system

Abstract

We provide the system. [Solution] A means of acquiring audio and video data on the user's terminal, A means of preprocessing the acquired data and sending it to the server, A method for determining the risk of harassment by performing natural language processing and sentiment analysis on a server, A means of visualizing the judgment results and providing them to the user as feedback, A system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, the method including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, 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 modern society, various types of harassment have a negative impact on people's physical and mental health and have become a major problem hindering the freedom of communication. In particular, since the recognition of harassment is subjective and depends on how the victim feels, its determination is ambiguous. Due to this ambiguity, unnecessary self-censorship due to fear of harassment and atrophy of dialogue have occurred. The present invention aims to solve such problems and promote sound communication.

Means for Solving the Problems

[0005] To address this challenge, the present invention provides a system that acquires audio and video data at a user terminal, preprocesses the data, and then transmits it to a server. The server analyzes the data using natural language processing and sentiment analysis to determine the risk of harassment. The system visualizes the results of this assessment and provides feedback to the user, supporting real-time behavioral improvement based on scores of discomfort and anxiety. This system aims to prevent harassment and facilitate smooth communication.

[0006] "Audio and video data" refers to digital data acquired as audio and video in order to record the user's speech and actions.

[0007] A "user terminal" refers to an electronic device used to acquire audio and video data and transmit that data to a server.

[0008] "Preprocessing" refers to the data formatting process performed on audio and video data to make it easier to analyze.

[0009] A "server" refers to a central computer system that receives data sent from user terminals, performs analysis, and provides feedback on the results.

[0010] "Natural language processing" is a technology that enables computers to understand human language and analyze its meaning.

[0011] "Emotional analysis" is a technology that evaluates a user's emotional state from data such as voice and facial expressions.

[0012] "Harassment risk" is an indicator that shows the degree to which a particular statement or action is likely to constitute harassment.

[0013] "Visualization" refers to the process of displaying analysis results in a way that is easy for users to understand.

[0014] "Feedback" refers to advice or information based on the analysis results provided by the system to the user.

[0015] "Score of discomfort or anxiety" is an indicator of the user's emotional state quantitatively expressed based on the analysis results.

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] 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 to which multiple emotions are mapped. [Figure 10] It shows an emotion map to which 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 the data processing system in Application Example 2 when the emotion engine is combined.

Mode for Carrying Out the Invention

[0017] Hereinafter, an example of an embodiment of the 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, the numbered processor (hereinafter simply referred to as "processor") may be one arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be one 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, the 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, the 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 relates to a system that acquires audio and video data on a user's terminal, analyzes that data on a server, and determines the risk of harassment. This system assists user communication and plays a role in preventing harassment from occurring.

[0038] The user uses a device to collect audio and video data during a conversation. The device preprocesses this data and sends it to a server. The server uses natural language processing techniques and sentiment analysis algorithms to analyze the received data.

[0039] In natural language processing technology, the server analyzes the text data of the conversation to understand the meaning and context of the statements. This allows it to determine whether specific phrases or expressions constitute harassment. Furthermore, sentiment analysis algorithms are used to evaluate voice tone and facial expression data, quantifying the user's emotional state. Based on this data, the server scores the risk of harassment and visualizes it.

[0040] As a concrete example, consider a situation where a user is speaking in a workplace meeting. In this situation, the user's device records the conversation and sends it to the server in real time. The server quickly analyzes the statements made during the meeting and determines whether certain statements are offensive or whether other participants are showing discomfort with the statements. If the risk of harassment increases as a result, the server immediately sends an alert to the user's device to draw their attention to the issue.

[0041] The user's device provides this feedback visually or audibly, allowing the user to adjust the tone and content of the conversation in real time. This facilitates smoother communication and ensures that all participants can comfortably continue the conversation.

[0042] This system is designed for use in a variety of environments, including workplaces, educational institutions, and public organizations, and aims to contribute to creating an environment where people can communicate with peace of mind.

[0043] The following describes the processing flow.

[0044] Step 1:

[0045] The user uses their device to collect audio and video data during conversations. The device's microphone and camera capture the data in real time.

[0046] Step 2:

[0047] The device preprocesses the collected data. Noise reduction and filtering of unnecessary information are performed, and the audio data is converted into text data using speech recognition technology.

[0048] Step 3:

[0049] The terminal sends the pre-processed data to the server using a secure communication protocol. This data includes text data and audio / video features.

[0050] Step 4:

[0051] The server inputs the received text data into a natural language processing (NLP) model, which performs grammatical and semantic analysis. This helps the server understand the context in which the utterance is presented.

[0052] Step 5:

[0053] The server performs sentiment analysis using audio and video data. It analyzes voice tone and facial expressions to quantify the user's emotional state and reactions.

[0054] Step 6:

[0055] The server integrates the results of natural language processing and sentiment analysis to assess the risk of each statement constituting harassment. It then scores the risk and visualizes the results.

[0056] Step 7:

[0057] The server sends the analysis results to the user's device. The results include visualized feedback, which the user can use to identify areas for improvement in their conversation.

[0058] Step 8:

[0059] The user's device provides feedback to the user through visual display or audio output. Based on the feedback, the user can immediately modify their words and actions, facilitating smoother communication.

[0060] (Example 1)

[0061] 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."

[0062] In modern society, harassment in communication is a serious problem, and monitoring and prevention are particularly difficult in the digital realm. To prevent such problems, an effective monitoring system is necessary. Traditional methods struggle with real-time analysis and immediate feedback, and lack the means to accurately capture feelings of discomfort and risk. There is a need to provide solutions to these problems.

[0063] 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.

[0064] In this invention, the server includes means for recording audio and video digital data, means for pre-processing the recorded digital data and transmitting it to the server, and means for performing natural language processing and sentiment analysis to determine harassment risk. This makes it possible to evaluate the risk of harassment in digital communication in real time and provide immediate feedback.

[0065] "Digital audio and video data" refers to a collection of information recorded in digital format, such as audio and video. This data possesses characteristics that make it suitable for storage and processing on digital devices.

[0066] "Information processing equipment" refers to a general term for electronic devices equipped with the functions of recording, processing, and transmitting digital data, including audio and video. Generally, this includes computers and smartphones.

[0067] A "data processing device" refers to an integrated system of hardware and software that functions as a server, analyzing and evaluating received digital data.

[0068] "Information processing technology" refers to all technologies that use natural language processing and data analysis techniques to extract meaning and emotion from audio and video data.

[0069] "State analysis" refers to a method of evaluating a user's emotional state and the quality of their communication based on audio and video data.

[0070] "Relationship risks" refer to problems and incidents that may arise from harassment or misunderstandings in communication. These can be predicted through analysis and quantified as risks.

[0071] "Visual display" refers to a format in which numerical analysis results or warnings are output as text or graphics on the screen to convey them to the user.

[0072] A "notification" refers to a message that a system sends to a user to quickly convey important information or warnings.

[0073] This invention relates to a system that uses an advanced digital data analysis system to predict harassment risks in real-time communication and provides immediate feedback to users. The following specifically describes embodiments for carrying out this invention.

[0074] First, the user records digital audio and video data using an information processing device (e.g., a smartphone or computer). This uses the built-in microphone and camera. The user activates these functions during the conversation, continuously collecting data.

[0075] Next, the terminal preprocesses the recorded digital data and transmits it to a data processing unit (server) via the internet. This process applies noise reduction algorithms and data compression techniques. The data processing unit analyzes the received data using natural language processing techniques and sentiment analysis algorithms that utilize generative AI models. This extracts risk factors from the data and quantifies the user's emotional state.

[0076] For example, this system could be used when a user is speaking in a workplace meeting. The user's speech is acquired as digital data in real time and immediately analyzed on the server. Based on this analysis, the server evaluates whether the conversation is aggressive or causing discomfort to participants. As a result, an alert is sent to the information processing device if necessary.

[0077] An example of a prompt would be, "Please describe the specific steps taken to analyze audio and video data to determine harassment risk." This prompt provides guidance for understanding in detail how the system processes the data.

[0078] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0079] Step 1:

[0080] The user uses the device to acquire digital audio and video data.

[0081] Specifically, the device launches an application and begins recording and video recording. The input is the user's voice and video, and the output is audio and video data in digital format. This data is stored in the device's storage.

[0082] Step 2:

[0083] The device performs preprocessing on the audio and video data it acquires.

[0084] Specifically, the terminal applies a noise reduction algorithm to compress the data. The input is the unprocessed digital data obtained in step 1, and the output is the noise-reduced and compressed data. The preprocessed data is then ready to be sent to the server.

[0085] Step 3:

[0086] The terminal sends the pre-processed data to the server.

[0087] In terms of specific operations, the terminal uploads data to the server via an internet connection. The input is the data preprocessed in step 2, and the output is the data transferred to the server. The server receives this data and prepares for the next processing step.

[0088] Step 4:

[0089] The server uses natural language processing techniques to convert the received audio data into text.

[0090] In terms of specific operation, the server uses a generative AI model to convert audio data into text. The input is the audio data sent to the server in step 3, and the output is text data. This text forms the basis for the next analysis.

[0091] Step 5:

[0092] The server performs data analysis based on text data, using natural language processing techniques.

[0093] In practice, the server understands specific phrases and contexts and extracts key elements for risk assessment. The input is the text data from step 4, and the output is the risk assessment information resulting from the analysis.

[0094] Step 6:

[0095] The server performs emotion analysis based on video data and quantifies the user's emotional state.

[0096] Specifically, the server analyzes the video, evaluates the user's facial expressions and voice tone, and generates a numerical score. The input is the video data sent to the server in step 3, and the output is the emotion analysis score.

[0097] Step 7:

[0098] The server integrates the results of natural language processing and sentiment analysis to calculate an overall harassment risk score.

[0099] In practice, the server integrates this data and quantifies the risk level. The input is the analysis results obtained in steps 5 and 6, and the output is the overall risk score.

[0100] Step 8:

[0101] The server sends an alert to the user's device based on the risk score.

[0102] Specifically, the server sends alert data to the terminal in real time if it determines the risk to be high. The input is the risk score from step 7, and the output is feedback provided to the user as a warning display on the screen or an audio notification.

[0103] (Application Example 1)

[0104] 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."

