracy by receiving text data, image data, and audio data exchanged between users in real time and analyzing them using the latest natural language processing and multimodal analysis technologies. Furthermore, it includes a function to provide users with step-by-step warnings when bullying is detected and to suggest connecting to counseling services as needed. This enables real-time response, realizing early detection of bullying and prompt action based on that detection.

[0006] A "user" refers to an individual who communicates using digital devices.

[0007] "Text data" refers to digital documents and written information.

[0008] "Image data" refers to visual information or images expressed in digital format.

[0009] "Audio data" refers to sound information or recordings expressed in digital format.

[0010] "Natural language processing technology" refers to the technology that enables computers to understand and process human language.

[0011] "Multimodal analysis technology" refers to a technology that comprehensively analyzes multiple different data formats (text, images, audio, etc.).

[0012] "Bullying" refers to aggressive or harmful behavior towards others and is a form of inappropriate communication in interpersonal relationships.

[0013] "Real-time" refers to information processing occurring in real time, meaning that responses are immediate and without delay.

[0014] A "warning" refers to a message or notification that draws attention to a particular action.

[0015] "Counseling service" refers to a specialized service where users can receive psychological support and guidance.

Brief Explanation of Drawings

[0016] [Figure 1] It is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] It is a conceptual diagram showing an example of the main functions of a data processing device and a smart device according to the first embodiment. [Figure 3] It is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] It 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] It is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] It 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] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] [[ID= 30]]It 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] It shows an emotion map where multiple emotions are mapped. [Figure 10] It shows an emotion map where multiple emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when the emotion engine is combined. [Figure 14]It is a sequence diagram showing the processing flow of a data processing system in Application Example 2 when a sentiment engine is combined.

Embodiment for Carrying Out the Invention

[0017] 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.

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

[0019] 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.

[0020] 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.

[0021] 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, etc.

[0022] 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).

[0023] 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."

[0024] [First Embodiment]

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

[0026] 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.

[0027] 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).

[0028] 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.

[0029] 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.

[0030] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form 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.

[0031] 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.

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

[0033] 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.

[0034] 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.

[0035] 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.

[0036] 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".

[0037] This invention is a system for detecting bullying behavior in real time during user-to-user communication and providing appropriate responses. This system is implemented through a series of processes that operate on a server and user terminals.

[0038] System Configuration

[0039] The server receives text, image, and audio data collected from users in real time. This data is encrypted and transmitted from the user's device as each communication progresses. The server analyzes the received data using the latest natural language processing and multimodal analysis technologies. In this process, the server performs sentiment analysis and tone analysis to assess for signs of bullying or inappropriate content.

[0040] If the server detects signs of bullying, it will send a warning message to the user in question. The warnings are gradual, starting with minor cautions and progressing to more detailed warnings and suggestions for connecting to counseling services in more serious cases. Users can view these messages on their devices and utilize the suggested counseling services.

[0041] Specific example

[0042] For example, suppose user A and user B are exchanging messages. If user A's message is perceived as offensive towards the other user, the server analyzes the text and determines whether the level of aggression exceeds a certain threshold. If aggression is detected, the server immediately sends a warning to user A, and if it persists, provides a detailed warning along with a link to a counseling service.

[0043] Furthermore, if user C shares an inappropriate image in the chat, the server will analyze the image and, if it determines it to be inappropriate, will issue a warning to user C. By comprehensively handling various data formats in this way, the system can detect bullying in diverse situations and enable appropriate responses. This allows for the early detection of malicious behavior on the internet and ensures a safe communication environment for users.

[0044] The following describes the processing flow.

[0045] Step 1:

[0046] The device collects communication data between users, including chat messages, sent image files, and recorded audio data. This data is prepared to be securely processed using end-to-end encryption technology, with the user's consent.

[0047] Step 2:

[0048] The device sends the collected data to the server. The data is processed in real time, minimizing delays in communication between users.

[0049] Step 3:

[0050] The server first decodes the received data. Then, it uses natural language processing techniques to analyze the text data and detect aggressive expressions or negative emotions within the text. Image data is analyzed using image recognition algorithms, and audio data is converted to text using speech recognition technology before being analyzed.

[0051] Step 4:

[0052] The server integrates the analysis results for each data type and scores the signs of bullying based on each data point. If the score exceeds a certain threshold, an alert is generated.

[0053] Step 5:

[0054] The server generates a warning message for a user if their bullying score exceeds a certain threshold. This message includes a step-by-step warning based on the user's behavior.

[0055] Step 6:

[0056] The terminal displays the warning message received from the server to the relevant user. The user can then review the warning details and consider using the suggested counseling service.

[0057] Step 7:

[0058] Users can provide feedback regarding the content or accuracy of system warnings. This feedback is aggregated by the server and used to improve the accuracy of the system's analysis.

[0059] (Example 1)

[0060] 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."

[0061] In today's information society, online communication is increasing more and more. However, bullying and inappropriate behavior among users are becoming a problem. In particular, because these behaviors unfold in real time, early detection and response are required. However, current technology does not have a sufficient means to do so. Therefore, in order to ensure a safe and healthy communication environment, there is a need for a system that can perform more advanced analysis and respond flexibly.

[0062] 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.

[0063] In this invention, the server includes means for acquiring information data extracted from users, means for analyzing the received information data using analysis techniques to evaluate aggression within interpersonal relationships, and means for providing a warning to the user if the analysis results in aggression exceeding a specified standard. This makes it possible to detect bullying between users in real time, provide staged warnings, and build a safe communication environment that enables prompt response and advice.

[0064] "User" refers to a person who uses the system to communicate.

[0065] "Information data" includes digital information expressed in forms such as text, images, and audio.

[0066] "Means of acquisition" refers to the processes and technologies used to receive information data from users.

[0067] "Analysis techniques" refer to methods such as natural language processing and multimodal analysis used to evaluate received information data.

[0068] "Means of assessing aggression" refers to criteria and techniques for determining inappropriate content or signs of bullying in information data.

[0069] "Means of providing warnings" refers to processes and technologies for notifying users of information when an attack is detected.

[0070] "Gradual warnings" refer to the practice of issuing different levels of caution or warnings to users when bullying or inappropriate behavior is observed, depending on the severity of the situation.

[0071] "Advice services" refer to resources and platforms that provide counseling and support to users who may have engaged in inappropriate behavior.

[0072] This invention is an information processing device for detecting bullying behavior in user-to-user communication in real time and providing appropriate responses. This is achieved by linking the user's terminal with a server.

[0073] The server receives encrypted information data, including text, image, and audio data, transmitted from the user's device. The user's device is responsible for generating this data and securely transferring it to the server.

[0074] The server utilizes natural language processing and deep learning-based multimodal analysis techniques to analyze the received data. This analysis uses state-of-the-art generative AI models to evaluate signs of aggression and inappropriate behavior in the data. Specifically, it understands the context of text messages and analyzes sentiment and tone. For image data, image analysis techniques are applied to check for the presence of inappropriate content. Similarly, audio data is converted to text using speech recognition technology and then subjected to text analysis.

[0075] For example, in a chat between user A and user B, if user A's message is deemed offensive, the server analyzes the message and compares its level of aggression to a baseline. If the baseline is exceeded, the server immediately sends a warning message to user A and provides a link to an advisory service if necessary. This warning message is designed to have different content depending on the level of aggression.

[0076] (Example of a prompt message)

[0077] "Evaluate whether User A's message is offensive, and if it is determined to be bullying, issue a warning."

[0078] This allows the system to detect malicious activity between users early on, helping to ensure a continuous and secure communication environment.

[0079] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0080] Step 1:

[0081] The server receives information data transmitted from the user's terminal. As input, the server obtains encrypted text data, image data, and audio data from the terminal. This data originates from communication between users. The server decrypts the encrypted data and prepares it for the next analysis step.

[0082] Step 2:

[0083] The server analyzes the received text data using natural language processing techniques. This process uses the decoded text data as input, employs a generative AI model to understand the context, and performs sentiment and tone analysis. The output is an aggression score, which is compared against criteria to determine if the text constitutes bullying.

[0084] Step 3:

[0085] The server performs image analysis on image data using deep learning. It uses decoded image data as input to evaluate whether the image contains inappropriate content. The output includes a determination of whether the image contains inappropriate elements and a score based on that evaluation. This analysis contributes to determining the level of aggression.

[0086] Step 4:

[0087] The server converts audio data into text using speech recognition technology and then performs text analysis on the converted data. Decoded audio data is used as input, and a generative AI model is used for the conversion process. The resulting text is analyzed using the same method as in step 2 to evaluate its aggression potential.

[0088] Step 5:

[0089] The server comprehensively analyzes text, images, and audio data and, if it detects signs of bullying, sends a warning message to the user in question. The analysis results for each data format are used as input, and if the level of aggression exceeds a predetermined threshold, a warning is issued to the user, and a link to an advisory service is provided as needed. The output generates the warning message the user receives and the advisory service recommendations.

[0090] (Application Example 1)

[0091] 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."

[0092] There is a need to proactively detect aggressive or inappropriate behavior in online communication within the home, ensuring user safety and providing a smooth communication environment. However, current technology lacks the means to do this in real time, making it difficult to monitor and control digital communication within the home.

[0093] 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.

[0094] In this invention, the server includes means for receiving data, image information, and sound information acquired from an information terminal; means for analyzing the received data using language processing technology and various data format analysis technologies to evaluate the threat level in information exchange; means for providing a warning to the user if the threat level exceeds a set standard based on the analysis results; means for recommending advisory services to the user based on the warning; and means for monitoring the security of digital communications within the home. This makes it possible to detect inappropriate communication behavior that may occur within the home in real time and take appropriate action.

[0095] An "information terminal" is a device used by users to send and receive digital information, and mainly includes smartphones and tablets.

[0096] "Data" refers to a collection of information obtained from users, and includes various formats such as text, images, and audio.

[0097] "Image information" refers to data formats that include visual information, such as photographs and illustrations.

[0098] "Acoustic information" refers to a data format that includes speech and sound, and includes conversations and other sound sources.

[0099] "Language processing technology" refers to technologies that perform analysis and understanding of natural language, including sentiment analysis and tone evaluation of text.

[0100] "Diverse data format analysis technology" refers to technology that comprehensively analyzes data in different formats, making it possible to evaluate the interrelationships between text, images, and audio.

[0101] "Threat level" refers to the degree of aggressive or inappropriate behavior or content in digital communication.

[0102] A "criteria" is a comparative value used to assess the level of threat, and a warning is issued when that value is exceeded.

[0103] A "warning" is a cautionary action given to a user, urging them to correct inappropriate behavior.

[0104] "Advice services" are services that provide users with counseling and other forms of support.

[0105] "Monitoring the security of digital communications" refers to continuous checks conducted to ensure that online communication within the home is conducted appropriately.

[0106] In an embodiment of the present invention, when digital communication is initiated by a user's information terminal, the terminal collects text, image, and audio data in real time and securely encrypts this data end-to-end. Next, the collected data is sent to a server, which performs analysis using natural language processing and multimodal analysis techniques. Specifically, the server uses Hugging Face's Transformers to perform sentiment and tone analysis of text, OpenCV to perform image analysis, and Google® Cloud Speech-to-Text API to transcribe and analyze audio data.

[0107] Based on the analysis results, the server assesses the threat level of the information exchange and immediately sends a warning message to the user's device if the threshold is exceeded. Furthermore, if necessary, it provides appropriate advice services by presenting the user with a link to a counseling service.

[0108] For example, if a child makes an inappropriate comment while playing an online game, that comment may be recognized as an attack on a friend. In this case, the system immediately notifies the parent and displays a warning on the child's device. In this way, the system can monitor and ensure that digital communication within the home is conducted safely.

[0109] An example of a prompt message would be: "Analyze the following chat message and evaluate whether it contains offensive content: 'I hate you, never talk to me again!'"

[0110] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0111] Step 1:

[0112] When an information terminal initiates digital communication, it collects text, image, and audio data in real time. Input includes user messages, transmitted images, and recorded audio. The data is encrypted end-to-end and securely transmitted to a server over the network.

[0113] Step 2:

[0114] The server classifies the received data, and the text data is fed into a natural language processing model (Hugging Face's Transformers). The input is encrypted text data, and prompt sentences are used to analyze sentiment and tone. The output is an aggression and threat rating score.

[0115] Step 3:

[0116] Image data is analyzed by the server using OpenCV. The input is encrypted image data, and various filters and models are applied to each image to determine whether it contains inappropriate content. The output is a determination of whether or not the image contains inappropriate content.

[0117] Step 4:

[0118] The audio data is converted to text by the server using the Google Cloud Speech-to-Text API and then fed into a parsing pipeline for text analysis. The input is encrypted audio data, which is then converted to text. The output is text data in a parseable format.

[0119] Step 5:

[0120] The server integrates evaluation scores and judgment results obtained from each data format to perform a comprehensive threat analysis. The input consists of individual evaluations from each data set, which are then combined to generate the final threat score. The output provides a final judgment on whether or not the threshold is exceeded.