[0105] In modern workplaces and educational environments, harassment is a major obstacle to interpersonal relationships, and its prevention is essential. However, conventional methods make it difficult to detect inappropriate behavior during communication in real time and provide immediate feedback. Therefore, a system is needed that can instantly detect risks during conversations and take appropriate action.

[0106] 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.

[0107] In this invention, the server includes means for acquiring audio and video data on the user's computing device, means for preprocessing the acquired information and transmitting it to a remote computer, means for performing unstructured data analysis and emotional analysis on the remote computer to determine the risk of inappropriate behavior in communication, means for visualizing the determination results and providing them to the user as feedback, and means for monitoring communication in a scene specified by the user, calculating emotional risk in real time, and providing feedback. This makes it possible to prevent harassment during communication and to realize smooth dialogue.

[0108] "Audio data" refers to information recorded in digital format from the user's voice and surrounding sounds.

[0109] "Video data" refers to information that visually captures the user's movements and environment.

[0110] "User's computing device" refers to an electronic device used to acquire and process audio and video data.

[0111] "Preprocessing" refers to the initial stage of processing data to convert it into a format suitable for analysis.

[0112] A "remote computer" is a network-connected electronic device that receives data transmitted from a user's computing device and performs analysis on it.

[0113] "Unstructured data analysis" is a technique for analyzing unstructured data formats such as text and audio using specific methods.

[0114] "Emotional analysis" is the process of determining a person's emotional state based on their tone of voice and facial expressions.

[0115] "Risk of inappropriate behavior" refers to an assessment of the likelihood of causing problems such as harassment in communication.

[0116] "Visualization" refers to a technique that displays analysis results in a format that is easy for users to understand.

[0117] "Feedback" refers to advice or instructions based on information that a system provides to the user.

[0118] The system of this invention consists of a user-operated computing device and a server. It primarily acquires and analyzes audio and video data to assess harassment risk. The user's computing device is a device such as a smartphone or smart glasses, which collects audio and video using a microphone and camera. A speech recognition engine (e.g., Google® Speech-to-Text API) is used to convert the audio data into text data. Furthermore, the OpenCV library is used for facial expression analysis.

[0119] The user's computing device can send preprocessed data to a remote server via a secure protocol (e.g., HTTPS). The server performs unstructured data analysis on the received data, which includes natural language processing libraries (e.g., NLTK, spaCy). For emotion analysis, machine learning techniques (e.g., TENSORFLOW®, PyTorch) are used to evaluate emotional states based on voice tone and facial expression data.

[0120] Based on the evaluation, the server calculates the risk of inappropriate behavior, visualizes the results, and provides feedback to the user. The user is presented with an emotional state score along with specific points to be mindful of during communication.

[0121] For example, a computer might monitor a conversation during a discussion at an educational institution. In this case, if the computer detects that a particular phrase might be offensive based on what is being said and the user's facial expressions, it will immediately provide feedback. This feedback serves as a guide for the user to react appropriately to the situation.

[0122] Examples of prompts for a generative AI model are as follows:

[0123] "Analyze the conversation content and assess the risk of harassment based on the following: phrases used, tone of voice, and participants' facial expressions. Issue an alert if specific factors are deemed high-risk."

[0124] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0125] Step 1:

[0126] The system acquires audio and video data from the user's device. The input is ambient sound and video information, and the output is digital audio and video files. In this step, the device's microphone and camera are used to collect data in real time.

[0127] Step 2:

[0128] The device preprocesses the acquired data and sends it to the server. It converts audio data to text (e.g., Google Speech-to-Text API) and extracts facial expression data from video (e.g., OpenCV). The input is raw audio and video data, and the output is preprocessed text and image data.

[0129] Step 3:

[0130] The server performs unstructured data analysis on text data to understand the meaning and context of the statements. The input is text data, and the output is the analyzed contextual information. At this stage, natural language processing techniques (e.g., NLTK, spaCy) are used to analyze the conversation.

[0131] Step 4:

[0132] The server performs emotion analysis, quantifying the user's emotional state based on their voice tone and facial expressions. The input is voice tone and facial expression data, and the output is an emotion score based on this data. Here, machine learning techniques (e.g., TensorFlow, PyTorch) are used for emotion evaluation.

[0133] Step 5:

[0134] The server calculates the risk of inappropriate behavior based on the analysis results, visualizes this risk, and provides feedback to the user. The input is contextual information and sentiment score, and the output is warning information provided to the user. This feedback is communicated to the user in real time by the terminal through display and audio.

[0135] Step 6:

[0136] Based on the feedback received, users adjust their communication methods and content on the spot. They revise specific actions based on the feedback, reducing the risks of the interaction. The input is feedback information, and the output is the result of the adjusted communication.

[0137] 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.

[0138] This invention is a system for determining the risk of harassment by utilizing user voice and video data. Furthermore, by combining it with an emotion engine, it achieves more accurate emotion recognition, enabling a detailed evaluation of the user's emotional state and the provision of appropriate feedback.

[0139] The user launches an application installed on their device to collect voice and facial expression data. The device records this data in real time, preprocesses it, and then sends it to a server. The server analyzes the data using natural language processing and emotion recognition technologies.

[0140] As an example, let's consider a workplace meeting. In this scenario, the user's statements and reactions are recorded in real time, and an emotion engine evaluates their current emotional state and stress level. Based on this analysis, the server determines how the user's statements are affecting others and generates a risk alert if inappropriate language is found.

[0141] Furthermore, the emotion engine accumulates past emotional history and learns how users react in specific situations. This allows the server to predict the user's emotional state and, if necessary, proactively suggest stress management and relaxation methods.

[0142] Users receive feedback sent from the server to their device, gaining clues to improve their communication. This feedback is provided as visual feedback through text and diagrams, or as audio guidance. In this way, users can achieve more flexible and adaptive communication in their interactions with others.

[0143] This invention can be used in a wide range of settings, including workplaces, educational institutions, and public spaces, and provides support for people to interact with each other in a safe and secure environment.

[0144] The following describes the processing flow.

[0145] Step 1:

[0146] The user launches a dedicated application on their device and begins capturing audio and video. The device's microphone and camera record the conversation and the user's facial expressions in real time.

[0147] Step 2:

[0148] The device preprocesses the recorded audio and video data. The audio data is de-noised and converted into text data using speech recognition. For the video data, facial recognition technology is used to extract features for sentiment analysis.

[0149] Step 3:

[0150] The device sends pre-processed speech-to-text data and sentiment features to the server. The transmission uses an encrypted communication protocol to ensure data security.

[0151] Step 4:

[0152] The server uses a natural language processing (NLP) model to analyze the received data, extracting the context and meaning of the conversation from the text data.

[0153] Step 5:

[0154] The server activates an emotion engine to evaluate the user's emotional state based on voice intonation and facial expression data. This evaluation also includes the user's stress and tension levels.

[0155] Step 6:

[0156] The server integrates NLP analysis and emotion engine results to score the harassment risk of a statement. If necessary, it generates alerts for statements that require improvement.

[0157] Step 7:

[0158] The server sends the generated risk assessment and feedback to the user's device. The feedback is presented as specific advice and visualized data.

[0159] Step 8:

[0160] The user's device receives feedback from the server and notifies the user visually and audibly. Based on the feedback received, the user can adjust their actions on the spot to improve communication.

[0161] (Example 2)

[0162] 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".

[0163] In recent years, harassment and inappropriate communication in interpersonal interactions in the workplace and educational settings have become a significant social problem. Such behavior negatively impacts individuals' mental and physical health and reduces overall organizational performance; therefore, systems are needed to prevent it. Furthermore, there is a need for technology that can provide appropriate feedback in diverse interaction settings and alert individuals to risky behaviors in real time.

[0164] 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.

[0165] In this invention, the server includes means for acquiring voice and image data with an information processing device, means for organizing the acquired data for data analysis and transmitting it to a remote processing device, and means for performing automatic processing and sentiment evaluation with the remote processing device to determine the risk of interpersonal behavior. This makes it possible to identify harassment risks in complex interpersonal interactions in real time and provide information users with guidelines for appropriate behavior.

[0166] "Audio and image data" refers to audio and visual information obtained from the user.

[0167] An "information processing device" is an electronic device used to acquire data and perform initial data organization.

[0168] "Data analysis" refers to the process of processing acquired data and transforming it into a useful form of information.

[0169] A "remote processing device" is a server or cloud system that receives acquired data and performs further advanced analysis on it.

[0170] "Automated processing" refers to computational processes that analyze and evaluate data without human intervention.

[0171] "Emotional assessment" is a process for determining a user's emotional state based on data.

[0172] "Interpersonal behavior risk" refers to the possibility that a user's behavior may be inappropriate or harmful to others.

[0173] "Information users" refer to users who receive the system's results and feedback.

[0174] "Visualization" refers to the process of displaying the results of data analysis in a visually easy-to-understand format.

[0175] "Evaluation value" refers to an indicator that quantifies a specific emotional state or risk.

[0176] A "warning" is the act of providing users with information that draws attention to a specific risk or issue.

[0177] "Real-time" refers to a process where data analysis and feedback are performed immediately on the spot.

[0178] This invention relates to a system that performs sentiment analysis and interpersonal behavior risk assessment using a user's voice and image data. This system consists of an information processing device, a remote processing device, and software modules that work in conjunction with them.

[0179] The user collects their own voice and facial expressions through an information processing device. This device consists of a typical mobile device or computer, and uses a camera and microphone as needed. This allows the user to acquire their own communication data. For example, it would record voice and facial expressions in real time during meetings or conversations.

[0180] The terminal processes the collected audio and image data, performing tasks such as noise reduction and facial recognition. Existing software is used for noise reduction, and libraries such as the OpenCV are used for facial recognition. The processed data is transmitted to the remote processing unit via encrypted communication.

[0181] The server acts as a remote processing unit, analyzing the received data. The server implements advanced automated processing technologies for natural language processing and sentiment evaluation. It performs speech recognition and evaluates emotional states based on the extracted text data. Specifically, it can utilize natural language processing APIs from common cloud services.

[0182] Based on the analysis results, the server determines the impact and risks of the user's statements on others and generates warnings as needed. Furthermore, it can learn from past data and predict the user's response in specific scenarios. This information is provided as real-time feedback to the device. Users can use the received feedback to try to improve their communication. The feedback is provided in text and visual formats, taking usability into consideration.

[0183] As a concrete example, in a workplace meeting, if a user speaks in an inappropriate tone, a warning will appear on their device immediately afterward. This warning includes information suggesting ways to improve their speech and explaining the potential future benefits.

[0184] Using a generative AI model, an example of a prompt message directed at a user might be, "Please suggest ways for the user to relax when they feel stressed in a specific situation."

[0185] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0186] Step 1:

[0187] The user launches an application on the information processing device. Audio and image data are acquired as input via the device's microphone and camera. This data is primary data recording the user's speech and facial expressions. Specifically, the information processing device begins recording and video, capturing the user's communication in real time.

[0188] Step 2:

[0189] The device performs noise reduction on the acquired audio data and applies a face recognition algorithm to the image data. This removes ambient noise from the audio and extracts only the user's facial expressions from the image. The output consists of pre-processed audio waveform data and clipped image data of the face region. The specific operation involves executing an audio processing algorithm and face recognition using OpenCV.

[0190] Step 3:

[0191] The terminal sends pre-processed audio and image data to the server. The data is encrypted using a secure protocol and transmitted via the server's API endpoint. The input is an encrypted data packet, and the output is a notification to the server indicating that data reception is complete. Specifically, the terminal establishes an HTTPS connection and sends the data to the server.

[0192] Step 4:

[0193] The server passes the received audio data to a natural language processing engine for conversion into text. Next, it passes the image data to an emotion evaluation engine to extract emotion information based on facial expressions. Inputs include audio waveforms and facial images, while outputs generate text data and emotion scores. Specific operations include calling speech recognition APIs and executing emotion analysis algorithms.

[0194] Step 5:

[0195] Based on the analysis results, the server evaluates the impact of a user's statements on others and generates risk alerts as needed. Furthermore, it predicts the user's emotional state by combining this with historical sentiment data. The inputs are text data and sentiment scores, and the outputs are a risk assessment report and predictive data. The specific operation involves applying statistical algorithms and predictive models.

[0196] Step 6:

[0197] The server generates and sends feedback to the user's terminal. The feedback is provided in a visual and audio instructional format and includes specific improvement suggestions. Inputs include risk assessment reports and predictive data, while output is the generation of feedback messages. Specific actions include creating the feedback format and sending the messages.

[0198] Step 7:

[0199] The user receives feedback displayed on their device and uses it to improve communication. The input is feedback messages from the server, and the output is the formation of a concrete action plan for future interactions. These actions include reviewing the feedback and formulating individual countermeasures.

[0200] (Application Example 2)

[0201] 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".

[0202] In brick-and-mortar stores, staff often struggle to accurately assess customers' emotions and concerns, resulting in a failure to prevent customer dissatisfaction and harassment. Traditional methods often rely on the individual experience and judgment of staff, lacking the ability to adjust customer service in response to changes in emotions. Therefore, technical solutions are needed to improve the quality of customer service in brick-and-mortar stores.

[0203] 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.

[0204] In this invention, the server includes means for acquiring audio and video data at a user terminal, means for initial processing the acquired information and transferring it to an information processing device, and means for performing contextual understanding and emotion evaluation on the information processing device and determining risk. This makes it possible to immediately analyze customer emotions and provide customer service support information that allows staff to respond appropriately.

[0205] "Audio data" refers to information about speech and related sounds, including user speech and other audio signals.

[0206] "Video data" refers to visual information acquired by cameras and other devices, including data such as the user's facial expressions, gestures, and other visual elements.

[0207] "User terminal" refers to a computer or smart device used to acquire and process data, specifically a device operated by the user.

[0208] An "information processing device" refers to a computer system that analyzes and processes data, and is a device that processes information using a specific algorithm.

[0209] "Contextual understanding" is the process of analyzing the meaning and intent of audio and text data to grasp linguistic and situational implications.

[0210] "Emotional assessment" is a technical process that identifies emotions from a user's voice and video data and evaluates that emotional state quantitatively or qualitatively.

[0211] "Assessing risk" is the process of evaluating potential risks and problems based on analyzed data, and determining their presence and degree.

[0212] "Customer service support information" refers to information provided to support smooth communication with customers, specifically real-time response strategies and guidance.

[0213] To implement this invention, users (staff) in a physical store wear a terminal such as smart glasses to acquire conversation data with customers. The terminal collects acoustic data using a microphone and records video data using a camera. After initial processing within the terminal, this data is transferred to an information processing device (server) in the cloud.

[0214] The server uses natural language processing software to understand the context of the data. It analyzes the linguistic features of audio data using Google Cloud's "Cloud Natural Language API." Furthermore, for video data, it estimates the emotional state from the user's facial expressions using an emotion assessment algorithm. Machine learning frameworks such as TensorFlow are utilized in this process.

[0215] Subsequently, based on the analyzed emotional and contextual information, the server determines the level of risk and sends the result as real-time feedback to the user's device. For example, if a customer expresses dissatisfaction with the service, customer service support information such as "Try to remain calm in this situation" will be displayed on the device's screen. This allows staff to respond instantly to the situation.

[0216] An example of a prompt for a generative AI model is, "Suggest how to handle a situation where a customer is angry." This is used by the server to provide appropriate advice to staff in real time.

[0217] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0218] Step 1:

[0219] The user wears a smart device and begins interacting with customers in the store. Input consists of audio data collected by the device's microphone and video data recorded by the camera. Output is this data in a format prepared for initial processing.

[0220] Step 2:

[0221] The terminal performs initial processing of the collected audio and video data. Specifically, it performs processes such as noise reduction and identifying areas of interest (faces) within the video using face detection algorithms. The output is pre-processed data that is converted into a format that can be sent to the server.

[0222] Step 3:

[0223] The server receives pre-processed data sent from the terminal. The input consists of pre-processed audio and video data. The server uses Google Cloud's "Cloud Natural Language API" to perform natural language processing on the audio data and analyze its linguistic features. For the video data, it performs sentiment evaluation using TensorFlow to estimate the user's emotional state. The output is a set of contextual information and sentiment information.

[0224] Step 4:

[0225] The server determines the risk level based on contextual and sentiment information. It uses a machine learning model to compare the results with historical data. The output is the determined risk level and the corresponding real-time countermeasures.

[0226] Step 5:

[0227] The server sends the generated risk assessment results and customer service support information to the user's terminal. The input consists of the assessment results and customer service support information. The output is customer service support information displayed on the terminal's screen, which specifically includes advice such as "If the customer expresses dissatisfaction, respond calmly."

[0228] Step 6:

[0229] Users optimize their customer service based on the support information provided. In practice, they are required to provide customer-friendly service while following the instructions on the terminal display.

[0230] 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.

[0231] 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.

[0232] 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.

[0233] [Second Embodiment]

[0234] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0235] 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.

[0236] 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).

[0237] 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.

[0238] 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.

[0239] 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).

[0240] 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.

[0241] 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.

[0242] 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.

[0243] 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.

[0244] 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.

[0245] 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".

[0246] This invention relates to a system that acquires audio and video data on a user's terminal, analyzes that data on a server, and determines the risk of harassment. This system assists user communication and plays a role in preventing harassment from occurring.

[0247] The user uses a device to collect audio and video data during a conversation. The device preprocesses this data and sends it to a server. The server uses natural language processing techniques and sentiment analysis algorithms to analyze the received data.

[0248] In natural language processing technology, the server analyzes the text data of the conversation to understand the meaning and context of the statements. This allows it to determine whether specific phrases or expressions constitute harassment. Furthermore, sentiment analysis algorithms are used to evaluate voice tone and facial expression data, quantifying the user's emotional state. Based on this data, the server scores the risk of harassment and visualizes it.

[0249] As a concrete example, consider a situation where a user is speaking in a workplace meeting. In this situation, the user's device records the conversation and sends it to the server in real time. The server quickly analyzes the statements made during the meeting and determines whether certain statements are offensive or whether other participants are showing discomfort with the statements. If the risk of harassment increases as a result, the server immediately sends an alert to the user's device to draw their attention to the issue.

[0250] The user's device provides this feedback visually or audibly, allowing the user to adjust the tone and content of the conversation in real time. This facilitates smoother communication and ensures that all participants can comfortably continue the conversation.

[0251] This system is designed for use in a variety of environments, including workplaces, educational institutions, and public organizations, and aims to contribute to creating an environment where people can communicate with peace of mind.

[0252] The following describes the processing flow.

[0253] Step 1:

[0254] The user uses their device to collect audio and video data during conversations. The device's microphone and camera capture the data in real time.

[0255] Step 2:

[0256] The device preprocesses the collected data. Noise reduction and filtering of unnecessary information are performed, and the audio data is converted into text data using speech recognition technology.

[0257] Step 3:

[0258] The terminal sends the pre-processed data to the server using a secure communication protocol. This data includes text data and audio / video features.

[0259] Step 4:

[0260] The server inputs the received text data into a natural language processing (NLP) model, which performs grammatical and semantic analysis. This helps the server understand the context in which the utterance is presented.

[0261] Step 5:

[0262] The server performs sentiment analysis using audio and video data. It analyzes voice tone and facial expressions to quantify the user's emotional state and reactions.

[0263] Step 6:

[0264] The server integrates the results of natural language processing and sentiment analysis to assess the risk of each statement constituting harassment. It then scores the risk and visualizes the results.

[0265] Step 7:

[0266] The server sends the analysis results to the user's device. The results include visualized feedback, which the user can use to identify areas for improvement in their conversation.

[0267] Step 8:

[0268] The user's device provides feedback to the user through visual display or audio output. Based on the feedback, the user can immediately modify their words and actions, facilitating smoother communication.

[0269] (Example 1)

[0270] 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."

[0271] In modern society, harassment in communication is a serious problem, and monitoring and prevention are particularly difficult in the digital realm. To prevent such problems, an effective monitoring system is necessary. Traditional methods struggle with real-time analysis and immediate feedback, and lack the means to accurately capture feelings of discomfort and risk. There is a need to provide solutions to these problems.

[0272] 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.