[0121] Step 6:

[0122] If the server determines that the threat level exceeds the threshold, it sends a warning message to the device. The input is the result of the threat assessment, and the output is a message prompting the user to take action. A pop-up message will appear on the device, requesting that the inappropriate behavior be corrected.

[0123] Step 7:

[0124] If necessary, the server will offer the user additional advice services. Input is the user's response and actions after the warning, and output is provided with links and information about counseling services. This ensures continued support for the user.

[0125] 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.

[0126] This invention is a system that analyzes data acquired from users from multiple perspectives and detects aggression in communications between users with high accuracy. This system incorporates an emotion engine and has functions that enhance the recognition and analysis of emotions compared to conventional systems.

[0127] System Configuration

[0128] In this system, the server and the user's terminal are the main components. The terminal collects text data, image data, and audio data generated by the user, encrypts them, and sends them to the server. The server analyzes the received data using a multi-functional analysis engine. Specifically, it first analyzes the text using natural language processing technology. Next, image and audio data are processed using multimodal analysis. Furthermore, an emotion engine is incorporated to recognize the user's emotional state from this data and reinforce the analysis results.

[0129] The server evaluates the aggressiveness of the communication based on the output of the emotion engine and determines whether a warning is necessary. If aggression is detected, a warning message is generated for the user. This warning includes step-by-step instructions and, if necessary, provides a link to counseling services. The emotion engine data is also provided to the counseling services, allowing for more individualized and appropriate support.

[0130] Specific example

[0131] For example, suppose user X and user Y are communicating through chat. When user X types a specific phrase, the emotion expressing that feeling is detected from the text. The server, using its emotion engine, detects that user X is experiencing strong anger. If this emotion is determined to be aggressive, the system immediately sends a warning to user X saying, "Please be careful. This communication may be aggressive."

[0132] Furthermore, when user Z shares an image, the image analysis and sentiment engine work together to detect if the image may contain an inappropriate message. As a result, user Z receives a warning and can be guided to counseling services where they can seek support if necessary.

[0133] Thus, by integrating with an emotion engine, this system goes beyond mere data analysis to understand the user's emotional state and achieve more accurate responses.

[0134] The following describes the processing flow.

[0135] Step 1:

[0136] The device collects user-generated text, image, and audio data in real time. This data is collected by the device during user communication and is end-to-end encrypted to prepare it for secure data transmission.

[0137] Step 2:

[0138] The device sends encrypted data to the server. This transmission occurs in real time, ensuring that conversations between users continue smoothly.

[0139] Step 3:

[0140] After decrypting the received encrypted data, the server applies natural language processing and multimodal analysis techniques for data analysis. For text data, natural language processing is performed to identify context, keywords, and sentiment tone.

[0141] Step 4:

[0142] The server analyzes image data using an image recognition algorithm to check for any inappropriate content. Audio data is converted to text using speech recognition technology and then analyzed similarly as text data.

[0143] Step 5:

[0144] The server then uses an emotion engine to recognize the user's emotional state from all the collected data. This emotional information is integrated with other analysis results to provide a multifaceted assessment of the aggressiveness of the communication.

[0145] Step 6:

[0146] The server calculates an attack score based on the analysis results and generates a warning message if it exceeds a threshold. The warning is adjusted to an appropriate level depending on the situation, and in some cases, a link to a counseling service is also provided.

[0147] Step 7:

[0148] The terminal displays a warning message sent from the server to the user. The user can review the warning and choose to utilize the suggested counseling service.

[0149] Step 8:

[0150] When a user provides feedback, the device sends that feedback to the server. The server uses this feedback to refine its analysis techniques and improve the overall accuracy of the system.

[0151] (Example 2)

[0152] 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".

[0153] In modern digital communication, aggressive messages and inappropriate content are sometimes exchanged between users. This can degrade the quality of communication and potentially damage relationships. Traditional systems have struggled to detect and appropriately address this aggression, making it difficult to prevent user stress and problems.

[0154] 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.

[0155] In this invention, the server includes means for receiving information data extracted from the user's terminal, means for utilizing a generative AI model with the received information data to perform natural language processing and complex data analysis to identify emotional states in communication and evaluate aggression using an improved emotion engine, and means for providing the user with step-by-step warnings and referrals to counseling services if the analysis results show that aggression exceeds a certain threshold. This makes it possible to detect aggressive communication between users with high accuracy and provide prompt and appropriate feedback and support.

[0156] A "user's terminal" is an electronic device used by a user to generate information and communicate with the system.

[0157] "Means of receiving" refers to a function or device for receiving data transmitted from a user's terminal.

[0158] A "generative AI model" is a form of artificial intelligence that learns from large amounts of data and enables the understanding of natural language and the analysis of images and sounds.

[0159] "Natural language processing" is a technology that enables computers to understand and analyze the language that humans use in everyday life.

[0160] "Combined data analysis" is a process that integrates and analyzes data in multiple formats, such as text, images, and audio.

[0161] An "emotion engine" is a specific algorithm or model used to analyze and evaluate emotional elements from user data.

[0162] A "means for evaluating aggression" refers to a function that identifies elements indicating hostility or aggression from data and measures their degree.

[0163] A "threshold" is a standard value or boundary that is required for a certain condition to be met.

[0164] A "warning" is a message issued by a system to draw the user's attention.

[0165] "Counseling services" are services that provide professional support to alleviate the problems and stress that users face.

[0166] "Protective measures" refer to technologies and processes used to ensure the privacy and security of information.

[0167] To implement this invention, it is necessary to utilize a user's terminal and a server. The user's terminal first collects information data such as text data, image data, and audio data. This information is generated through everyday communication tools. The terminal encrypts this information data using AES encryption technology to ensure security, and then transmits it to the server.

[0168] The server decrypts the received encrypted information data and analyzes it using a generative AI model. It analyzes text data using natural language processing techniques to determine grammatical structure and emotional tone. Furthermore, it performs composite data analysis on image data using computer vision techniques to identify features and objects within the images. For audio data, it uses speech recognition technology to convert speech to text and analyze tone.

[0169] The server incorporates an emotion engine that comprehensively evaluates the user's emotional state based on information extracted from natural language processing and complex data analysis. The emotion engine plays a particularly important role in assessing aggression. If aggression is detected, the server immediately generates a series of warning messages for the user and provides a link to a counseling service. Furthermore, as a protective measure, the system is designed to maintain privacy throughout the data analysis process.

[0170] (Specific example)

[0171] For example, when user A types a phrase like "I really can't stand this!" in an online chat, the device encrypts the message and sends it to the server. The server receives this message through natural language processing and detects strong anger from the phrase "can't stand it." If the emotion engine determines that this strong emotion may be aggressive towards other users, it immediately sends a warning message to user A such as "Your emotions are running high. Please calm down." At this point, user A is also provided with a link to a counseling service if necessary.

[0172] (Example of a prompt message)

[0173] "When a user sends a message while in a state where they cannot control their emotions, please generate an appropriate warning message."

[0174] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0175] Step 1:

[0176] The device collects information such as text, images, and audio generated by the user. During this process, data is captured on the device based on the user's input actions. The collected data is encrypted using AES encryption technology to protect user privacy. The input consists of raw data resulting from user actions, and the output is encrypted data.

[0177] Step 2:

[0178] The terminal sends encrypted information data to the server. This transmission process utilizes security protocols to ensure the safe transmission of data. The input is encrypted information data, and the output is encrypted data that reaches the server.

[0179] Step 3:

[0180] The server decrypts the received encrypted data. This decryption process separates the text, image, and audio data formats, making them available for individual analysis. The input is encrypted information data, and the output is decrypted information data.

[0181] Step 4:

[0182] The server performs natural language processing on the decoded text data using a generative AI model. Specifically, it performs syntactic analysis and sentiment assessment of the text. The input is the decoded text data, and the output is the sentiment analysis result of the text.

[0183] Step 5:

[0184] The server analyzes image data using computer vision technology. This analysis detects objects and text elements within the image. The input is decoded image data, and the output is the image analysis results.

[0185] Step 6:

[0186] The server analyzes audio data using speech recognition technology, converting the speech to text and evaluating the speech tone. The input is decoded audio data, and the output is the speech-to-text conversion result and the tone evaluation result.

[0187] Step 7:

[0188] The server integrates the results of natural language processing, image analysis, and speech analysis, and uses an emotion engine to evaluate the user's overall emotional state. This process specifically assesses aggression. The input consists of various analysis results, and the output is the aggression evaluation result.

[0189] Step 8:

[0190] If an attack is detected, the server generates a series of warning messages for the user, including information about counseling services. The input is the attack assessment result, and the output is a warning message and guidance.

[0191] Step 9:

[0192] The server sends the generated warning message to the user's terminal. Immediacy and delivery confirmation of the message are crucial in this process. The input is a warning message, and the output is the warning message displayed on the user's terminal.

[0193] (Application Example 2)

[0194] 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".

[0195] Online verbal aggression has become a social problem. The objective of this invention is to maintain a safe and healthy communication environment by detecting aggression in online communication with high accuracy and issuing timely warnings to users.

[0196] 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.

[0197] In this invention, the server includes a configuration means for receiving information, still images, and audio information extracted from a user; a configuration means for processing the received information using natural language processing technology and integrated analysis technology to evaluate the aggression in the dialogue; and a configuration means for monitoring user statements in real time on an online exchange platform and immediately notifying the user if aggression is detected. This makes it possible to quickly and accurately evaluate the aggression of language, images, and audio uttered by users online and provide appropriate warnings.

[0198] "Information extracted from users" refers to text, still images, and audio information generated or shared by users.

[0199] A "still image" is digital data representing a visual expression captured or shared by a user.

[0200] "Audio information" refers to audio data that a user has spoken or recorded.

[0201] "Natural language processing technology" is a technology that enables computers to understand human language, and involves the analysis and interpretation of text data.

[0202] "Integrated analysis techniques" are technologies that integrate and analyze multiple data formats to gain deeper insights.

[0203] An "online exchange platform" refers to a virtual space on the internet where users can exchange information and opinions with each other.

[0204] Aggression is an index that assesses hostile or harmful intentions or attitudes expressed through language, images, or sounds.

[0205] A "real-time monitoring configuration" refers to a system mechanism that instantly monitors user activity and analyzes events that occur almost simultaneously.

[0206] "Configuration for notifying users" refers to a function that immediately informs users of information detected by the system.

[0207] The system for realizing this invention consists of a user terminal and a server. The user terminal has the function of collecting text, still images, and audio information generated or shared by the user, encrypting them, and sending them to the server. For natural language processing, Python and its libraries, NLTK and spaCy, can be used, and for sentiment analysis, Hugging Face's Transformers are used. This makes it possible to effectively detect the aggressiveness of the user's statements.

[0208] The server receives encrypted data and analyzes the information. Here, natural language processing techniques are used to analyze text, and integrated analysis techniques are used to evaluate still images and audio information. This makes it possible to assess the aggression level of the communication content in real time and with high accuracy.

[0209] For example, if a user makes an inappropriate comment towards another user in an online forum, and the server determines that the comment is offensive, a warning message such as "This comment is inappropriate. Please choose better words to maintain the quality of communication" will be sent immediately. This is expected to encourage users to be more mindful of their online communication.

[0210] An example of a prompt message is, "Analyze the user's posts and evaluate the aggression level. If it is deemed aggression, generate a warning message." This is how instructions can be given to the generation AI model. By using this prompt message, the system can efficiently evaluate the aggression level between users and issue warnings.

[0211] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0212] Step 1:

[0213] The user's device collects text, still images, and audio information generated or shared by the user. This data is encrypted within the device to ensure security and prepared for transmission to the server. Data input occurs through the user interface, and output is in the form of encrypted data files.

[0214] Step 2:

[0215] The terminal sends encrypted data to the server. Secure communication methods such as the SSL / TLS protocol are used to prevent data eavesdropping and tampering. The input is an encrypted data file, and the output is the completion of the data transfer to the server.

[0216] Step 3:

[0217] The server decrypts the received encrypted data, thereby obtaining text, still images, and audio information in their original form. The input is an encrypted data file, and the output is the decrypted raw data.

[0218] Step 4:

[0219] The server analyzes text data using natural language processing techniques. Specifically, it uses Hugging Face's Transformers to extract emotions from the text and evaluate their aggression level. The input is text data, and the output is an aggression evaluation score.

[0220] Step 5:

[0221] The server processes decoded still images and audio information using integrated analysis techniques to assess their aggression. This is done using image recognition and audio analysis techniques to identify the meaning and emotion of each data point. The input is still images and audio information, and the output is an evaluation score of their aggression.

[0222] Step 6:

[0223] The server integrates evaluation scores from text, still images, and audio information to calculate an overall level of aggression. Based on this evaluation, it generates warnings and instructions if necessary. The input is the aggression score from each medium, and the output is the overall aggression score and warning messages.

[0224] Step 7:

[0225] The server sends the generated warning message to the user's terminal. The warning message is created using a generation AI model based on the prompt text and is provided in a user-friendly format. The input is the overall attack score, and the output is the warning message displayed on the user interface.

[0226] 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.

[0227] 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.

[0228] 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.

[0229] [Second Embodiment]

[0230] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0231] 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.

[0232] 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).

[0233] 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.

[0234] 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.

[0235] 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).

[0236] 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.

[0237] 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.

[0238] 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.

[0239] 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.

[0240] 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.

[0241] 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".

[0242] This invention is a system for detecting bullying behavior in real time during user-to-user communication and providing appropriate responses. This system is implemented through a series of processes that operate on a server and user terminals.

[0243] System Configuration

[0244] The server receives text, image, and audio data collected from users in real time. This data is encrypted and transmitted from the user's device as each communication progresses. The server analyzes the received data using the latest natural language processing and multimodal analysis technologies. In this process, the server performs sentiment analysis and tone analysis to assess for signs of bullying or inappropriate content.

[0245] If the server detects signs of bullying, it will send a warning message to the user in question. The warnings are gradual, starting with minor cautions and progressing to more detailed warnings and suggestions for connecting to counseling services in more serious cases. Users can view these messages on their devices and utilize the suggested counseling services.

[0246] Specific example

[0247] For example, suppose user A and user B are exchanging messages. If user A's message is perceived as offensive towards the other user, the server analyzes the text and determines whether the level of aggression exceeds a certain threshold. If aggression is detected, the server immediately sends a warning to user A, and if it persists, provides a detailed warning along with a link to a counseling service.

[0248] Furthermore, if user C shares an inappropriate image in the chat, the server will analyze the image and, if it determines it to be inappropriate, will issue a warning to user C. By comprehensively handling various data formats in this way, the system can detect bullying in diverse situations and enable appropriate responses. This allows for the early detection of malicious behavior on the internet and ensures a safe communication environment for users.

[0249] The following describes the processing flow.

[0250] Step 1:

[0251] The device collects communication data between users, including chat messages, sent image files, and recorded audio data. This data is prepared to be securely processed using end-to-end encryption technology, with the user's consent.

[0252] Step 2:

[0253] The device sends the collected data to the server. The data is processed in real time, minimizing delays in communication between users.

[0254] Step 3:

[0255] The server first decodes the received data. Then, it uses natural language processing techniques to analyze the text data and detect aggressive expressions or negative emotions within the text. Image data is analyzed using image recognition algorithms, and audio data is converted to text using speech recognition technology before being analyzed.

[0256] Step 4:

[0257] The server integrates the analysis results for each data type and scores the signs of bullying based on each data point. If the score exceeds a certain threshold, an alert is generated.

[0258] Step 5:

[0259] The server generates a warning message for a user if their bullying score exceeds a certain threshold. This message includes a step-by-step warning based on the user's behavior.

[0260] Step 6:

[0261] The terminal displays the warning message received from the server to the relevant user. The user can then review the warning details and consider using the suggested counseling service.

[0262] Step 7:

[0263] Users can provide feedback regarding the content or accuracy of system warnings. This feedback is aggregated by the server and used to improve the accuracy of the system's analysis.

[0264] (Example 1)

[0265] 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."

[0266] In today's information society, online communication is increasing more and more. However, bullying and inappropriate behavior among users are becoming a problem. In particular, because these behaviors unfold in real time, early detection and response are required. However, current technology does not have a sufficient means to do so. Therefore, in order to ensure a safe and healthy communication environment, there is a need for a system that can perform more advanced analysis and respond flexibly.

[0267] 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.

[0268] In this invention, the server includes means for acquiring information data extracted from users, means for analyzing the received information data using analysis techniques to evaluate aggression within interpersonal relationships, and means for providing a warning to the user if the analysis results in aggression exceeding a specified standard. This makes it possible to detect bullying between users in real time, provide staged warnings, and build a safe communication environment that enables prompt response and advice.

[0269] "User" refers to a person who uses the system to communicate.

[0270] "Information data" includes digital information expressed in forms such as text, images, and audio.

[0271] "Means of acquisition" refers to the processes and technologies used to receive information data from users.

[0272] "Analysis techniques" refer to methods such as natural language processing and multimodal analysis used to evaluate received information data.

[0273] "Means of assessing aggression" refers to criteria and techniques for determining inappropriate content or signs of bullying in information data.

[0274] "Means of providing warnings" refers to processes and technologies for notifying users of information when an attack is detected.

[0275] "Gradual warnings" refer to the practice of issuing different levels of caution or warnings to users when bullying or inappropriate behavior is observed, depending on the severity of the situation.

[0276] "Advice services" refer to resources and platforms that provide counseling and support to users who may have engaged in inappropriate behavior.

[0277] This invention is an information processing device for detecting bullying behavior in user-to-user communication in real time and providing appropriate responses. This is achieved by linking the user's terminal with a server.

[0278] The server receives encrypted information data, including text, image, and audio data, transmitted from the user's device. The user's device is responsible for generating this data and securely transferring it to the server.

[0279] The server utilizes natural language processing technology and multimodal analysis technology based on deep learning to analyze the received data. In this analysis, the latest generative AI model is used to evaluate signs of aggression and inappropriate behavior in the data. Specifically, it understands the context of text messages and conducts sentiment and tone analysis. For image data, image analysis technology is applied to check for inappropriate content. Also, voice data is converted to text by voice recognition technology and then text analysis is performed similarly.

[0280] For example, in a situation where User A and User B are chatting, if User A's message is judged to be aggressive, the server analyzes the speech and compares the degree of aggression with a reference value. If it exceeds the reference value, the server immediately sends a warning message to User A and provides a link to an advice service if necessary. This warning message is designed to have different contents according to different levels of aggression.

[0281] (Example of a prompt sentence)

[0282] "Evaluate whether User A is aggressive in the content of the message and send a warning if it is judged to be bullying behavior."

[0283] This enables the system to detect malicious behavior among users at an early stage and support the establishment of a continuous and safe communication environment.

[0284] The flow of the specific process in Example 1 will be described using FIG. 11.

[0285] Step 1:

[0286] The server receives information data transmitted from the user's terminal. As input, the server obtains text data, image data, and audio data in an encrypted form from the terminal. This data is generated in the communication between users. The server decrypts the encrypted data and prepares for the next analysis step.

[0287] Step 2:

[0288] The server analyzes the received text data using natural language processing techniques. In this process, the decrypted text data is used as input, and the generative AI model is used to understand the context and perform sentiment analysis and tone analysis. As output, an aggressiveness score is obtained and compared with a criterion for determining whether the text corresponds to bullying behavior.

[0289] Step 3:

[0290] The server performs image analysis on the image data using deep learning. Using the decrypted image data as input, it evaluates whether there is inappropriate content in the image. As output, a determination result of whether it contains inappropriate elements and a score based on that evaluation are obtained. This analysis result contributes to the determination of aggressiveness.

[0291] Step 4:

[0292] The server converts the audio data into text using speech recognition technology and performs text analysis on the converted text. Using the decrypted audio data as input, the conversion process is performed using the generative AI model. The text obtained as output is analyzed using the same method as in Step 2 to evaluate aggressiveness.

[0293] Step 5:

[0294] The server comprehensively analyzes text, images, and audio data and, if it detects signs of bullying, sends a warning message to the user in question. The analysis results for each data format are used as input, and if the level of aggression exceeds a predetermined threshold, a warning is issued to the user, and a link to an advisory service is provided as needed. The output generates the warning message the user receives and the advisory service recommendations.

[0295] (Application Example 1)

[0296] 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."

[0297] There is a need to proactively detect aggressive or inappropriate behavior in online communication within the home, ensuring user safety and providing a smooth communication environment. However, current technology lacks the means to do this in real time, making it difficult to monitor and control digital communication within the home.

[0298] 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.

[0299] In this invention, the server includes means for receiving data, image information, and sound information acquired from an information terminal; means for analyzing the received data using language processing technology and various data format analysis technologies to evaluate the threat level in information exchange; means for providing a warning to the user if the threat level exceeds a set standard based on the analysis results; means for recommending advisory services to the user based on the warning; and means for monitoring the security of digital communications within the home. This makes it possible to detect inappropriate communication behavior that may occur within the home in real time and take appropriate action.

[0300] An "information terminal" is a device for a user to send and receive digital information, mainly including smartphones, tablets, etc.

[0301] "Data" is a collection of information obtained from users, including various forms such as text, images, and sounds.

[0302] "Image information" is a data format including visual information, corresponding to photos and illustrations.

[0303] "Acoustic information" is a data format including voices and sounds, including conversations and other sound sources.

[0304] "Language processing technology" is a technology for analyzing and understanding natural languages, performing text sentiment analysis and tone evaluation.

[0305] "Diverse data format analysis technology" is a technology for comprehensively analyzing data in different formats, enabling the evaluation of the interrelationships among text, images, and sounds.

[0306] "Threat level" refers to the degree of aggressive or inappropriate actions and content in digital communication.

[0307] "Criterion" is a comparison value when evaluating the threat level, and a warning will be issued if the value is exceeded.

[0308] "Attention" is a warning action taken towards users, prompting them to correct inappropriate actions.

[0309] "Advisory service" is a service for providing counseling and other support to users.

[0310] "Monitoring the security of digital communication" is a continuous check conducted to maintain proper online communication within a household.

[0311] In an embodiment of the present invention, when digital communication is initiated by a user's information terminal, the terminal collects text, image, and audio data in real time and securely encrypts this data end-to-end. Next, the collected data is sent to a server, which performs analysis using natural language processing and multimodal analysis techniques. Specifically, the server uses Hugging Face's Transformers to perform sentiment and tone analysis of text, OpenCV to perform image analysis, and the Google Cloud Speech-to-Text API to transcribe and analyze audio data.

[0312] Based on the analysis results, the server assesses the threat level of the information exchange and immediately sends a warning message to the user's device if the threshold is exceeded. Furthermore, if necessary, it provides appropriate advice services by presenting the user with a link to a counseling service.

[0313] For example, if a child makes an inappropriate comment while playing an online game, that comment may be recognized as an attack on a friend. In this case, the system immediately notifies the parent and displays a warning on the child's device. In this way, the system can monitor and ensure that digital communication within the home is conducted safely.

[0314] An example of a prompt message would be: "Analyze the following chat message and evaluate whether it contains offensive content: 'I hate you, never talk to me again!'"

[0315] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0316] Step 1:

[0317] When an information terminal initiates digital communication, it collects text, image, and audio data in real time. Input includes user messages, transmitted images, and recorded audio. The data is encrypted end-to-end and securely transmitted to a server over the network.

[0318] Step 2:

[0319] The server classifies the received data, and the text data is fed into a natural language processing model (Hugging Face's Transformers). The input is encrypted text data, and prompt sentences are used to analyze sentiment and tone. The output is an aggression and threat rating score.

[0320] Step 3:

[0321] Image data is analyzed by the server using OpenCV. The input is encrypted image data, and various filters and models are applied to each image to determine whether it contains inappropriate content. The output is a determination of whether or not the image contains inappropriate content.

[0322] Step 4:

[0323] The audio data is converted to text by the server using the Google Cloud Speech-to-Text API and then fed into a parsing pipeline for text analysis. The input is encrypted audio data, which is then converted to text. The output is text data in a parseable format.

[0324] Step 5:

[0325] The server integrates evaluation scores and judgment results obtained from each data format to perform a comprehensive threat analysis. The input consists of individual evaluations from each data set, which are then combined to generate the final threat score. The output provides a final judgment on whether or not the threshold is exceeded.

[0326] Step 6:

[0327] If the server determines that the threat level exceeds the threshold, it sends a warning message to the device. The input is the result of the threat assessment, and the output is a message prompting the user to take action. A pop-up message will appear on the device, requesting that the inappropriate behavior be corrected.

[0328] Step 7:

[0329] If necessary, the server will offer the user additional advice services. Input is the user's response and actions after the warning, and output is provided with links and information about counseling services. This ensures continued support for the user.

[0330] 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.

[0331] This invention is a system that analyzes data acquired from users from multiple perspectives and detects aggression in communications between users with high accuracy. This system incorporates an emotion engine and has functions that enhance the recognition and analysis of emotions compared to conventional systems.

[0332] System Configuration

[0333] In this system, the server and the user's terminal are the main components. The terminal collects text data, image data, and audio data generated by the user, encrypts them, and sends them to the server. The server analyzes the received data using a multi-functional analysis engine. Specifically, it first analyzes the text using natural language processing technology. Next, image and audio data are processed using multimodal analysis. Furthermore, an emotion engine is incorporated to recognize the user's emotional state from this data and reinforce the analysis results.

[0334] The server evaluates the aggressiveness of the communication based on the output of the emotion engine and determines whether a warning is necessary. If aggression is detected, a warning message is generated for the user. This warning includes step-by-step instructions and, if necessary, provides a link to counseling services. The emotion engine data is also provided to the counseling services, allowing for more individualized and appropriate support.