[0273] In this invention, the server includes means for recording audio and video digital data, means for pre-processing the recorded digital data and transmitting it to the server, and means for performing natural language processing and sentiment analysis to determine harassment risk. This makes it possible to evaluate the risk of harassment in digital communication in real time and provide immediate feedback.

[0274] "Digital audio and video data" refers to a collection of information recorded in digital format, such as audio and video. This data possesses characteristics that make it suitable for storage and processing on digital devices.

[0275] "Information processing equipment" refers to a general term for electronic devices equipped with the functions of recording, processing, and transmitting digital data, including audio and video. Generally, this includes computers and smartphones.

[0276] A "data processing device" refers to an integrated system of hardware and software that functions as a server, analyzing and evaluating received digital data.

[0277] "Information processing technology" refers to all technologies that use natural language processing and data analysis techniques to extract meaning and emotion from audio and video data.

[0278] "State analysis" refers to a method of evaluating a user's emotional state and the quality of their communication based on audio and video data.

[0279] "Relationship risks" refer to problems and incidents that may arise from harassment or misunderstandings in communication. These can be predicted through analysis and quantified as risks.

[0280] "Visual display" refers to a format in which numerical analysis results or warnings are output as text or graphics on the screen to convey them to the user.

[0281] A "notification" refers to a message that a system sends to a user to quickly convey important information or warnings.

[0282] This invention relates to a system that uses an advanced digital data analysis system to predict harassment risks in real-time communication and provides immediate feedback to users. The following specifically describes embodiments for carrying out this invention.

[0283] First, the user records digital audio and video data using an information processing device (e.g., a smartphone or computer). This uses the built-in microphone and camera. The user activates these functions during the conversation, continuously collecting data.

[0284] Next, the terminal preprocesses the recorded digital data and transmits it to a data processing device (server) via the Internet. In this process, noise removal algorithms and data compression techniques are applied. The data processing device analyzes the received data using natural language processing techniques and sentiment analysis algorithms that utilize a generated AI model. As a result, elements that are risk factors are extracted from the data, and the user's emotional state is quantified.

[0285] For example, it is conceivable to utilize this system when a user is speaking during a workplace meeting. The content of the user's speech is acquired as digital data in real time and immediately analyzed on the server. As a result of this analysis, the server evaluates whether the conversation is aggressive or is causing discomfort to the participants. Depending on the result, an alert is sent to the information processing device as necessary.

[0286] An example of a prompt sentence is "Please explain the specific steps for analyzing voice and video data to determine harassment risk." This prompt serves as a guide for understanding in detail how the system processes data.

[0287] The flow of the specific process in Example 1 will be described using FIG. 11.

[0288] Step 1:

[0289] The user uses the terminal to acquire digital data of voice and video.

[0290] As a specific operation, the application on the terminal is launched, and recording and video recording are started. The input is the user's voice and video, and the output is digital format voice and video data. This data is stored in the terminal's storage.

[0291] Step 2:

[0292] The terminal preprocesses the acquired voice and video data.

[0293] Specifically, the terminal applies a noise reduction algorithm to compress the data. The input is the unprocessed digital data obtained in step 1, and the output is the noise-reduced and compressed data. The preprocessed data is then ready to be sent to the server.

[0294] Step 3:

[0295] The terminal sends the pre-processed data to the server.

[0296] In terms of specific operations, the terminal uploads data to the server via an internet connection. The input is the data preprocessed in step 2, and the output is the data transferred to the server. The server receives this data and prepares for the next processing step.

[0297] Step 4:

[0298] The server uses natural language processing techniques to convert the received audio data into text.

[0299] In terms of specific operation, the server uses a generative AI model to convert audio data into text. The input is the audio data sent to the server in step 3, and the output is text data. This text forms the basis for the next analysis.

[0300] Step 5:

[0301] The server performs data analysis based on text data, using natural language processing techniques.

[0302] In practice, the server understands specific phrases and contexts and extracts key elements for risk assessment. The input is the text data from step 4, and the output is the risk assessment information resulting from the analysis.

[0303] Step 6:

[0304] The server performs sentiment analysis based on video data and quantifies the user's emotional state.

[0305] As a specific operation, the server analyzes the video, evaluates the user's facial expressions and voice tones, and generates a quantification score. The input is the video data sent to the server in step 3, and the output is the sentiment analysis score.

[0306] Step 7:

[0307] The server integrates the results of natural language processing and sentiment analysis and calculates a comprehensive harassment risk score.

[0308] As a specific operation, the server integrates these data and quantifies the risk level. The input is each analysis result obtained in step 5 and step 6, and the output is the comprehensive risk score.

[0309] Step 8:

[0310] The server sends an alert to the user's terminal based on the risk score.

[0311] As a specific operation, the server sends alert data to the terminal in real time when it is determined to be a high risk. The input is the risk score in step 7, and the output is the feedback provided to the user as a warning display on the screen or an audio notification.

[0312] (Application Example 1)

[0313] Next, Application Example 1 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart glasses 214 are referred to as the "terminal".

[0314] In modern workplaces and educational environments, harassment is a major obstacle to interpersonal relationships, and its prevention is essential. However, conventional methods make it difficult to detect inappropriate behavior during communication in real time and provide immediate feedback. Therefore, a system is needed that can instantly detect risks during conversations and take appropriate action.

[0315] 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.

[0316] In this invention, the server includes means for acquiring audio and video data on the user's computing device, means for preprocessing the acquired information and transmitting it to a remote computer, means for performing unstructured data analysis and emotional analysis on the remote computer to determine the risk of inappropriate behavior in communication, means for visualizing the determination results and providing them to the user as feedback, and means for monitoring communication in a scene specified by the user, calculating emotional risk in real time, and providing feedback. This makes it possible to prevent harassment during communication and to realize smooth dialogue.

[0317] "Audio data" refers to information recorded in digital format from the user's voice and surrounding sounds.

[0318] "Video data" refers to information that visually captures the user's movements and environment.

[0319] "User's computing device" refers to an electronic device used to acquire and process audio and video data.

[0320] "Preprocessing" refers to the initial stage of processing data to convert it into a format suitable for analysis.

[0321] A "remote computer" is a network-connected electronic device that receives data transmitted from a user's computing device and performs analysis on it.

[0322] "Unstructured data analysis" is a technique for analyzing unstructured data formats such as text and audio using specific methods.

[0323] "Emotional analysis" is the process of determining a person's emotional state based on their tone of voice and facial expressions.

[0324] "Risk of inappropriate behavior" refers to an assessment of the likelihood of causing problems such as harassment in communication.

[0325] "Visualization" refers to a technique that displays analysis results in a format that is easy for users to understand.

[0326] "Feedback" refers to advice or instructions based on information that a system provides to the user.

[0327] The system of this invention consists of a user-operated computing device and a server. It primarily acquires and analyzes audio and video data to assess harassment risk. The user's computing device is a device such as a smartphone or smart glasses, which collects audio and video using a microphone and camera. A speech recognition engine (e.g., Google Speech-to-Text API) is used to convert the audio data into text data. Furthermore, the OpenCV library is used for facial expression analysis.

[0328] The user's computing device can send preprocessed data to a remote server via a secure protocol (e.g., HTTPS). The server performs unstructured data analysis on the received data, including natural language processing libraries (e.g., NLTK, spaCy). For emotion analysis, machine learning techniques (e.g., TensorFlow, PyTorch) are used to evaluate emotional states based on voice tone and facial expression data.

[0329] Based on the evaluation, the server calculates the risk of inappropriate behavior, visualizes the results, and provides feedback to the user. The user is presented with an emotional state score along with specific points to be mindful of during communication.

[0330] For example, a computer might monitor a conversation during a discussion at an educational institution. In this case, if the computer detects that a particular phrase might be offensive based on what is being said and the user's facial expressions, it will immediately provide feedback. This feedback serves as a guide for the user to react appropriately to the situation.

[0331] Examples of prompts for a generative AI model are as follows:

[0332] "Analyze the conversation content and assess the risk of harassment based on the following: phrases used, tone of voice, and participants' facial expressions. Issue an alert if specific factors are deemed high-risk."

[0333] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0334] Step 1:

[0335] The system acquires audio and video data from the user's device. The input is ambient sound and video information, and the output is digital audio and video files. In this step, the device's microphone and camera are used to collect data in real time.

[0336] Step 2:

[0337] The device preprocesses the acquired data and sends it to the server. It converts audio data to text (e.g., Google Speech-to-Text API) and extracts facial expression data from video (e.g., OpenCV). The input is raw audio and video data, and the output is preprocessed text and image data.

[0338] Step 3:

[0339] The server performs unstructured data analysis on text data to understand the meaning and context of the statements. The input is text data, and the output is the analyzed contextual information. At this stage, natural language processing techniques (e.g., NLTK, spaCy) are used to analyze the conversation.

[0340] Step 4:

[0341] The server performs emotion analysis, quantifying the user's emotional state based on their voice tone and facial expressions. The input is voice tone and facial expression data, and the output is an emotion score based on this data. Here, machine learning techniques (e.g., TensorFlow, PyTorch) are used for emotion evaluation.

[0342] Step 5:

[0343] The server calculates the risk of inappropriate behavior based on the analysis results, visualizes this risk, and provides feedback to the user. The input is contextual information and sentiment score, and the output is warning information provided to the user. This feedback is communicated to the user in real time by the terminal through display and audio.

[0344] Step 6:

[0345] Based on the feedback received, users adjust their communication methods and content on the spot. They revise specific actions based on the feedback, reducing the risks of the interaction. The input is feedback information, and the output is the result of the adjusted communication.

[0346] 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.

[0347] This invention is a system for determining the risk of harassment by utilizing user voice and video data. Furthermore, by combining it with an emotion engine, it achieves more accurate emotion recognition, enabling a detailed evaluation of the user's emotional state and the provision of appropriate feedback.

[0348] The user launches an application installed on their device to collect voice and facial expression data. The device records this data in real time, preprocesses it, and then sends it to a server. The server analyzes the data using natural language processing and emotion recognition technologies.

[0349] As an example, let's consider a workplace meeting. In this scenario, the user's statements and reactions are recorded in real time, and an emotion engine evaluates their current emotional state and stress level. Based on this analysis, the server determines how the user's statements are affecting others and generates a risk alert if inappropriate language is found.

[0350] Furthermore, the emotion engine accumulates past emotional history and learns how users react in specific situations. This allows the server to predict the user's emotional state and, if necessary, proactively suggest stress management and relaxation methods.

[0351] Users receive feedback sent from the server to their device, gaining clues to improve their communication. This feedback is provided as visual feedback through text and diagrams, or as audio guidance. In this way, users can achieve more flexible and adaptive communication in their interactions with others.

[0352] This invention can be used in a wide range of settings, including workplaces, educational institutions, and public spaces, and provides support for people to interact with each other in a safe and secure environment.

[0353] The following describes the processing flow.

[0354] Step 1:

[0355] The user launches a dedicated application on their device and begins capturing audio and video. The device's microphone and camera record the conversation and the user's facial expressions in real time.

[0356] Step 2:

[0357] The device preprocesses the recorded audio and video data. The audio data is de-noised and converted into text data using speech recognition. For the video data, facial recognition technology is used to extract features for sentiment analysis.

[0358] Step 3:

[0359] The device sends pre-processed speech-to-text data and sentiment features to the server. The transmission uses an encrypted communication protocol to ensure data security.

[0360] Step 4:

[0361] The server uses a natural language processing (NLP) model to analyze the received data, extracting the context and meaning of the conversation from the text data.

[0362] Step 5:

[0363] The server activates an emotion engine to evaluate the user's emotional state based on voice intonation and facial expression data. This evaluation also includes the user's stress and tension levels.

[0364] Step 6:

[0365] The server integrates NLP analysis and emotion engine results to score the harassment risk of a statement. If necessary, it generates alerts for statements that require improvement.

[0366] Step 7:

[0367] The server sends the generated risk assessment and feedback to the user's device. The feedback is presented as specific advice and visualized data.

[0368] Step 8:

[0369] The user's device receives feedback from the server and notifies the user visually and audibly. Based on the feedback received, the user can adjust their actions on the spot to improve communication.

[0370] (Example 2)

[0371] 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".

[0372] In recent years, harassment and inappropriate communication in interpersonal interactions in the workplace and educational settings have become a significant social problem. Such behavior negatively impacts individuals' mental and physical health and reduces overall organizational performance; therefore, systems are needed to prevent it. Furthermore, there is a need for technology that can provide appropriate feedback in diverse interaction settings and alert individuals to risky behaviors in real time.

[0373] 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.

[0374] In this invention, the server includes means for acquiring voice and image data with an information processing device, means for organizing the acquired data for data analysis and transmitting it to a remote processing device, and means for performing automatic processing and sentiment evaluation with the remote processing device to determine the risk of interpersonal behavior. This makes it possible to identify harassment risks in complex interpersonal interactions in real time and provide information users with guidelines for appropriate behavior.

[0375] "Audio and image data" refers to audio and visual information obtained from the user.

[0376] An "information processing device" is an electronic device used to acquire data and perform initial data organization.

[0377] "Data analysis" refers to the process of processing acquired data and transforming it into a useful form of information.

[0378] A "remote processing device" is a server or cloud system that receives acquired data and performs further advanced analysis on it.

[0379] "Automated processing" refers to computational processes that analyze and evaluate data without human intervention.

[0380] "Emotional assessment" is a process for determining a user's emotional state based on data.

[0381] "Interpersonal behavior risk" refers to the possibility that a user's behavior may be inappropriate or harmful to others.

[0382] "Information users" refer to users who receive the system's results and feedback.

[0383] "Visualization" refers to the process of displaying the results of data analysis in a visually easy-to-understand format.

[0384] "Evaluation value" refers to an indicator that quantifies a specific emotional state or risk.

[0385] A "warning" is the act of providing users with information that draws attention to a specific risk or issue.

[0386] "Real-time" refers to a process where data analysis and feedback are performed immediately on the spot.

[0387] This invention relates to a system that performs sentiment analysis and interpersonal behavior risk assessment using a user's voice and image data. This system consists of an information processing device, a remote processing device, and software modules that work in conjunction with them.

[0388] The user collects their own voice and facial expressions through an information processing device. This device consists of a typical mobile device or computer, and uses a camera and microphone as needed. This allows the user to acquire their own communication data. For example, it would record voice and facial expressions in real time during meetings or conversations.

[0389] The terminal processes the collected audio and image data, performing tasks such as noise reduction and facial recognition. Existing software is used for noise reduction, and libraries such as the OpenCV are used for facial recognition. The processed data is transmitted to the remote processing unit via encrypted communication.

[0390] The server acts as a remote processing unit, analyzing the received data. The server implements advanced automated processing technologies for natural language processing and sentiment evaluation. It performs speech recognition and evaluates emotional states based on the extracted text data. Specifically, it can utilize natural language processing APIs from common cloud services.

[0391] Based on the analysis results, the server determines the impact and risks of the user's statements on others and generates warnings as needed. Furthermore, it can learn from past data and predict the user's response in specific scenarios. This information is provided as real-time feedback to the device. Users can use the received feedback to try to improve their communication. The feedback is provided in text and visual formats, taking usability into consideration.

[0392] As a concrete example, in a workplace meeting, if a user speaks in an inappropriate tone, a warning will appear on their device immediately afterward. This warning includes information suggesting ways to improve their speech and explaining the potential future benefits.

[0393] Using a generative AI model, an example of a prompt message directed at a user might be, "Please suggest ways for the user to relax when they feel stressed in a specific situation."

[0394] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0395] Step 1:

[0396] The user launches an application on the information processing device. Audio and image data are acquired as input via the device's microphone and camera. This data is primary data recording the user's speech and facial expressions. Specifically, the information processing device begins recording and video, capturing the user's communication in real time.

[0397] Step 2:

[0398] The device performs noise reduction on the acquired audio data and applies a face recognition algorithm to the image data. This removes ambient noise from the audio and extracts only the user's facial expressions from the image. The output consists of pre-processed audio waveform data and clipped image data of the face region. The specific operation involves executing an audio processing algorithm and face recognition using OpenCV.

[0399] Step 3:

[0400] The terminal sends pre-processed audio and image data to the server. The data is encrypted using a secure protocol and transmitted via the server's API endpoint. The input is an encrypted data packet, and the output is a notification to the server indicating that data reception is complete. Specifically, the terminal establishes an HTTPS connection and sends the data to the server.

[0401] Step 4:

[0402] The server passes the received audio data to a natural language processing engine for conversion into text. Next, it passes the image data to an emotion evaluation engine to extract emotion information based on facial expressions. Inputs include audio waveforms and facial images, while outputs generate text data and emotion scores. Specific operations include calling speech recognition APIs and executing emotion analysis algorithms.

[0403] Step 5:

[0404] Based on the analysis results, the server evaluates the impact of a user's statements on others and generates risk alerts as needed. Furthermore, it predicts the user's emotional state by combining this with historical sentiment data. The inputs are text data and sentiment scores, and the outputs are a risk assessment report and predictive data. The specific operation involves applying statistical algorithms and predictive models.

[0405] Step 6:

[0406] The server generates and sends feedback to the user's terminal. The feedback is provided in a visual and audio instructional format and includes specific improvement suggestions. Inputs include risk assessment reports and predictive data, while output is the generation of feedback messages. Specific actions include creating the feedback format and sending the messages.

[0407] Step 7:

[0408] The user receives feedback displayed on their device and uses it to improve communication. The input is feedback messages from the server, and the output is the formation of a concrete action plan for future interactions. These actions include reviewing the feedback and formulating individual countermeasures.

[0409] (Application Example 2)

[0410] 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."

[0411] In brick-and-mortar stores, staff often struggle to accurately assess customers' emotions and concerns, resulting in a failure to prevent customer dissatisfaction and harassment. Traditional methods often rely on the individual experience and judgment of staff, lacking the ability to adjust customer service in response to changes in emotions. Therefore, technical solutions are needed to improve the quality of customer service in brick-and-mortar stores.

[0412] 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.

[0413] In this invention, the server includes means for acquiring audio and video data at a user terminal, means for initial processing the acquired information and transferring it to an information processing device, and means for performing contextual understanding and emotion evaluation on the information processing device and determining risk. This makes it possible to immediately analyze customer emotions and provide customer service support information that allows staff to respond appropriately.

[0414] "Audio data" refers to information about speech and related sounds, including user speech and other audio signals.

[0415] "Video data" refers to visual information acquired by cameras and other devices, including data such as the user's facial expressions, gestures, and other visual elements.

[0416] "User terminal" refers to a computer or smart device used to acquire and process data, specifically a device operated by the user.

[0417] An "information processing device" refers to a computer system that analyzes and processes data, and is a device that processes information using a specific algorithm.

[0418] "Contextual understanding" is the process of analyzing the meaning and intent of audio and text data to grasp linguistic and situational implications.

[0419] "Emotional assessment" is a technical process that identifies emotions from a user's voice and video data and evaluates that emotional state quantitatively or qualitatively.

[0420] "Assessing risk" is the process of evaluating potential risks and problems based on analyzed data, and determining their presence and degree.

[0421] "Customer service support information" refers to information provided to support smooth communication with customers, specifically real-time response strategies and guidance.

[0422] To implement this invention, users (staff) in a physical store wear a terminal such as smart glasses to acquire conversation data with customers. The terminal collects acoustic data using a microphone and records video data using a camera. After initial processing within the terminal, this data is transferred to an information processing device (server) in the cloud.

[0423] The server uses natural language processing software to understand the context of the data. It analyzes the linguistic features of audio data using Google Cloud's "Cloud Natural Language API." Furthermore, for video data, it estimates the emotional state from the user's facial expressions using an emotion assessment algorithm. Machine learning frameworks such as TensorFlow are utilized in this process.

[0424] Subsequently, based on the analyzed emotional and contextual information, the server determines the level of risk and sends the result as real-time feedback to the user's device. For example, if a customer expresses dissatisfaction with the service, customer service support information such as "Try to remain calm in this situation" will be displayed on the device's screen. This allows staff to respond instantly to the situation.

[0425] An example of a prompt for a generative AI model is, "Suggest how to handle a situation where a customer is angry." This is used by the server to provide appropriate advice to staff in real time.

[0426] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0427] Step 1:

[0428] The user wears a smart device and begins interacting with customers in the store. Input consists of audio data collected by the device's microphone and video data recorded by the camera. Output is this data in a format prepared for initial processing.