[0335] Specific example

[0336] For example, suppose user X and user Y are communicating through chat. When user X types a specific phrase, the emotion expressing that feeling is detected from the text. The server, using its emotion engine, detects that user X is experiencing strong anger. If this emotion is determined to be aggressive, the system immediately sends a warning to user X saying, "Please be careful. This communication may be aggressive."

[0337] Furthermore, when user Z shares an image, the image analysis and sentiment engine work together to detect if the image may contain an inappropriate message. As a result, user Z receives a warning and can be guided to counseling services where they can seek support if necessary.

[0338] Thus, by integrating with an emotion engine, this system goes beyond mere data analysis to understand the user's emotional state and achieve more accurate responses.

[0339] The following describes the processing flow.

[0340] Step 1:

[0341] The device collects user-generated text, image, and audio data in real time. This data is collected by the device during user communication and is end-to-end encrypted to prepare it for secure data transmission.

[0342] Step 2:

[0343] The device sends encrypted data to the server. This transmission occurs in real time, ensuring that conversations between users continue smoothly.

[0344] Step 3:

[0345] After decrypting the received encrypted data, the server applies natural language processing and multimodal analysis techniques for data analysis. For text data, natural language processing is performed to identify context, keywords, and sentiment tone.

[0346] Step 4:

[0347] The server analyzes image data using an image recognition algorithm to check for any inappropriate content. Audio data is converted to text using speech recognition technology and then analyzed similarly as text data.

[0348] Step 5:

[0349] The server then uses an emotion engine to recognize the user's emotional state from all the collected data. This emotional information is integrated with other analysis results to provide a multifaceted assessment of the aggressiveness of the communication.

[0350] Step 6:

[0351] The server calculates an attack score based on the analysis results and generates a warning message if it exceeds a threshold. The warning is adjusted to an appropriate level depending on the situation, and in some cases, a link to a counseling service is also provided.

[0352] Step 7:

[0353] The terminal displays a warning message sent from the server to the user. The user can review the warning and choose to utilize the suggested counseling service.

[0354] Step 8:

[0355] When a user provides feedback, the device sends that feedback to the server. The server uses this feedback to refine its analysis techniques and improve the overall accuracy of the system.

[0356] (Example 2)

[0357] 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".

[0358] In modern digital communication, aggressive messages and inappropriate content are sometimes exchanged between users. This can degrade the quality of communication and potentially damage relationships. Traditional systems have struggled to detect and appropriately address this aggression, making it difficult to prevent user stress and problems.

[0359] 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.

[0360] In this invention, the server includes means for receiving information data extracted from the user's terminal, means for utilizing a generative AI model with the received information data to perform natural language processing and complex data analysis to identify emotional states in communication and evaluate aggression using an improved emotion engine, and means for providing the user with step-by-step warnings and referrals to counseling services if the analysis results show that aggression exceeds a certain threshold. This makes it possible to detect aggressive communication between users with high accuracy and provide prompt and appropriate feedback and support.

[0361] A "user's terminal" is an electronic device used by a user to generate information and communicate with the system.

[0362] "Means of receiving" refers to a function or device for receiving data transmitted from a user's terminal.

[0363] A "generative AI model" is a form of artificial intelligence that learns from large amounts of data and enables the understanding of natural language and the analysis of images and sounds.

[0364] "Natural language processing" is a technology that enables computers to understand and analyze the language that humans use in everyday life.

[0365] "Combined data analysis" is a process that integrates and analyzes data in multiple formats, such as text, images, and audio.

[0366] An "emotion engine" is a specific algorithm or model used to analyze and evaluate emotional elements from user data.

[0367] A "means for evaluating aggression" refers to a function that identifies elements indicating hostility or aggression from data and measures their degree.

[0368] A "threshold" is a standard value or boundary that is required for a certain condition to be met.

[0369] A "warning" is a message issued by a system to draw the user's attention.

[0370] "Counseling services" are services that provide professional support to alleviate the problems and stress that users face.

[0371] "Protective measures" refer to technologies and processes used to ensure the privacy and security of information.

[0372] To implement this invention, it is necessary to utilize a user's terminal and a server. The user's terminal first collects information data such as text data, image data, and audio data. This information is generated through everyday communication tools. The terminal encrypts this information data using AES encryption technology to ensure security, and then transmits it to the server.

[0373] The server decrypts the received encrypted information data and analyzes it using a generative AI model. It analyzes text data using natural language processing techniques to determine grammatical structure and emotional tone. Furthermore, it performs composite data analysis on image data using computer vision techniques to identify features and objects within the images. For audio data, it uses speech recognition technology to convert speech to text and analyze tone.

[0374] The server incorporates an emotion engine that comprehensively evaluates the user's emotional state based on information extracted from natural language processing and complex data analysis. The emotion engine plays a particularly important role in assessing aggression. If aggression is detected, the server immediately generates a series of warning messages for the user and provides a link to a counseling service. Furthermore, as a protective measure, the system is designed to maintain privacy throughout the data analysis process.

[0375] (Specific example)

[0376] For example, when user A types a phrase like "I really can't stand this!" in an online chat, the device encrypts the message and sends it to the server. The server receives this message through natural language processing and detects strong anger from the phrase "can't stand it." If the emotion engine determines that this strong emotion may be aggressive towards other users, it immediately sends a warning message to user A such as "Your emotions are running high. Please calm down." At this point, user A is also provided with a link to a counseling service if necessary.

[0377] (Example of a prompt message)

[0378] "When a user sends a message while in a state where they cannot control their emotions, please generate an appropriate warning message."

[0379] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0380] Step 1:

[0381] The device collects information such as text, images, and audio generated by the user. During this process, data is captured on the device based on the user's input actions. The collected data is encrypted using AES encryption technology to protect user privacy. The input consists of raw data resulting from user actions, and the output is encrypted data.

[0382] Step 2:

[0383] The terminal sends encrypted information data to the server. This transmission process utilizes security protocols to ensure the safe transmission of data. The input is encrypted information data, and the output is encrypted data that reaches the server.

[0384] Step 3:

[0385] The server decrypts the received encrypted data. This decryption process separates the text, image, and audio data formats, making them available for individual analysis. The input is encrypted information data, and the output is decrypted information data.

[0386] Step 4:

[0387] The server performs natural language processing on the decoded text data using a generative AI model. Specifically, it performs syntactic analysis and sentiment assessment of the text. The input is the decoded text data, and the output is the sentiment analysis result of the text.

[0388] Step 5:

[0389] The server analyzes image data using computer vision technology. This analysis detects objects and text elements within the image. The input is decoded image data, and the output is the image analysis results.

[0390] Step 6:

[0391] The server analyzes audio data using speech recognition technology, converting the speech to text and evaluating the speech tone. The input is decoded audio data, and the output is the speech-to-text conversion result and the tone evaluation result.

[0392] Step 7:

[0393] The server integrates the results of natural language processing, image analysis, and speech analysis, and uses an emotion engine to evaluate the user's overall emotional state. This process specifically assesses aggression. The input consists of various analysis results, and the output is the aggression evaluation result.

[0394] Step 8:

[0395] If an attack is detected, the server generates a series of warning messages for the user, including information about counseling services. The input is the attack assessment result, and the output is a warning message and guidance.

[0396] Step 9:

[0397] The server sends the generated warning message to the user's terminal. Immediacy and delivery confirmation of the message are crucial in this process. The input is a warning message, and the output is the warning message displayed on the user's terminal.

[0398] (Application Example 2)

[0399] 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."

[0400] Online verbal aggression has become a social problem. The objective of this invention is to maintain a safe and healthy communication environment by detecting aggression in online communication with high accuracy and issuing timely warnings to users.

[0401] 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.

[0402] In this invention, the server includes a configuration means for receiving information, still images, and audio information extracted from a user; a configuration means for processing the received information using natural language processing technology and integrated analysis technology to evaluate the aggression in the dialogue; and a configuration means for monitoring user statements in real time on an online exchange platform and immediately notifying the user if aggression is detected. This makes it possible to quickly and accurately evaluate the aggression of language, images, and audio uttered by users online and provide appropriate warnings.

[0403] "Information extracted from users" refers to text, still images, and audio information generated or shared by users.

[0404] A "still image" is digital data representing a visual expression captured or shared by a user.

[0405] "Audio information" refers to audio data that a user has spoken or recorded.

[0406] "Natural language processing technology" is a technology that enables computers to understand human language, and involves the analysis and interpretation of text data.

[0407] "Integrated analysis techniques" are technologies that integrate and analyze multiple data formats to gain deeper insights.

[0408] An "online exchange platform" refers to a virtual space on the internet where users can exchange information and opinions with each other.

[0409] Aggression is an index that assesses hostile or harmful intentions or attitudes expressed through language, images, or sounds.

[0410] A "real-time monitoring configuration" refers to a system mechanism that instantly monitors user activity and analyzes events that occur almost simultaneously.

[0411] "Configuration for notifying users" refers to a function that immediately informs users of information detected by the system.

[0412] The system for realizing this invention consists of a user terminal and a server. The user terminal has the function of collecting text, still images, and audio information generated or shared by the user, encrypting them, and sending them to the server. For natural language processing, Python and its libraries, NLTK and spaCy, can be used, and for sentiment analysis, Hugging Face's Transformers are used. This makes it possible to effectively detect the aggressiveness of the user's statements.

[0413] The server receives encrypted data and analyzes the information. Here, natural language processing techniques are used to analyze text, and integrated analysis techniques are used to evaluate still images and audio information. This makes it possible to assess the aggression level of the communication content in real time and with high accuracy.

[0414] For example, if a user makes an inappropriate comment towards another user in an online forum, and the server determines that the comment is offensive, a warning message such as "This comment is inappropriate. Please choose better words to maintain the quality of communication" will be sent immediately. This is expected to encourage users to be more mindful of their online communication.

[0415] An example of a prompt message is, "Analyze the user's posts and evaluate the aggression level. If it is deemed aggression, generate a warning message." This is how instructions can be given to the generation AI model. By using this prompt message, the system can efficiently evaluate the aggression level between users and issue warnings.

[0416] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0417] Step 1:

[0418] The user's device collects text, still images, and audio information generated or shared by the user. This data is encrypted within the device to ensure security and prepared for transmission to the server. Data input occurs through the user interface, and output is in the form of encrypted data files.

[0419] Step 2:

[0420] The terminal sends encrypted data to the server. Secure communication methods such as the SSL / TLS protocol are used to prevent data eavesdropping and tampering. The input is an encrypted data file, and the output is the completion of the data transfer to the server.

[0421] Step 3:

[0422] The server decrypts the received encrypted data, thereby obtaining text, still images, and audio information in their original form. The input is an encrypted data file, and the output is the decrypted raw data.

[0423] Step 4:

[0424] The server analyzes text data using natural language processing techniques. Specifically, it uses Hugging Face's Transformers to extract emotions from the text and evaluate their aggression level. The input is text data, and the output is an aggression evaluation score.

[0425] Step 5:

[0426] The server processes decoded still images and audio information using integrated analysis techniques to assess their aggression. This is done using image recognition and audio analysis techniques to identify the meaning and emotion of each data point. The input is still images and audio information, and the output is an evaluation score of their aggression.

[0427] Step 6:

[0428] The server integrates evaluation scores from text, still images, and audio information to calculate an overall level of aggression. Based on this evaluation, it generates warnings and instructions if necessary. The input is the aggression score from each medium, and the output is the overall aggression score and warning messages.

[0429] Step 7:

[0430] The server sends the generated warning message to the user's terminal. The warning message is created using a generation AI model based on the prompt text and is provided in a user-friendly format. The input is the overall attack score, and the output is the warning message displayed on the user interface.

[0431] 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.

[0432] 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.

[0433] 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.

[0434] [Third Embodiment]

[0435] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0436] 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.

[0437] 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).

[0438] 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.

[0439] 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.

[0440] 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).

[0441] 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.

[0442] 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.

[0443] 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.

[0444] 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.

[0445] 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.

[0446] 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".

[0447] This invention is a system for detecting bullying behavior in real time during user-to-user communication and providing appropriate responses. This system is implemented through a series of processes that operate on a server and user terminals.

[0448] System Configuration

[0449] The server receives text, image, and audio data collected from users in real time. This data is encrypted and transmitted from the user's device as each communication progresses. The server analyzes the received data using the latest natural language processing and multimodal analysis technologies. In this process, the server performs sentiment analysis and tone analysis to assess for signs of bullying or inappropriate content.

[0450] If the server detects signs of bullying, it will send a warning message to the user in question. The warnings are gradual, starting with minor cautions and progressing to more detailed warnings and suggestions for connecting to counseling services in more serious cases. Users can view these messages on their devices and utilize the suggested counseling services.

[0451] Specific example

[0452] For example, suppose user A and user B are exchanging messages. If user A's message is perceived as offensive towards the other user, the server analyzes the text and determines whether the level of aggression exceeds a certain threshold. If aggression is detected, the server immediately sends a warning to user A, and if it persists, provides a detailed warning along with a link to a counseling service.