[0429] Step 2:

[0430] The terminal performs initial processing of the collected audio and video data. Specifically, it performs processes such as noise reduction and identifying areas of interest (faces) within the video using face detection algorithms. The output is pre-processed data that is converted into a format that can be sent to the server.

[0431] Step 3:

[0432] The server receives pre-processed data sent from the terminal. The input consists of pre-processed audio and video data. The server uses Google Cloud's "Cloud Natural Language API" to perform natural language processing on the audio data and analyze its linguistic features. For the video data, it performs sentiment evaluation using TensorFlow to estimate the user's emotional state. The output is a set of contextual information and sentiment information.

[0433] Step 4:

[0434] The server determines the risk level based on contextual and sentiment information. It uses a machine learning model to compare the results with historical data. The output is the determined risk level and the corresponding real-time countermeasures.

[0435] Step 5:

[0436] The server sends the generated risk assessment results and customer service support information to the user's terminal. The input consists of the assessment results and customer service support information. The output is customer service support information displayed on the terminal's screen, which specifically includes advice such as "If the customer expresses dissatisfaction, respond calmly."

[0437] Step 6:

[0438] Users optimize their customer service based on the support information provided. In practice, they are required to provide customer-friendly service while following the instructions on the terminal display.

[0439] 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.

[0440] 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.

[0441] 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.

[0442] [Third Embodiment]

[0443] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0444] 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.

[0445] 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).

[0446] 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.

[0447] 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.

[0448] 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).

[0449] 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.

[0450] 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.

[0451] 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.

[0452] 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.

[0453] 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.

[0454] 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".

[0455] This invention relates to a system that acquires audio and video data on a user's terminal, analyzes that data on a server, and determines the risk of harassment. This system assists user communication and plays a role in preventing harassment from occurring.

[0456] The user uses a device to collect audio and video data during a conversation. The device preprocesses this data and sends it to a server. The server uses natural language processing techniques and sentiment analysis algorithms to analyze the received data.

[0457] In natural language processing technology, the server analyzes the text data of the conversation to understand the meaning and context of the statements. This allows it to determine whether specific phrases or expressions constitute harassment. Furthermore, sentiment analysis algorithms are used to evaluate voice tone and facial expression data, quantifying the user's emotional state. Based on this data, the server scores the risk of harassment and visualizes it.

[0458] As a concrete example, consider a situation where a user is speaking in a workplace meeting. In this situation, the user's device records the conversation and sends it to the server in real time. The server quickly analyzes the statements made during the meeting and determines whether certain statements are offensive or whether other participants are showing discomfort with the statements. If the risk of harassment increases as a result, the server immediately sends an alert to the user's device to draw their attention to the issue.

[0459] The user's device provides this feedback visually or audibly, allowing the user to adjust the tone and content of the conversation in real time. This facilitates smoother communication and ensures that all participants can comfortably continue the conversation.

[0460] This system is designed for use in a variety of environments, including workplaces, educational institutions, and public organizations, and aims to contribute to creating an environment where people can communicate with peace of mind.

[0461] The following describes the processing flow.

[0462] Step 1:

[0463] The user uses their device to collect audio and video data during conversations. The device's microphone and camera capture the data in real time.

[0464] Step 2:

[0465] The device preprocesses the collected data. Noise reduction and filtering of unnecessary information are performed, and the audio data is converted into text data using speech recognition technology.

[0466] Step 3:

[0467] The terminal sends the pre-processed data to the server using a secure communication protocol. This data includes text data and audio / video features.

[0468] Step 4:

[0469] The server inputs the received text data into a natural language processing (NLP) model, which performs grammatical and semantic analysis. This helps the server understand the context in which the utterance is presented.

[0470] Step 5:

[0471] The server performs sentiment analysis using audio and video data. It analyzes voice tone and facial expressions to quantify the user's emotional state and reactions.

[0472] Step 6:

[0473] The server integrates the results of natural language processing and sentiment analysis to assess the risk of each statement constituting harassment. It then scores the risk and visualizes the results.

[0474] Step 7:

[0475] The server sends the analysis results to the user's device. The results include visualized feedback, which the user can use to identify areas for improvement in their conversation.

[0476] Step 8:

[0477] The user's device provides feedback to the user through visual display or audio output. Based on the feedback, the user can immediately modify their words and actions, facilitating smoother communication.

[0478] (Example 1)

[0479] 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."

[0480] In modern society, harassment in communication is a serious problem, and monitoring and prevention are particularly difficult in the digital realm. To prevent such problems, an effective monitoring system is necessary. Traditional methods struggle with real-time analysis and immediate feedback, and lack the means to accurately capture feelings of discomfort and risk. There is a need to provide solutions to these problems.

[0481] 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.

[0482] In this invention, the server includes means for recording audio and video digital data, means for pre-processing the recorded digital data and transmitting it to the server, and means for performing natural language processing and sentiment analysis to determine harassment risk. This makes it possible to evaluate the risk of harassment in digital communication in real time and provide immediate feedback.

[0483] "Digital audio and video data" refers to a collection of information recorded in digital format, such as audio and video. This data possesses characteristics that make it suitable for storage and processing on digital devices.

[0484] "Information processing equipment" refers to a general term for electronic devices equipped with the functions of recording, processing, and transmitting digital data, including audio and video. Generally, this includes computers and smartphones.

[0485] A "data processing device" refers to an integrated system of hardware and software that functions as a server, analyzing and evaluating received digital data.

[0486] "Information processing technology" refers to all technologies that use natural language processing and data analysis techniques to extract meaning and emotion from audio and video data.

[0487] "State analysis" refers to a method of evaluating a user's emotional state and the quality of their communication based on audio and video data.

[0488] "Relationship risks" refer to problems and incidents that may arise from harassment or misunderstandings in communication. These can be predicted through analysis and quantified as risks.

[0489] "Visual display" refers to a format in which numerical analysis results or warnings are output as text or graphics on the screen to convey them to the user.

[0490] A "notification" refers to a message that a system sends to a user to quickly convey important information or warnings.

[0491] This invention relates to a system that uses an advanced digital data analysis system to predict harassment risks in real-time communication and provides immediate feedback to users. The following specifically describes embodiments for carrying out this invention.

[0492] First, the user records digital audio and video data using an information processing device (e.g., a smartphone or computer). This uses the built-in microphone and camera. The user activates these functions during the conversation, continuously collecting data.

[0493] Next, the terminal preprocesses the recorded digital data and transmits it to a data processing unit (server) via the internet. This process applies noise reduction algorithms and data compression techniques. The data processing unit analyzes the received data using natural language processing techniques and sentiment analysis algorithms that utilize generative AI models. This extracts risk factors from the data and quantifies the user's emotional state.

[0494] For example, this system could be used when a user is speaking in a workplace meeting. The user's speech is acquired as digital data in real time and immediately analyzed on the server. Based on this analysis, the server evaluates whether the conversation is aggressive or causing discomfort to participants. As a result, an alert is sent to the information processing device if necessary.

[0495] An example of a prompt would be, "Please describe the specific steps taken to analyze audio and video data to determine harassment risk." This prompt provides guidance for understanding in detail how the system processes the data.

[0496] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0497] Step 1:

[0498] The user uses the device to acquire digital audio and video data.

[0499] Specifically, the device launches an application and begins recording and video recording. The input is the user's voice and video, and the output is audio and video data in digital format. This data is stored in the device's storage.

[0500] Step 2:

[0501] The device performs preprocessing on the audio and video data it acquires.

[0502] Specifically, the terminal applies a noise reduction algorithm to compress the data. The input is the unprocessed digital data obtained in step 1, and the output is the noise-reduced and compressed data. The preprocessed data is then ready to be sent to the server.

[0503] Step 3:

[0504] The terminal sends the pre-processed data to the server.

[0505] In terms of specific operations, the terminal uploads data to the server via an internet connection. The input is the data preprocessed in step 2, and the output is the data transferred to the server. The server receives this data and prepares for the next processing step.

[0506] Step 4:

[0507] The server uses natural language processing techniques to convert the received audio data into text.

[0508] In terms of specific operation, the server uses a generative AI model to convert audio data into text. The input is the audio data sent to the server in step 3, and the output is text data. This text forms the basis for the next analysis.

[0509] Step 5:

[0510] The server performs data analysis based on text data, using natural language processing techniques.

[0511] In practice, the server understands specific phrases and contexts and extracts key elements for risk assessment. The input is the text data from step 4, and the output is the risk assessment information resulting from the analysis.

[0512] Step 6:

[0513] The server performs emotion analysis based on video data and quantifies the user's emotional state.

[0514] Specifically, the server analyzes the video, evaluates the user's facial expressions and voice tone, and generates a numerical score. The input is the video data sent to the server in step 3, and the output is the emotion analysis score.

[0515] Step 7:

[0516] The server integrates the results of natural language processing and sentiment analysis to calculate an overall harassment risk score.

[0517] In practice, the server integrates this data and quantifies the risk level. The input is the analysis results obtained in steps 5 and 6, and the output is the overall risk score.

[0518] Step 8:

[0519] The server sends an alert to the user's device based on the risk score.

[0520] Specifically, the server sends alert data to the terminal in real time if it determines the risk to be high. The input is the risk score from step 7, and the output is feedback provided to the user as a warning display on the screen or an audio notification.

[0521] (Application Example 1)

[0522] 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."

[0523] In modern workplaces and educational environments, harassment is a major obstacle to interpersonal relationships, and its prevention is essential. However, conventional methods make it difficult to detect inappropriate behavior during communication in real time and provide immediate feedback. Therefore, a system is needed that can instantly detect risks during conversations and take appropriate action.

[0524] 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.

[0525] In this invention, the server includes means for acquiring audio and video data on the user's computing device, means for preprocessing the acquired information and transmitting it to a remote computer, means for performing unstructured data analysis and emotional analysis on the remote computer to determine the risk of inappropriate behavior in communication, means for visualizing the determination results and providing them to the user as feedback, and means for monitoring communication in a scene specified by the user, calculating emotional risk in real time, and providing feedback. This makes it possible to prevent harassment during communication and to realize smooth dialogue.