[0453] Furthermore, if user C shares an inappropriate image in the chat, the server will analyze the image and, if it determines it to be inappropriate, will issue a warning to user C. By comprehensively handling various data formats in this way, the system can detect bullying in diverse situations and enable appropriate responses. This allows for the early detection of malicious behavior on the internet and ensures a safe communication environment for users.

[0454] The following describes the processing flow.

[0455] Step 1:

[0456] The device collects communication data between users, including chat messages, sent image files, and recorded audio data. This data is prepared to be securely processed using end-to-end encryption technology, with the user's consent.

[0457] Step 2:

[0458] The device sends the collected data to the server. The data is processed in real time, minimizing delays in communication between users.

[0459] Step 3:

[0460] The server first decodes the received data. Then, it uses natural language processing techniques to analyze the text data and detect aggressive expressions or negative emotions within the text. Image data is analyzed using image recognition algorithms, and audio data is converted to text using speech recognition technology before being analyzed.

[0461] Step 4:

[0462] The server integrates the analysis results for each data type and scores the signs of bullying based on each data point. If the score exceeds a certain threshold, an alert is generated.

[0463] Step 5:

[0464] The server generates a warning message for a user if their bullying score exceeds a certain threshold. This message includes a step-by-step warning based on the user's behavior.

[0465] Step 6:

[0466] The terminal displays the warning message received from the server to the relevant user. The user can then review the warning details and consider using the suggested counseling service.

[0467] Step 7:

[0468] Users can provide feedback regarding the content or accuracy of system warnings. This feedback is aggregated by the server and used to improve the accuracy of the system's analysis.

[0469] (Example 1)

[0470] 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."

[0471] In today's information society, online communication is increasing more and more. However, bullying and inappropriate behavior among users are becoming a problem. In particular, because these behaviors unfold in real time, early detection and response are required. However, current technology does not have a sufficient means to do so. Therefore, in order to ensure a safe and healthy communication environment, there is a need for a system that can perform more advanced analysis and respond flexibly.

[0472] 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.

[0473] In this invention, the server includes means for acquiring information data extracted from users, means for analyzing the received information data using analysis techniques to evaluate aggression within interpersonal relationships, and means for providing a warning to the user if the analysis results in aggression exceeding a specified standard. This makes it possible to detect bullying between users in real time, provide staged warnings, and build a safe communication environment that enables prompt response and advice.

[0474] "User" refers to a person who uses the system to communicate.

[0475] "Information data" includes digital information expressed in forms such as text, images, and audio.

[0476] "Means of acquisition" refers to the processes and technologies used to receive information data from users.

[0477] "Analysis techniques" refer to methods such as natural language processing and multimodal analysis used to evaluate received information data.

[0478] "Means of assessing aggression" refers to criteria and techniques for determining inappropriate content or signs of bullying in information data.

[0479] "Means of providing warnings" refers to processes and technologies for notifying users of information when an attack is detected.

[0480] "Gradual warnings" refer to the practice of issuing different levels of caution or warnings to users when bullying or inappropriate behavior is observed, depending on the severity of the situation.

[0481] "Advice services" refer to resources and platforms that provide counseling and support to users who may have engaged in inappropriate behavior.

[0482] This invention is an information processing device for detecting bullying behavior in user-to-user communication in real time and providing appropriate responses. This is achieved by linking the user's terminal with a server.

[0483] The server receives encrypted information data, including text, image, and audio data, transmitted from the user's device. The user's device is responsible for generating this data and securely transferring it to the server.

[0484] The server utilizes natural language processing and deep learning-based multimodal analysis techniques to analyze the received data. This analysis uses state-of-the-art generative AI models to evaluate signs of aggression and inappropriate behavior in the data. Specifically, it understands the context of text messages and analyzes sentiment and tone. For image data, image analysis techniques are applied to check for the presence of inappropriate content. Similarly, audio data is converted to text using speech recognition technology and then subjected to text analysis.

[0485] For example, in a chat between user A and user B, if user A's message is deemed offensive, the server analyzes the message and compares its level of aggression to a baseline. If the baseline is exceeded, the server immediately sends a warning message to user A and provides a link to an advisory service if necessary. This warning message is designed to have different content depending on the level of aggression.

[0486] (Example of a prompt message)

[0487] "Evaluate whether User A's message is offensive, and if it is determined to be bullying, issue a warning."

[0488] This allows the system to detect malicious activity between users early on, helping to ensure a continuous and secure communication environment.

[0489] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0490] Step 1:

[0491] The server receives information data transmitted from the user's terminal. As input, the server obtains encrypted text data, image data, and audio data from the terminal. This data originates from communication between users. The server decrypts the encrypted data and prepares it for the next analysis step.

[0492] Step 2:

[0493] The server analyzes the received text data using natural language processing techniques. This process uses the decoded text data as input, employs a generative AI model to understand the context, and performs sentiment and tone analysis. The output is an aggression score, which is compared against criteria to determine if the text constitutes bullying.

[0494] Step 3:

[0495] The server performs image analysis on image data using deep learning. It uses decoded image data as input to evaluate whether the image contains inappropriate content. The output includes a determination of whether the image contains inappropriate elements and a score based on that evaluation. This analysis contributes to determining the level of aggression.

[0496] Step 4:

[0497] The server converts audio data into text using speech recognition technology and then performs text analysis on the converted data. Decoded audio data is used as input, and a generative AI model is used for the conversion process. The resulting text is analyzed using the same method as in step 2 to evaluate its aggression potential.

[0498] Step 5:

[0499] The server comprehensively analyzes text, images, and audio data and, if it detects signs of bullying, sends a warning message to the user in question. The analysis results for each data format are used as input, and if the level of aggression exceeds a predetermined threshold, a warning is issued to the user, and a link to an advisory service is provided as needed. The output generates the warning message the user receives and the advisory service recommendations.

[0500] (Application Example 1)

[0501] 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."

[0502] There is a need to proactively detect aggressive or inappropriate behavior in online communication within the home, ensuring user safety and providing a smooth communication environment. However, current technology lacks the means to do this in real time, making it difficult to monitor and control digital communication within the home.

[0503] 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.

[0504] In this invention, the server includes means for receiving data, image information, and sound information acquired from an information terminal; means for analyzing the received data using language processing technology and various data format analysis technologies to evaluate the threat level in information exchange; means for providing a warning to the user if the threat level exceeds a set standard based on the analysis results; means for recommending advisory services to the user based on the warning; and means for monitoring the security of digital communications within the home. This makes it possible to detect inappropriate communication behavior that may occur within the home in real time and take appropriate action.

[0505] An "information terminal" is a device used by users to send and receive digital information, and mainly includes smartphones and tablets.

[0506] "Data" refers to a collection of information obtained from users, and includes various formats such as text, images, and audio.

[0507] "Image information" refers to data formats that include visual information, such as photographs and illustrations.

[0508] "Acoustic information" refers to a data format that includes speech and sound, and includes conversations and other sound sources.

[0509] "Language processing technology" refers to technologies that perform analysis and understanding of natural language, including sentiment analysis and tone evaluation of text.

[0510] "Diverse data format analysis technology" refers to technology that comprehensively analyzes data in different formats, making it possible to evaluate the interrelationships between text, images, and audio.

[0511] "Threat level" refers to the degree of aggressive or inappropriate behavior or content in digital communication.

[0512] A "criteria" is a comparative value used to assess the level of threat, and a warning is issued when that value is exceeded.

[0513] A "warning" is a cautionary action given to a user, urging them to correct inappropriate behavior.

[0514] "Advice services" are services that provide users with counseling and other forms of support.

[0515] "Monitoring the security of digital communications" refers to continuous checks conducted to ensure that online communication within the home is conducted appropriately.

[0516] In an embodiment of the present invention, when digital communication is initiated by a user's information terminal, the terminal collects text, image, and audio data in real time and securely encrypts this data end-to-end. Next, the collected data is sent to a server, which performs analysis using natural language processing and multimodal analysis techniques. Specifically, the server uses Hugging Face's Transformers to perform sentiment and tone analysis of text, OpenCV to perform image analysis, and the Google Cloud Speech-to-Text API to transcribe and analyze audio data.

[0517] Based on the analysis results, the server assesses the threat level of the information exchange and immediately sends a warning message to the user's device if the threshold is exceeded. Furthermore, if necessary, it provides appropriate advice services by presenting the user with a link to a counseling service.

[0518] For example, if a child makes an inappropriate comment while playing an online game, that comment may be recognized as an attack on a friend. In this case, the system immediately notifies the parent and displays a warning on the child's device. In this way, the system can monitor and ensure that digital communication within the home is conducted safely.

[0519] An example of a prompt message would be: "Analyze the following chat message and evaluate whether it contains offensive content: 'I hate you, never talk to me again!'"

[0520] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0521] Step 1:

[0522] When an information terminal initiates digital communication, it collects text, image, and audio data in real time. Input includes user messages, transmitted images, and recorded audio. The data is encrypted end-to-end and securely transmitted to a server over the network.

[0523] Step 2:

[0524] The server classifies the received data, and the text data is fed into a natural language processing model (Hugging Face's Transformers). The input is encrypted text data, and prompt sentences are used to analyze sentiment and tone. The output is an aggression and threat rating score.

[0525] Step 3:

[0526] Image data is analyzed by the server using OpenCV. The input is encrypted image data, and various filters and models are applied to each image to determine whether it contains inappropriate content. The output is a determination of whether or not the image contains inappropriate content.

[0527] Step 4:

[0528] The audio data is converted to text by the server using the Google Cloud Speech-to-Text API and then fed into a parsing pipeline for text analysis. The input is encrypted audio data, which is then converted to text. The output is text data in a parseable format.

[0529] Step 5:

[0530] The server integrates evaluation scores and judgment results obtained from each data format to perform a comprehensive threat analysis. The input consists of individual evaluations from each data set, which are then combined to generate the final threat score. The output provides a final judgment on whether or not the threshold is exceeded.

[0531] Step 6:

[0532] If the server determines that the threat level exceeds the threshold, it sends a warning message to the device. The input is the result of the threat assessment, and the output is a message prompting the user to take action. A pop-up message will appear on the device, requesting that the inappropriate behavior be corrected.

[0533] Step 7:

[0534] If necessary, the server will offer the user additional advice services. Input is the user's response and actions after the warning, and output is provided with links and information about counseling services. This ensures continued support for the user.

[0535] 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.

[0536] This invention is a system that analyzes data acquired from users from multiple perspectives and detects aggression in communications between users with high accuracy. This system incorporates an emotion engine and has functions that enhance the recognition and analysis of emotions compared to conventional systems.

[0537] System Configuration

[0538] In this system, the server and the user's terminal are the main components. The terminal collects text data, image data, and audio data generated by the user, encrypts them, and sends them to the server. The server analyzes the received data using a multi-functional analysis engine. Specifically, it first analyzes the text using natural language processing technology. Next, image and audio data are processed using multimodal analysis. Furthermore, an emotion engine is incorporated to recognize the user's emotional state from this data and reinforce the analysis results.

[0539] The server evaluates the aggressiveness of the communication based on the output of the emotion engine and determines whether a warning is necessary. If aggression is detected, a warning message is generated for the user. This warning includes step-by-step instructions and, if necessary, provides a link to counseling services. The emotion engine data is also provided to the counseling services, allowing for more individualized and appropriate support.

[0540] Specific example

[0541] For example, suppose user X and user Y are communicating through chat. When user X types a specific phrase, the emotion expressing that feeling is detected from the text. The server, using its emotion engine, detects that user X is experiencing strong anger. If this emotion is determined to be aggressive, the system immediately sends a warning to user X saying, "Please be careful. This communication may be aggressive."

[0542] Furthermore, when user Z shares an image, the image analysis and sentiment engine work together to detect if the image may contain an inappropriate message. As a result, user Z receives a warning and can be guided to counseling services where they can seek support if necessary.

[0543] Thus, by integrating with an emotion engine, this system goes beyond mere data analysis to understand the user's emotional state and achieve more accurate responses.

[0544] The following describes the processing flow.

[0545] Step 1:

[0546] The device collects user-generated text, image, and audio data in real time. This data is collected by the device during user communication and is end-to-end encrypted to prepare it for secure data transmission.

[0547] Step 2:

[0548] The device sends encrypted data to the server. This transmission occurs in real time, ensuring that conversations between users continue smoothly.

[0549] Step 3:

[0550] After decrypting the received encrypted data, the server applies natural language processing and multimodal analysis techniques for data analysis. For text data, natural language processing is performed to identify context, keywords, and sentiment tone.

[0551] Step 4:

[0552] The server analyzes image data using an image recognition algorithm to check for any inappropriate content. Audio data is converted to text using speech recognition technology and then analyzed similarly as text data.

[0553] Step 5:

[0554] The server then uses an emotion engine to recognize the user's emotional state from all the collected data. This emotional information is integrated with other analysis results to provide a multifaceted assessment of the aggressiveness of the communication.

[0555] Step 6:

[0556] The server calculates an attack score based on the analysis results and generates a warning message if it exceeds a threshold. The warning is adjusted to an appropriate level depending on the situation, and in some cases, a link to a counseling service is also provided.

[0557] Step 7:

[0558] The terminal displays a warning message sent from the server to the user. The user can review the warning and choose to utilize the suggested counseling service.

[0559] Step 8:

[0560] When a user provides feedback, the device sends that feedback to the server. The server uses this feedback to refine its analysis techniques and improve the overall accuracy of the system.

[0561] (Example 2)

[0562] 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."

[0563] In modern digital communication, aggressive messages and inappropriate content are sometimes exchanged between users. This can degrade the quality of communication and potentially damage relationships. Traditional systems have struggled to detect and appropriately address this aggression, making it difficult to prevent user stress and problems.

[0564] 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.

[0565] In this invention, the server includes means for receiving information data extracted from the user's terminal, means for utilizing a generative AI model with the received information data to perform natural language processing and complex data analysis to identify emotional states in communication and evaluate aggression using an improved emotion engine, and means for providing the user with step-by-step warnings and referrals to counseling services if the analysis results show that aggression exceeds a certain threshold. This makes it possible to detect aggressive communication between users with high accuracy and provide prompt and appropriate feedback and support.

[0566] A "user's terminal" is an electronic device used by a user to generate information and communicate with the system.

[0567] "Means of receiving" refers to a function or device for receiving data transmitted from a user's terminal.

[0568] A "generative AI model" is a form of artificial intelligence that learns from large amounts of data and enables the understanding of natural language and the analysis of images and sounds.

[0569] "Natural language processing" is a technology that enables computers to understand and analyze the language that humans use in everyday life.

[0570] "Combined data analysis" is a process that integrates and analyzes data in multiple formats, such as text, images, and audio.

[0571] An "emotion engine" is a specific algorithm or model used to analyze and evaluate emotional elements from user data.

[0572] A "means for evaluating aggression" refers to a function that identifies elements indicating hostility or aggression from data and measures their degree.

[0573] A "threshold" is a standard value or boundary that is required for a certain condition to be met.

[0574] A "warning" is a message issued by a system to draw the user's attention.

[0575] "Counseling services" are services that provide professional support to alleviate the problems and stress that users face.

[0576] "Protective measures" refer to technologies and processes used to ensure the privacy and security of information.

[0577] To implement this invention, it is necessary to utilize a user's terminal and a server. The user's terminal first collects information data such as text data, image data, and audio data. This information is generated through everyday communication tools. The terminal encrypts this information data using AES encryption technology to ensure security, and then transmits it to the server.

[0578] The server decrypts the received encrypted information data and analyzes it using a generative AI model. It analyzes text data using natural language processing techniques to determine grammatical structure and emotional tone. Furthermore, it performs composite data analysis on image data using computer vision techniques to identify features and objects within the images. For audio data, it uses speech recognition technology to convert speech to text and analyze tone.

[0579] The server incorporates an emotion engine that comprehensively evaluates the user's emotional state based on information extracted from natural language processing and complex data analysis. The emotion engine plays a particularly important role in assessing aggression. If aggression is detected, the server immediately generates a series of warning messages for the user and provides a link to a counseling service. Furthermore, as a protective measure, the system is designed to maintain privacy throughout the data analysis process.

[0580] (Specific example)

[0581] For example, when user A types a phrase like "I really can't stand this!" in an online chat, the device encrypts the message and sends it to the server. The server receives this message through natural language processing and detects strong anger from the phrase "can't stand it." If the emotion engine determines that this strong emotion may be aggressive towards other users, it immediately sends a warning message to user A such as "Your emotions are running high. Please calm down." At this point, user A is also provided with a link to a counseling service if necessary.

[0582] (Example of a prompt message)

[0583] "When a user sends a message while in a state where they cannot control their emotions, please generate an appropriate warning message."

[0584] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0585] Step 1:

[0586] The device collects information such as text, images, and audio generated by the user. During this process, data is captured on the device based on the user's input actions. The collected data is encrypted using AES encryption technology to protect user privacy. The input consists of raw data resulting from user actions, and the output is encrypted data.

[0587] Step 2:

[0588] The terminal sends encrypted information data to the server. This transmission process utilizes security protocols to ensure the safe transmission of data. The input is encrypted information data, and the output is encrypted data that reaches the server.

[0589] Step 3:

[0590] The server decrypts the received encrypted data. This decryption process separates the text, image, and audio data formats, making them available for individual analysis. The input is encrypted information data, and the output is decrypted information data.

[0591] Step 4:

[0592] The server performs natural language processing on the decoded text data using a generative AI model. Specifically, it performs syntactic analysis and sentiment assessment of the text. The input is the decoded text data, and the output is the sentiment analysis result of the text.

[0593] Step 5:

[0594] The server analyzes image data using computer vision technology. This analysis detects objects and text elements within the image. The input is decoded image data, and the output is the image analysis results.

[0595] Step 6:

[0596] The server analyzes audio data using speech recognition technology, converting the speech to text and evaluating the speech tone. The input is decoded audio data, and the output is the speech-to-text conversion result and the tone evaluation result.

[0597] Step 7:

[0598] The server integrates the results of natural language processing, image analysis, and speech analysis, and uses an emotion engine to evaluate the user's overall emotional state. This process specifically assesses aggression. The input consists of various analysis results, and the output is the aggression evaluation result.

[0599] Step 8:

[0600] If an attack is detected, the server generates a series of warning messages for the user, including information about counseling services. The input is the attack assessment result, and the output is a warning message and guidance.

[0601] Step 9:

[0602] The server sends the generated warning message to the user's terminal. Immediacy and delivery confirmation of the message are crucial in this process. The input is a warning message, and the output is the warning message displayed on the user's terminal.

[0603] (Application Example 2)

[0604] 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."

[0605] Online verbal aggression has become a social problem. The objective of this invention is to maintain a safe and healthy communication environment by detecting aggression in online communication with high accuracy and issuing timely warnings to users.

[0606] 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.

[0607] In this invention, the server includes a configuration means for receiving information, still images, and audio information extracted from a user; a configuration means for processing the received information using natural language processing technology and integrated analysis technology to evaluate the aggression in the dialogue; and a configuration means for monitoring user statements in real time on an online exchange platform and immediately notifying the user if aggression is detected. This makes it possible to quickly and accurately evaluate the aggression of language, images, and audio uttered by users online and provide appropriate warnings.

[0608] "Information extracted from users" refers to text, still images, and audio information generated or shared by users.

[0609] A "still image" is digital data representing a visual expression captured or shared by a user.

[0610] "Audio information" refers to audio data that a user has spoken or recorded.

[0611] "Natural language processing technology" is a technology that enables computers to understand human language, and involves the analysis and interpretation of text data.

[0612] "Integrated analysis techniques" are technologies that integrate and analyze multiple data formats to gain deeper insights.

[0613] An "online exchange platform" refers to a virtual space on the internet where users can exchange information and opinions with each other.

[0614] Aggression is an index that assesses hostile or harmful intentions or attitudes expressed through language, images, or sounds.

[0615] A "real-time monitoring configuration" refers to a system mechanism that instantly monitors user activity and analyzes events that occur almost simultaneously.

[0616] "Configuration for notifying users" refers to a function that immediately informs users of information detected by the system.

[0617] The system for realizing this invention consists of a user terminal and a server. The user terminal has the function of collecting text, still images, and audio information generated or shared by the user, encrypting them, and sending them to the server. For natural language processing, Python and its libraries, NLTK and spaCy, can be used, and for sentiment analysis, Hugging Face's Transformers are used. This makes it possible to effectively detect the aggressiveness of the user's statements.

[0618] The server receives encrypted data and analyzes the information. Here, natural language processing techniques are used to analyze text, and integrated analysis techniques are used to evaluate still images and audio information. This makes it possible to assess the aggression level of the communication content in real time and with high accuracy.

[0619] For example, if a user makes an inappropriate comment towards another user in an online forum, and the server determines that the comment is offensive, a warning message such as "This comment is inappropriate. Please choose better words to maintain the quality of communication" will be sent immediately. This is expected to encourage users to be more mindful of their online communication.

[0620] An example of a prompt message is, "Analyze the user's posts and evaluate the aggression level. If it is deemed aggression, generate a warning message." This is how instructions can be given to the generation AI model. By using this prompt message, the system can efficiently evaluate the aggression level between users and issue warnings.

[0621] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0622] Step 1:

[0623] The user's device collects text, still images, and audio information generated or shared by the user. This data is encrypted within the device to ensure security and prepared for transmission to the server. Data input occurs through the user interface, and output is in the form of encrypted data files.

[0624] Step 2:

[0625] The terminal sends encrypted data to the server. Secure communication methods such as the SSL / TLS protocol are used to prevent data eavesdropping and tampering. The input is an encrypted data file, and the output is the completion of the data transfer to the server.

[0626] Step 3:

[0627] The server decrypts the received encrypted data, thereby obtaining text, still images, and audio information in their original form. The input is an encrypted data file, and the output is the decrypted raw data.

[0628] Step 4:

[0629] The server analyzes text data using natural language processing techniques. Specifically, it uses Hugging Face's Transformers to extract emotions from the text and evaluate their aggression level. The input is text data, and the output is an aggression evaluation score.

[0630] Step 5:

[0631] The server processes decoded still images and audio information using integrated analysis techniques to assess their aggression. This is done using image recognition and audio analysis techniques to identify the meaning and emotion of each data point. The input is still images and audio information, and the output is an evaluation score of their aggression.

[0632] Step 6:

[0633] The server integrates evaluation scores from text, still images, and audio information to calculate an overall level of aggression. Based on this evaluation, it generates warnings and instructions if necessary. The input is the aggression score from each medium, and the output is the overall aggression score and warning messages.

[0634] Step 7:

[0635] The server sends the generated warning message to the user's terminal. The warning message is created using a generation AI model based on the prompt text and is provided in a user-friendly format. The input is the overall attack score, and the output is the warning message displayed on the user interface.

[0636] 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.

[0637] 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.

[0638] 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.

[0639] [Fourth Embodiment]

[0640] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0641] 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.

[0642] 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).

[0643] 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.

[0644] 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.

[0645] 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).

[0646] 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.

[0647] 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.

[0648] 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.

[0649] 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.

[0650] 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.

[0651] 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.

[0652] 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".

[0653] This invention is a system for detecting bullying behavior in real time during user-to-user communication and providing appropriate responses. This system is implemented through a series of processes that operate on a server and user terminals.

[0654] System Configuration

[0655] The server receives text, image, and audio data collected from users in real time. This data is encrypted and transmitted from the user's device as each communication progresses. The server analyzes the received data using the latest natural language processing and multimodal analysis technologies. In this process, the server performs sentiment analysis and tone analysis to assess for signs of bullying or inappropriate content.

[0656] If the server detects signs of bullying, it will send a warning message to the user in question. The warnings are gradual, starting with minor cautions and progressing to more detailed warnings and suggestions for connecting to counseling services in more serious cases. Users can view these messages on their devices and utilize the suggested counseling services.

[0657] Specific example

[0658] For example, suppose user A and user B are exchanging messages. If user A's message is perceived as offensive towards the other user, the server analyzes the text and determines whether the level of aggression exceeds a certain threshold. If aggression is detected, the server immediately sends a warning to user A, and if it persists, provides a detailed warning along with a link to a counseling service.

[0659] Furthermore, if user C shares an inappropriate image in the chat, the server will analyze the image and, if it determines it to be inappropriate, will issue a warning to user C. By comprehensively handling various data formats in this way, the system can detect bullying in diverse situations and enable appropriate responses. This allows for the early detection of malicious behavior on the internet and ensures a safe communication environment for users.

[0660] The following describes the processing flow.

[0661] Step 1:

[0662] The device collects communication data between users, including chat messages, sent image files, and recorded audio data. This data is prepared to be securely processed using end-to-end encryption technology, with the user's consent.

[0663] Step 2:

[0664] The device sends the collected data to the server. The data is processed in real time, minimizing delays in communication between users.

[0665] Step 3:

[0666] The server first decodes the received data. Then, it uses natural language processing techniques to analyze the text data and detect aggressive expressions or negative emotions within the text. Image data is analyzed using image recognition algorithms, and audio data is converted to text using speech recognition technology before being analyzed.

[0667] Step 4:

[0668] The server integrates the analysis results for each data type and scores the signs of bullying based on each data point. If the score exceeds a certain threshold, an alert is generated.

[0669] Step 5:

[0670] The server generates a warning message for a user if their bullying score exceeds a certain threshold. This message includes a step-by-step warning based on the user's behavior.

[0671] Step 6:

[0672] The terminal displays the warning message received from the server to the relevant user. The user can then review the warning details and consider using the suggested counseling service.

[0673] Step 7:

[0674] Users can provide feedback regarding the content or accuracy of system warnings. This feedback is aggregated by the server and used to improve the accuracy of the system's analysis.

[0675] (Example 1)

[0676] 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".