[0526] "Audio data" refers to information recorded in digital format from the user's voice and surrounding sounds.

[0527] "Video data" refers to information that visually captures the user's movements and environment.

[0528] "User's computing device" refers to an electronic device used to acquire and process audio and video data.

[0529] "Preprocessing" refers to the initial stage of processing data to convert it into a format suitable for analysis.

[0530] A "remote computer" is a network-connected electronic device that receives data transmitted from a user's computing device and performs analysis on it.

[0531] "Unstructured data analysis" is a technique for analyzing unstructured data formats such as text and audio using specific methods.

[0532] "Emotional analysis" is the process of determining a person's emotional state based on their tone of voice and facial expressions.

[0533] "Risk of inappropriate behavior" refers to an assessment of the likelihood of causing problems such as harassment in communication.

[0534] "Visualization" refers to a technique that displays analysis results in a format that is easy for users to understand.

[0535] "Feedback" refers to advice or instructions based on information that a system provides to the user.

[0536] The system of this invention consists of a user-operated computing device and a server. It primarily acquires and analyzes audio and video data to assess harassment risk. The user's computing device is a device such as a smartphone or smart glasses, which collects audio and video using a microphone and camera. A speech recognition engine (e.g., Google Speech-to-Text API) is used to convert the audio data into text data. Furthermore, the OpenCV library is used for facial expression analysis.

[0537] The user's computing device can send preprocessed data to a remote server via a secure protocol (e.g., HTTPS). The server performs unstructured data analysis on the received data, including natural language processing libraries (e.g., NLTK, spaCy). For emotion analysis, machine learning techniques (e.g., TensorFlow, PyTorch) are used to evaluate emotional states based on voice tone and facial expression data.

[0538] Based on the evaluation, the server calculates the risk of inappropriate behavior, visualizes the results, and provides feedback to the user. The user is presented with an emotional state score along with specific points to be mindful of during communication.

[0539] For example, a computer might monitor a conversation during a discussion at an educational institution. In this case, if the computer detects that a particular phrase might be offensive based on what is being said and the user's facial expressions, it will immediately provide feedback. This feedback serves as a guide for the user to react appropriately to the situation.

[0540] Examples of prompts for a generative AI model are as follows:

[0541] "Analyze the conversation content and assess the risk of harassment based on the following: phrases used, tone of voice, and participants' facial expressions. Issue an alert if specific factors are deemed high-risk."

[0542] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0543] Step 1:

[0544] The system acquires audio and video data from the user's device. The input is ambient sound and video information, and the output is digital audio and video files. In this step, the device's microphone and camera are used to collect data in real time.

[0545] Step 2:

[0546] The device preprocesses the acquired data and sends it to the server. It converts audio data to text (e.g., Google Speech-to-Text API) and extracts facial expression data from video (e.g., OpenCV). The input is raw audio and video data, and the output is preprocessed text and image data.

[0547] Step 3:

[0548] The server performs unstructured data analysis on text data to understand the meaning and context of the statements. The input is text data, and the output is the analyzed contextual information. At this stage, natural language processing techniques (e.g., NLTK, spaCy) are used to analyze the conversation.

[0549] Step 4:

[0550] The server performs emotion analysis, quantifying the user's emotional state based on their voice tone and facial expressions. The input is voice tone and facial expression data, and the output is an emotion score based on this data. Here, machine learning techniques (e.g., TensorFlow, PyTorch) are used for emotion evaluation.

[0551] Step 5:

[0552] The server calculates the risk of inappropriate behavior based on the analysis results, visualizes this risk, and provides feedback to the user. The input is contextual information and sentiment score, and the output is warning information provided to the user. This feedback is communicated to the user in real time by the terminal through display and audio.

[0553] Step 6:

[0554] Based on the feedback received, users adjust their communication methods and content on the spot. They revise specific actions based on the feedback, reducing the risks of the interaction. The input is feedback information, and the output is the result of the adjusted communication.

[0555] 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.

[0556] This invention is a system for determining the risk of harassment by utilizing user voice and video data. Furthermore, by combining it with an emotion engine, it achieves more accurate emotion recognition, enabling a detailed evaluation of the user's emotional state and the provision of appropriate feedback.

[0557] The user launches an application installed on their device to collect voice and facial expression data. The device records this data in real time, preprocesses it, and then sends it to a server. The server analyzes the data using natural language processing and emotion recognition technologies.

[0558] As an example, let's consider a workplace meeting. In this scenario, the user's statements and reactions are recorded in real time, and an emotion engine evaluates their current emotional state and stress level. Based on this analysis, the server determines how the user's statements are affecting others and generates a risk alert if inappropriate language is found.

[0559] Furthermore, the emotion engine accumulates past emotional history and learns how users react in specific situations. This allows the server to predict the user's emotional state and, if necessary, proactively suggest stress management and relaxation methods.

[0560] Users receive feedback sent from the server to their device, gaining clues to improve their communication. This feedback is provided as visual feedback through text and diagrams, or as audio guidance. In this way, users can achieve more flexible and adaptive communication in their interactions with others.

[0561] This invention can be used in a wide range of settings, including workplaces, educational institutions, and public spaces, and provides support for people to interact with each other in a safe and secure environment.

[0562] The following describes the processing flow.

[0563] Step 1:

[0564] The user launches a dedicated application on their device and begins capturing audio and video. The device's microphone and camera record the conversation and the user's facial expressions in real time.

[0565] Step 2:

[0566] The device preprocesses the recorded audio and video data. The audio data is de-noised and converted into text data using speech recognition. For the video data, facial recognition technology is used to extract features for sentiment analysis.

[0567] Step 3:

[0568] The device sends pre-processed speech-to-text data and sentiment features to the server. The transmission uses an encrypted communication protocol to ensure data security.

[0569] Step 4:

[0570] The server uses a natural language processing (NLP) model to analyze the received data, extracting the context and meaning of the conversation from the text data.

[0571] Step 5:

[0572] The server activates an emotion engine to evaluate the user's emotional state based on voice intonation and facial expression data. This evaluation also includes the user's stress and tension levels.

[0573] Step 6:

[0574] The server integrates NLP analysis and emotion engine results to score the harassment risk of a statement. If necessary, it generates alerts for statements that require improvement.

[0575] Step 7:

[0576] The server sends the generated risk assessment and feedback to the user's device. The feedback is presented as specific advice and visualized data.

[0577] Step 8:

[0578] The user's device receives feedback from the server and notifies the user visually and audibly. Based on the feedback received, the user can adjust their actions on the spot to improve communication.

[0579] (Example 2)

[0580] 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."

[0581] In recent years, harassment and inappropriate communication in interpersonal interactions in the workplace and educational settings have become a significant social problem. Such behavior negatively impacts individuals' mental and physical health and reduces overall organizational performance; therefore, systems are needed to prevent it. Furthermore, there is a need for technology that can provide appropriate feedback in diverse interaction settings and alert individuals to risky behaviors in real time.

[0582] 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.

[0583] In this invention, the server includes means for acquiring voice and image data with an information processing device, means for organizing the acquired data for data analysis and transmitting it to a remote processing device, and means for performing automatic processing and sentiment evaluation with the remote processing device to determine the risk of interpersonal behavior. This makes it possible to identify harassment risks in complex interpersonal interactions in real time and provide information users with guidelines for appropriate behavior.

[0584] "Audio and image data" refers to audio and visual information obtained from the user.

[0585] An "information processing device" is an electronic device used to acquire data and perform initial data organization.

[0586] "Data analysis" refers to the process of processing acquired data and transforming it into a useful form of information.

[0587] A "remote processing device" is a server or cloud system that receives acquired data and performs further advanced analysis on it.

[0588] "Automated processing" refers to computational processes that analyze and evaluate data without human intervention.

[0589] "Emotional assessment" is a process for determining a user's emotional state based on data.

[0590] "Interpersonal behavior risk" refers to the possibility that a user's behavior may be inappropriate or harmful to others.

[0591] "Information users" refer to users who receive the system's results and feedback.

[0592] "Visualization" refers to the process of displaying the results of data analysis in a visually easy-to-understand format.

[0593] "Evaluation value" refers to an indicator that quantifies a specific emotional state or risk.

[0594] A "warning" is the act of providing users with information that draws attention to a specific risk or issue.

[0595] "Real-time" refers to a process where data analysis and feedback are performed immediately on the spot.

[0596] This invention relates to a system that performs sentiment analysis and interpersonal behavior risk assessment using a user's voice and image data. This system consists of an information processing device, a remote processing device, and software modules that work in conjunction with them.

[0597] The user collects their own voice and facial expressions through an information processing device. This device consists of a typical mobile device or computer, and uses a camera and microphone as needed. This allows the user to acquire their own communication data. For example, it would record voice and facial expressions in real time during meetings or conversations.

[0598] The terminal processes the collected audio and image data, performing tasks such as noise reduction and facial recognition. Existing software is used for noise reduction, and libraries such as the OpenCV are used for facial recognition. The processed data is transmitted to the remote processing unit via encrypted communication.

[0599] The server acts as a remote processing unit, analyzing the received data. The server implements advanced automated processing technologies for natural language processing and sentiment evaluation. It performs speech recognition and evaluates emotional states based on the extracted text data. Specifically, it can utilize natural language processing APIs from common cloud services.

[0600] Based on the analysis results, the server determines the impact and risks of the user's statements on others and generates warnings as needed. Furthermore, it can learn from past data and predict the user's response in specific scenarios. This information is provided as real-time feedback to the device. Users can use the received feedback to try to improve their communication. The feedback is provided in text and visual formats, taking usability into consideration.

[0601] As a concrete example, in a workplace meeting, if a user speaks in an inappropriate tone, a warning will appear on their device immediately afterward. This warning includes information suggesting ways to improve their speech and explaining the potential future benefits.

[0602] Using a generative AI model, an example of a prompt message directed at a user might be, "Please suggest ways for the user to relax when they feel stressed in a specific situation."

[0603] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0604] Step 1:

[0605] The user launches an application on the information processing device. Audio and image data are acquired as input via the device's microphone and camera. This data is primary data recording the user's speech and facial expressions. Specifically, the information processing device begins recording and video, capturing the user's communication in real time.