[0677] In today's information society, online communication is increasing more and more. However, bullying and inappropriate behavior among users are becoming a problem. In particular, because these behaviors unfold in real time, early detection and response are required. However, current technology does not have a sufficient means to do so. Therefore, in order to ensure a safe and healthy communication environment, there is a need for a system that can perform more advanced analysis and respond flexibly.

[0678] 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.

[0679] In this invention, the server includes means for acquiring information data extracted from users, means for analyzing the received information data using analysis techniques to evaluate aggression within interpersonal relationships, and means for providing a warning to the user if the analysis results in aggression exceeding a specified standard. This makes it possible to detect bullying between users in real time, provide staged warnings, and build a safe communication environment that enables prompt response and advice.

[0680] "User" refers to a person who uses the system to communicate.

[0681] "Information data" includes digital information expressed in forms such as text, images, and audio.

[0682] "Means of acquisition" refers to the processes and technologies used to receive information data from users.

[0683] "Analysis techniques" refer to methods such as natural language processing and multimodal analysis used to evaluate received information data.

[0684] "Means of assessing aggression" refers to criteria and techniques for determining inappropriate content or signs of bullying in information data.

[0685] "Means of providing warnings" refers to processes and technologies for notifying users of information when an attack is detected.

[0686] "Gradual warnings" refer to the practice of issuing different levels of caution or warnings to users when bullying or inappropriate behavior is observed, depending on the severity of the situation.

[0687] "Advice services" refer to resources and platforms that provide counseling and support to users who may have engaged in inappropriate behavior.

[0688] This invention is an information processing device for detecting bullying behavior in user-to-user communication in real time and providing appropriate responses. This is achieved by linking the user's terminal with a server.

[0689] The server receives encrypted information data, including text, image, and audio data, transmitted from the user's device. The user's device is responsible for generating this data and securely transferring it to the server.

[0690] The server utilizes natural language processing and deep learning-based multimodal analysis techniques to analyze the received data. This analysis uses state-of-the-art generative AI models to evaluate signs of aggression and inappropriate behavior in the data. Specifically, it understands the context of text messages and analyzes sentiment and tone. For image data, image analysis techniques are applied to check for the presence of inappropriate content. Similarly, audio data is converted to text using speech recognition technology and then subjected to text analysis.

[0691] For example, in a chat between user A and user B, if user A's message is deemed offensive, the server analyzes the message and compares its level of aggression to a baseline. If the baseline is exceeded, the server immediately sends a warning message to user A and provides a link to an advisory service if necessary. This warning message is designed to have different content depending on the level of aggression.

[0692] (Example of a prompt message)

[0693] "Evaluate whether User A's message is offensive, and if it is determined to be bullying, issue a warning."

[0694] This allows the system to detect malicious activity between users early on, helping to ensure a continuous and secure communication environment.

[0695] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0696] Step 1:

[0697] The server receives information data transmitted from the user's terminal. As input, the server obtains encrypted text data, image data, and audio data from the terminal. This data originates from communication between users. The server decrypts the encrypted data and prepares it for the next analysis step.

[0698] Step 2:

[0699] The server analyzes the received text data using natural language processing techniques. This process uses the decoded text data as input, employs a generative AI model to understand the context, and performs sentiment and tone analysis. The output is an aggression score, which is compared against criteria to determine if the text constitutes bullying.

[0700] Step 3:

[0701] The server performs image analysis on image data using deep learning. It uses decoded image data as input to evaluate whether the image contains inappropriate content. The output includes a determination of whether the image contains inappropriate elements and a score based on that evaluation. This analysis contributes to determining the level of aggression.

[0702] Step 4:

[0703] The server converts audio data into text using speech recognition technology and then performs text analysis on the converted data. Decoded audio data is used as input, and a generative AI model is used for the conversion process. The resulting text is analyzed using the same method as in step 2 to evaluate its aggression potential.

[0704] Step 5:

[0705] The server comprehensively analyzes text, images, and audio data and, if it detects signs of bullying, sends a warning message to the user in question. The analysis results for each data format are used as input, and if the level of aggression exceeds a predetermined threshold, a warning is issued to the user, and a link to an advisory service is provided as needed. The output generates the warning message the user receives and the advisory service recommendations.

[0706] (Application Example 1)

[0707] 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".

[0708] There is a need to proactively detect aggressive or inappropriate behavior in online communication within the home, ensuring user safety and providing a smooth communication environment. However, current technology lacks the means to do this in real time, making it difficult to monitor and control digital communication within the home.

[0709] 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.

[0710] In this invention, the server includes means for receiving data, image information, and sound information acquired from an information terminal; means for analyzing the received data using language processing technology and various data format analysis technologies to evaluate the threat level in information exchange; means for providing a warning to the user if the threat level exceeds a set standard based on the analysis results; means for recommending advisory services to the user based on the warning; and means for monitoring the security of digital communications within the home. This makes it possible to detect inappropriate communication behavior that may occur within the home in real time and take appropriate action.

[0711] An "information terminal" is a device used by users to send and receive digital information, and mainly includes smartphones and tablets.

[0712] "Data" refers to a collection of information obtained from users, and includes various formats such as text, images, and audio.

[0713] "Image information" refers to data formats that include visual information, such as photographs and illustrations.

[0714] "Acoustic information" refers to a data format that includes speech and sound, and includes conversations and other sound sources.

[0715] "Language processing technology" refers to technologies that perform analysis and understanding of natural language, including sentiment analysis and tone evaluation of text.

[0716] "Diverse data format analysis technology" refers to technology that comprehensively analyzes data in different formats, making it possible to evaluate the interrelationships between text, images, and audio.

[0717] "Threat level" refers to the degree of aggressive or inappropriate behavior or content in digital communication.

[0718] A "criteria" is a comparative value used to assess the level of threat, and a warning is issued when that value is exceeded.

[0719] A "warning" is a cautionary action given to a user, urging them to correct inappropriate behavior.

[0720] "Advice services" are services that provide users with counseling and other forms of support.

[0721] "Monitoring the security of digital communications" refers to continuous checks conducted to ensure that online communication within the home is conducted appropriately.

[0722] In an embodiment of the present invention, when digital communication is initiated by a user's information terminal, the terminal collects text, image, and audio data in real time and securely encrypts this data end-to-end. Next, the collected data is sent to a server, which performs analysis using natural language processing and multimodal analysis techniques. Specifically, the server uses Hugging Face's Transformers to perform sentiment and tone analysis of text, OpenCV to perform image analysis, and the Google Cloud Speech-to-Text API to transcribe and analyze audio data.

[0723] Based on the analysis results, the server assesses the threat level of the information exchange and immediately sends a warning message to the user's device if the threshold is exceeded. Furthermore, if necessary, it provides appropriate advice services by presenting the user with a link to a counseling service.

[0724] For example, if a child makes an inappropriate comment while playing an online game, that comment may be recognized as an attack on a friend. In this case, the system immediately notifies the parent and displays a warning on the child's device. In this way, the system can monitor and ensure that digital communication within the home is conducted safely.

[0725] An example of a prompt message would be: "Analyze the following chat message and evaluate whether it contains offensive content: 'I hate you, never talk to me again!'"

[0726] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0727] Step 1:

[0728] When an information terminal initiates digital communication, it collects text, image, and audio data in real time. Input includes user messages, transmitted images, and recorded audio. The data is encrypted end-to-end and securely transmitted to a server over the network.

[0729] Step 2:

[0730] The server classifies the received data, and the text data is fed into a natural language processing model (Hugging Face's Transformers). The input is encrypted text data, and prompt sentences are used to analyze sentiment and tone. The output is an aggression and threat rating score.

[0731] Step 3:

[0732] Image data is analyzed by the server using OpenCV. The input is encrypted image data, and various filters and models are applied to each image to determine whether it contains inappropriate content. The output is a determination of whether or not the image contains inappropriate content.

[0733] Step 4:

[0734] The audio data is converted to text by the server using the Google Cloud Speech-to-Text API and then fed into a parsing pipeline for text analysis. The input is encrypted audio data, which is then converted to text. The output is text data in a parseable format.

[0735] Step 5:

[0736] The server integrates evaluation scores and judgment results obtained from each data format to perform a comprehensive threat analysis. The input consists of individual evaluations from each data set, which are then combined to generate the final threat score. The output provides a final judgment on whether or not the threshold is exceeded.

[0737] Step 6:

[0738] If the server determines that the threat level exceeds the threshold, it sends a warning message to the device. The input is the result of the threat assessment, and the output is a message prompting the user to take action. A pop-up message will appear on the device, requesting that the inappropriate behavior be corrected.

[0739] Step 7:

[0740] If necessary, the server will offer the user additional advice services. Input is the user's response and actions after the warning, and output is provided with links and information about counseling services. This ensures continued support for the user.

[0741] 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.

[0742] This invention is a system that analyzes data acquired from users from multiple perspectives and detects aggression in communications between users with high accuracy. This system incorporates an emotion engine and has functions that enhance the recognition and analysis of emotions compared to conventional systems.

[0743] System Configuration

[0744] In this system, the server and the user's terminal are the main components. The terminal collects text data, image data, and audio data generated by the user, encrypts them, and sends them to the server. The server analyzes the received data using a multi-functional analysis engine. Specifically, it first analyzes the text using natural language processing technology. Next, image and audio data are processed using multimodal analysis. Furthermore, an emotion engine is incorporated to recognize the user's emotional state from this data and reinforce the analysis results.

[0745] The server evaluates the aggressiveness of the communication based on the output of the emotion engine and determines whether a warning is necessary. If aggression is detected, a warning message is generated for the user. This warning includes step-by-step instructions and, if necessary, provides a link to counseling services. The emotion engine data is also provided to the counseling services, allowing for more individualized and appropriate support.

[0746] Specific example

[0747] For example, suppose user X and user Y are communicating through chat. When user X types a specific phrase, the emotion expressing that feeling is detected from the text. The server, using its emotion engine, detects that user X is experiencing strong anger. If this emotion is determined to be aggressive, the system immediately sends a warning to user X saying, "Please be careful. This communication may be aggressive."

[0748] Furthermore, when user Z shares an image, the image analysis and sentiment engine work together to detect if the image may contain an inappropriate message. As a result, user Z receives a warning and can be guided to counseling services where they can seek support if necessary.

[0749] Thus, by integrating with an emotion engine, this system goes beyond mere data analysis to understand the user's emotional state and achieve more accurate responses.

[0750] The following describes the processing flow.

[0751] Step 1:

[0752] The device collects user-generated text, image, and audio data in real time. This data is collected by the device during user communication and is end-to-end encrypted to prepare it for secure data transmission.

[0753] Step 2:

[0754] The device sends encrypted data to the server. This transmission occurs in real time, ensuring that conversations between users continue smoothly.

[0755] Step 3:

[0756] After decrypting the received encrypted data, the server applies natural language processing and multimodal analysis techniques for data analysis. For text data, natural language processing is performed to identify context, keywords, and sentiment tone.

[0757] Step 4:

[0758] The server analyzes image data using an image recognition algorithm to check for any inappropriate content. Audio data is converted to text using speech recognition technology and then analyzed similarly as text data.

[0759] Step 5:

[0760] The server then uses an emotion engine to recognize the user's emotional state from all the collected data. This emotional information is integrated with other analysis results to provide a multifaceted assessment of the aggressiveness of the communication.

[0761] Step 6:

[0762] The server calculates an attack score based on the analysis results and generates a warning message if it exceeds a threshold. The warning is adjusted to an appropriate level depending on the situation, and in some cases, a link to a counseling service is also provided.

[0763] Step 7:

[0764] The terminal displays a warning message sent from the server to the user. The user can review the warning and choose to utilize the suggested counseling service.

[0765] Step 8:

[0766] When a user provides feedback, the device sends that feedback to the server. The server uses this feedback to refine its analysis techniques and improve the overall accuracy of the system.

[0767] (Example 2)

[0768] 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".

[0769] In modern digital communication, aggressive messages and inappropriate content are sometimes exchanged between users. This can degrade the quality of communication and potentially damage relationships. Traditional systems have struggled to detect and appropriately address this aggression, making it difficult to prevent user stress and problems.

[0770] 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.

[0771] In this invention, the server includes means for receiving information data extracted from the user's terminal, means for utilizing a generative AI model with the received information data to perform natural language processing and complex data analysis to identify emotional states in communication and evaluate aggression using an improved emotion engine, and means for providing the user with step-by-step warnings and referrals to counseling services if the analysis results show that aggression exceeds a certain threshold. This makes it possible to detect aggressive communication between users with high accuracy and provide prompt and appropriate feedback and support.

[0772] A "user's terminal" is an electronic device used by a user to generate information and communicate with the system.

[0773] "Means of receiving" refers to a function or device for receiving data transmitted from a user's terminal.

[0774] A "generative AI model" is a form of artificial intelligence that learns from large amounts of data and enables the understanding of natural language and the analysis of images and sounds.

[0775] "Natural language processing" is a technology that enables computers to understand and analyze the language that humans use in everyday life.

[0776] "Combined data analysis" is a process that integrates and analyzes data in multiple formats, such as text, images, and audio.

[0777] An "emotion engine" is a specific algorithm or model used to analyze and evaluate emotional elements from user data.

[0778] A "means for evaluating aggression" refers to a function that identifies elements indicating hostility or aggression from data and measures their degree.

[0779] A "threshold" is a standard value or boundary that is required for a certain condition to be met.

[0780] A "warning" is a message issued by a system to draw the user's attention.

[0781] "Counseling services" are services that provide professional support to alleviate the problems and stress that users face.

[0782] "Protective measures" refer to technologies and processes used to ensure the privacy and security of information.

[0783] To implement this invention, it is necessary to utilize a user's terminal and a server. The user's terminal first collects information data such as text data, image data, and audio data. This information is generated through everyday communication tools. The terminal encrypts this information data using AES encryption technology to ensure security, and then transmits it to the server.

[0784] The server decrypts the received encrypted information data and analyzes it using a generative AI model. It analyzes text data using natural language processing techniques to determine grammatical structure and emotional tone. Furthermore, it performs composite data analysis on image data using computer vision techniques to identify features and objects within the images. For audio data, it uses speech recognition technology to convert speech to text and analyze tone.

[0785] The server incorporates an emotion engine that comprehensively evaluates the user's emotional state based on information extracted from natural language processing and complex data analysis. The emotion engine plays a particularly important role in assessing aggression. If aggression is detected, the server immediately generates a series of warning messages for the user and provides a link to a counseling service. Furthermore, as a protective measure, the system is designed to maintain privacy throughout the data analysis process.

[0786] (Specific example)

[0787] For example, when user A types a phrase like "I really can't stand this!" in an online chat, the device encrypts the message and sends it to the server. The server receives this message through natural language processing and detects strong anger from the phrase "can't stand it." If the emotion engine determines that this strong emotion may be aggressive towards other users, it immediately sends a warning message to user A such as "Your emotions are running high. Please calm down." At this point, user A is also provided with a link to a counseling service if necessary.

[0788] (Example of a prompt message)

[0789] "When a user sends a message while in a state where they cannot control their emotions, please generate an appropriate warning message."

[0790] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0791] Step 1:

[0792] The device collects information such as text, images, and audio generated by the user. During this process, data is captured on the device based on the user's input actions. The collected data is encrypted using AES encryption technology to protect user privacy. The input consists of raw data resulting from user actions, and the output is encrypted data.

[0793] Step 2:

[0794] The terminal sends encrypted information data to the server. This transmission process utilizes security protocols to ensure the safe transmission of data. The input is encrypted information data, and the output is encrypted data that reaches the server.

[0795] Step 3:

[0796] The server decrypts the received encrypted data. This decryption process separates the text, image, and audio data formats, making them available for individual analysis. The input is encrypted information data, and the output is decrypted information data.

[0797] Step 4:

[0798] The server performs natural language processing on the decoded text data using a generative AI model. Specifically, it performs syntactic analysis and sentiment assessment of the text. The input is the decoded text data, and the output is the sentiment analysis result of the text.

[0799] Step 5:

[0800] The server analyzes image data using computer vision technology. This analysis detects objects and text elements within the image. The input is decoded image data, and the output is the image analysis results.

[0801] Step 6:

[0802] The server analyzes audio data using speech recognition technology, converting the speech to text and evaluating the speech tone. The input is decoded audio data, and the output is the speech-to-text conversion result and the tone evaluation result.

[0803] Step 7:

[0804] The server integrates the results of natural language processing, image analysis, and speech analysis, and uses an emotion engine to evaluate the user's overall emotional state. This process specifically assesses aggression. The input consists of various analysis results, and the output is the aggression evaluation result.

[0805] Step 8:

[0806] If an attack is detected, the server generates a series of warning messages for the user, including information about counseling services. The input is the attack assessment result, and the output is a warning message and guidance.

[0807] Step 9:

[0808] The server sends the generated warning message to the user's terminal. Immediacy and delivery confirmation of the message are crucial in this process. The input is a warning message, and the output is the warning message displayed on the user's terminal.

[0809] (Application Example 2)

[0810] 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".

[0811] Online verbal aggression has become a social problem. The objective of this invention is to maintain a safe and healthy communication environment by detecting aggression in online communication with high accuracy and issuing timely warnings to users.

[0812] 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.

[0813] In this invention, the server includes a configuration means for receiving information, still images, and audio information extracted from a user; a configuration means for processing the received information using natural language processing technology and integrated analysis technology to evaluate the aggression in the dialogue; and a configuration means for monitoring user statements in real time on an online exchange platform and immediately notifying the user if aggression is detected. This makes it possible to quickly and accurately evaluate the aggression of language, images, and audio uttered by users online and provide appropriate warnings.

[0814] "Information extracted from users" refers to text, still images, and audio information generated or shared by users.

[0815] A "still image" is digital data representing a visual expression captured or shared by a user.

[0816] "Audio information" refers to audio data that a user has spoken or recorded.

[0817] "Natural language processing technology" is a technology that enables computers to understand human language, and involves the analysis and interpretation of text data.

[0818] "Integrated analysis techniques" are technologies that integrate and analyze multiple data formats to gain deeper insights.

[0819] An "online exchange platform" refers to a virtual space on the internet where users can exchange information and opinions with each other.

[0820] Aggression is an index that assesses hostile or harmful intentions or attitudes expressed through language, images, or sounds.

[0821] A "real-time monitoring configuration" refers to a system mechanism that instantly monitors user activity and analyzes events that occur almost simultaneously.

[0822] "Configuration for notifying users" refers to a function that immediately informs users of information detected by the system.

[0823] The system for realizing this invention consists of a user terminal and a server. The user terminal has the function of collecting text, still images, and audio information generated or shared by the user, encrypting them, and sending them to the server. For natural language processing, Python and its libraries, NLTK and spaCy, can be used, and for sentiment analysis, Hugging Face's Transformers are used. This makes it possible to effectively detect the aggressiveness of the user's statements.

[0824] The server receives encrypted data and analyzes the information. Here, natural language processing techniques are used to analyze text, and integrated analysis techniques are used to evaluate still images and audio information. This makes it possible to assess the aggression level of the communication content in real time and with high accuracy.

[0825] For example, if a user makes an inappropriate comment towards another user in an online forum, and the server determines that the comment is offensive, a warning message such as "This comment is inappropriate. Please choose better words to maintain the quality of communication" will be sent immediately. This is expected to encourage users to be more mindful of their online communication.

[0826] An example of a prompt message is, "Analyze the user's posts and evaluate the aggression level. If it is deemed aggression, generate a warning message." This is how instructions can be given to the generation AI model. By using this prompt message, the system can efficiently evaluate the aggression level between users and issue warnings.

[0827] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0828] Step 1:

[0829] The user's device collects text, still images, and audio information generated or shared by the user. This data is encrypted within the device to ensure security and prepared for transmission to the server. Data input occurs through the user interface, and output is in the form of encrypted data files.

[0830] Step 2:

[0831] The terminal sends encrypted data to the server. Secure communication methods such as the SSL / TLS protocol are used to prevent data eavesdropping and tampering. The input is an encrypted data file, and the output is the completion of the data transfer to the server.

[0832] Step 3:

[0833] The server decrypts the received encrypted data, thereby obtaining text, still images, and audio information in their original form. The input is an encrypted data file, and the output is the decrypted raw data.

[0834] Step 4:

[0835] The server analyzes text data using natural language processing techniques. Specifically, it uses Hugging Face's Transformers to extract emotions from the text and evaluate their aggression level. The input is text data, and the output is an aggression evaluation score.

[0836] Step 5:

[0837] The server processes decoded still images and audio information using integrated analysis techniques to assess their aggression. This is done using image recognition and audio analysis techniques to identify the meaning and emotion of each data point. The input is still images and audio information, and the output is an evaluation score of their aggression.

[0838] Step 6:

[0839] The server integrates evaluation scores from text, still images, and audio information to calculate an overall level of aggression. Based on this evaluation, it generates warnings and instructions if necessary. The input is the aggression score from each medium, and the output is the overall aggression score and warning messages.

[0840] Step 7:

[0841] The server sends the generated warning message to the user's terminal. The warning message is created using a generation AI model based on the prompt text and is provided in a user-friendly format. The input is the overall attack score, and the output is the warning message displayed on the user interface.

[0842] 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.

[0843] 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.

[0844] 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.

[0845] 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.

[0846] 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.

[0847] 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.

[0848] 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.

[0849] 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.

[0850] 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."

[0851] 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.

[0852] 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.

[0853] 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.

[0854] 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.

[0855] 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.

[0856] 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.

[0857] 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.

[0858] 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.

[0859] 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.

[0860] 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.

[0861] 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.

[0862] 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.

[0863] The following is further disclosed regarding the embodiments described above.

[0864] (Claim 1)

[0865] A means for receiving text data, image data, and audio data extracted from a user,

[0866] A means of analyzing received data using natural language processing and multimodal analysis techniques to evaluate aggression in communication,

[0867] The analysis results in a means of providing a warning to the user when the level of aggression exceeds a predetermined threshold,

[0868] A means of suggesting counseling services to users based on warnings,

[0869] A system that includes this.

[0870] (Claim 2)

[0871] The system according to claim 1, which provides means for improving the accuracy of the analysis by collecting user feedback and adjusting the analysis technique.

[0872] (Claim 3)

[0873] A security means for performing end-to-end encryption and analysis of communication data, as described in claim 1.

[0874] "Example 1"

[0875] (Claim 1)

[0876] A means of obtaining information data extracted from users,

[0877] A means of evaluating aggression in interpersonal relationships by analyzing received information data using analytical techniques,

[0878] The analysis results indicate a means of providing a warning to the user if the level of aggression exceeds a specified threshold,

[0879] A means of presenting users with advice services based on warnings,

[0880] A means for generating and sending step-by-step warnings,

[0881] Information processing device including

[0882] (Claim 2)

[0883] The information processing apparatus according to claim 1, which provides means for improving the accuracy of analysis by collecting feedback from users and adjusting the analysis technique.

[0884] (Claim 3)

[0885] The information processing apparatus according to claim 1, which provides protection for analyzing information data while encrypting it from start to finish.

[0886] "Application Example 1"

[0887] (Claim 1)

[0888] A means for receiving data acquired from an information terminal, as well as image information and sound information,

[0889] A means for analyzing received data using language processing techniques and various data format analysis techniques to evaluate the threat level in information exchange,

[0890] Based on the analysis results, a means of alerting users when the threat level exceeds a set standard,

[0891] Means of recommending advisory services to users based on caution, and means of monitoring the security of digital communications within the home,

[0892] A system that includes this.

[0893] (Claim 2)

[0894] The system according to claim 1, which provides means for improving the accuracy of the analysis by collecting feedback from users and modifying the analysis technique.

[0895] (Claim 3)

[0896] A protection means for performing analysis while encrypting communication information between terminals, according to claim 1.

[0897] "Example 2 of combining an emotion engine"

[0898] (Claim 1)

[0899] A means for receiving information data extracted from the user's device,

[0900] A means to utilize a generative AI model using received information data, perform natural language processing and complex data analysis to identify emotional states in communication, and evaluate aggression using an improved emotion engine,

[0901] The analysis revealed a means to provide users with step-by-step warnings and referrals to counseling services when aggression exceeds a certain threshold.

[0902] A means of offering additional assistance to the user based on the warning,

[0903] A system that includes this.

[0904] (Claim 2)

[0905] The system according to claim 1, which means improving the accuracy of analysis and emotion recognition by collecting user responses and adjusting the generation AI model and analysis techniques.

[0906] (Claim 3)

[0907] A protection means for performing end-to-end encryption and analysis of information data, as described in claim 1.

[0908] "Application example 2 when combining with an emotional engine"

[0909] (Claim 1)

[0910] A configuration means for receiving information, still images, and audio information extracted from the user,

[0911] A configuration means for processing received information using natural language processing technology and integrated analysis technology to evaluate aggression in dialogue,

[0912] A configuration means that provides a warning to the user if the processing results in an attack exceeding a set threshold,

[0913] A configuration means for suggesting support services to the user based on a warning,

[0914] An online exchange platform includes a configuration that monitors user statements in real time and immediately notifies users if aggression is detected,

[0915] A system that includes this.

[0916] (Claim 2)

[0917] The system according to claim 1, wherein the accuracy of processing is improved by collecting feedback from users and adjusting the processing technology.

[0918] (Claim 3)

[0919] A protection configuration for processing dialogue information while fully encrypting it mutually, according to claim 1. [Explanation of Symbols]

[0920] 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. A means for receiving text data, image data, and audio data extracted from a user, A means of analyzing received data using natural language processing and multimodal analysis techniques to evaluate aggression in communication, The analysis results in a means of providing a warning to the user when the level of aggression exceeds a predetermined threshold, A means of suggesting counseling services to users based on warnings, A system that includes this.

2. The system according to claim 1, further comprising means for improving the accuracy of the analysis by collecting user feedback and adjusting the analysis technique.

3. The system according to claim 1, including security means for performing end-to-end encryption and analysis of communication data.

Images