[0606] Step 2:

[0607] The device performs noise reduction on the acquired audio data and applies a face recognition algorithm to the image data. This removes ambient noise from the audio and extracts only the user's facial expressions from the image. The output consists of pre-processed audio waveform data and clipped image data of the face region. The specific operation involves executing an audio processing algorithm and face recognition using OpenCV.

[0608] Step 3:

[0609] The terminal sends pre-processed audio and image data to the server. The data is encrypted using a secure protocol and transmitted via the server's API endpoint. The input is an encrypted data packet, and the output is a notification to the server indicating that data reception is complete. Specifically, the terminal establishes an HTTPS connection and sends the data to the server.

[0610] Step 4:

[0611] The server passes the received audio data to a natural language processing engine for conversion into text. Next, it passes the image data to an emotion evaluation engine to extract emotion information based on facial expressions. Inputs include audio waveforms and facial images, while outputs generate text data and emotion scores. Specific operations include calling speech recognition APIs and executing emotion analysis algorithms.

[0612] Step 5:

[0613] Based on the analysis results, the server evaluates the impact of a user's statements on others and generates risk alerts as needed. Furthermore, it predicts the user's emotional state by combining this with historical sentiment data. The inputs are text data and sentiment scores, and the outputs are a risk assessment report and predictive data. The specific operation involves applying statistical algorithms and predictive models.

[0614] Step 6:

[0615] The server generates and sends feedback to the user's terminal. The feedback is provided in a visual and audio instructional format and includes specific improvement suggestions. Inputs include risk assessment reports and predictive data, while output is the generation of feedback messages. Specific actions include creating the feedback format and sending the messages.

[0616] Step 7:

[0617] The user receives feedback displayed on their device and uses it to improve communication. The input is feedback messages from the server, and the output is the formation of a concrete action plan for future interactions. These actions include reviewing the feedback and formulating individual countermeasures.

[0618] (Application Example 2)

[0619] 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."

[0620] In brick-and-mortar stores, staff often struggle to accurately assess customers' emotions and concerns, resulting in a failure to prevent customer dissatisfaction and harassment. Traditional methods often rely on the individual experience and judgment of staff, lacking the ability to adjust customer service in response to changes in emotions. Therefore, technical solutions are needed to improve the quality of customer service in brick-and-mortar stores.

[0621] 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.

[0622] In this invention, the server includes means for acquiring audio and video data at a user terminal, means for initial processing the acquired information and transferring it to an information processing device, and means for performing contextual understanding and emotion evaluation on the information processing device and determining risk. This makes it possible to immediately analyze customer emotions and provide customer service support information that allows staff to respond appropriately.

[0623] "Audio data" refers to information about speech and related sounds, including user speech and other audio signals.

[0624] "Video data" refers to visual information acquired by cameras and other devices, including data such as the user's facial expressions, gestures, and other visual elements.

[0625] "User terminal" refers to a computer or smart device used to acquire and process data, specifically a device operated by the user.

[0626] An "information processing device" refers to a computer system that analyzes and processes data, and is a device that processes information using a specific algorithm.

[0627] "Contextual understanding" is the process of analyzing the meaning and intent of audio and text data to grasp linguistic and situational implications.

[0628] "Emotional assessment" is a technical process that identifies emotions from a user's voice and video data and evaluates that emotional state quantitatively or qualitatively.

[0629] "Assessing risk" is the process of evaluating potential risks and problems based on analyzed data, and determining their presence and degree.

[0630] "Customer service support information" refers to information provided to support smooth communication with customers, specifically real-time response strategies and guidance.

[0631] To implement this invention, users (staff) in a physical store wear a terminal such as smart glasses to acquire conversation data with customers. The terminal collects acoustic data using a microphone and records video data using a camera. After initial processing within the terminal, this data is transferred to an information processing device (server) in the cloud.

[0632] The server uses natural language processing software to understand the context of the data. It analyzes the linguistic features of audio data using Google Cloud's "Cloud Natural Language API." Furthermore, for video data, it estimates the emotional state from the user's facial expressions using an emotion assessment algorithm. Machine learning frameworks such as TensorFlow are utilized in this process.

[0633] Subsequently, based on the analyzed emotional and contextual information, the server determines the level of risk and sends the result as real-time feedback to the user's device. For example, if a customer expresses dissatisfaction with the service, customer service support information such as "Try to remain calm in this situation" will be displayed on the device's screen. This allows staff to respond instantly to the situation.

[0634] An example of a prompt for a generative AI model is, "Suggest how to handle a situation where a customer is angry." This is used by the server to provide appropriate advice to staff in real time.

[0635] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0636] Step 1:

[0637] The user wears a smart device and begins interacting with customers in the store. Input consists of audio data collected by the device's microphone and video data recorded by the camera. Output is this data in a format prepared for initial processing.

[0638] Step 2:

[0639] The terminal performs initial processing of the collected audio and video data. Specifically, it performs processes such as noise reduction and identifying areas of interest (faces) within the video using face detection algorithms. The output is pre-processed data that is converted into a format that can be sent to the server.

[0640] Step 3:

[0641] The server receives pre-processed data sent from the terminal. The input consists of pre-processed audio and video data. The server uses Google Cloud's "Cloud Natural Language API" to perform natural language processing on the audio data and analyze its linguistic features. For the video data, it performs sentiment evaluation using TensorFlow to estimate the user's emotional state. The output is a set of contextual information and sentiment information.

[0642] Step 4:

[0643] The server determines the risk level based on contextual and sentiment information. It uses a machine learning model to compare the results with historical data. The output is the determined risk level and the corresponding real-time countermeasures.

[0644] Step 5:

[0645] The server sends the generated risk assessment results and customer service support information to the user's terminal. The input consists of the assessment results and customer service support information. The output is customer service support information displayed on the terminal's screen, which specifically includes advice such as "If the customer expresses dissatisfaction, respond calmly."

[0646] Step 6:

[0647] Users optimize their customer service based on the support information provided. In practice, they are required to provide customer-friendly service while following the instructions on the terminal display.

[0648] 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.

[0649] 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.

[0650] 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.

[0651] [Fourth Embodiment]

[0652] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0653] 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.

[0654] 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).

[0655] 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.

[0656] 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.

[0657] 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).

[0658] 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.

[0659] 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.

[0660] 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.

[0661] 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.

[0662] 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.

[0663] 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.

[0664] 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".

[0665] This invention relates to a system that acquires audio and video data on a user's terminal, analyzes that data on a server, and determines the risk of harassment. This system assists user communication and plays a role in preventing harassment from occurring.

[0666] The user uses a device to collect audio and video data during a conversation. The device preprocesses this data and sends it to a server. The server uses natural language processing techniques and sentiment analysis algorithms to analyze the received data.

[0667] In natural language processing technology, the server analyzes the text data of the conversation to understand the meaning and context of the statements. This allows it to determine whether specific phrases or expressions constitute harassment. Furthermore, sentiment analysis algorithms are used to evaluate voice tone and facial expression data, quantifying the user's emotional state. Based on this data, the server scores the risk of harassment and visualizes it.

[0668] As a concrete example, consider a situation where a user is speaking in a workplace meeting. In this situation, the user's device records the conversation and sends it to the server in real time. The server quickly analyzes the statements made during the meeting and determines whether certain statements are offensive or whether other participants are showing discomfort with the statements. If the risk of harassment increases as a result, the server immediately sends an alert to the user's device to draw their attention to the issue.

[0669] The user's device provides this feedback visually or audibly, allowing the user to adjust the tone and content of the conversation in real time. This facilitates smoother communication and ensures that all participants can comfortably continue the conversation.

[0670] This system is designed for use in a variety of environments, including workplaces, educational institutions, and public organizations, and aims to contribute to creating an environment where people can communicate with peace of mind.

[0671] The following describes the processing flow.

[0672] Step 1:

[0673] The user uses their device to collect audio and video data during conversations. The device's microphone and camera capture the data in real time.

[0674] Step 2:

[0675] The device preprocesses the collected data. Noise reduction and filtering of unnecessary information are performed, and the audio data is converted into text data using speech recognition technology.

[0676] Step 3:

[0677] The terminal sends the pre-processed data to the server using a secure communication protocol. This data includes text data and audio / video features.

[0678] Step 4:

[0679] The server inputs the received text data into a natural language processing (NLP) model, which performs grammatical and semantic analysis. This helps the server understand the context in which the utterance is presented.

[0680] Step 5:

[0681] The server performs sentiment analysis using audio and video data. It analyzes voice tone and facial expressions to quantify the user's emotional state and reactions.

[0682] Step 6:

[0683] The server integrates the results of natural language processing and sentiment analysis to assess the risk of each statement constituting harassment. It then scores the risk and visualizes the results.

[0684] Step 7:

[0685] The server sends the analysis results to the user's device. The results include visualized feedback, which the user can use to identify areas for improvement in their conversation.

[0686] Step 8:

[0687] The user's device provides feedback to the user through visual display or audio output. Based on the feedback, the user can immediately modify their words and actions, facilitating smoother communication.

[0688] (Example 1)

[0689] 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".

[0690] In modern society, harassment in communication is a serious problem, and monitoring and prevention are particularly difficult in the digital realm. To prevent such problems, an effective monitoring system is necessary. Traditional methods struggle with real-time analysis and immediate feedback, and lack the means to accurately capture feelings of discomfort and risk. There is a need to provide solutions to these problems.

[0691] 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.

[0692] In this invention, the server includes means for recording audio and video digital data, means for pre-processing the recorded digital data and transmitting it to the server, and means for performing natural language processing and sentiment analysis to determine harassment risk. This makes it possible to evaluate the risk of harassment in digital communication in real time and provide immediate feedback.

[0693] "Digital audio and video data" refers to a collection of information recorded in digital format, such as audio and video. This data possesses characteristics that make it suitable for storage and processing on digital devices.

[0694] "Information processing equipment" refers to a general term for electronic devices equipped with the functions of recording, processing, and transmitting digital data, including audio and video. Generally, this includes computers and smartphones.

[0695] A "data processing device" refers to an integrated system of hardware and software that functions as a server, analyzing and evaluating received digital data.

[0696] "Information processing technology" refers to all technologies that use natural language processing and data analysis techniques to extract meaning and emotion from audio and video data.

[0697] "State analysis" refers to a method of evaluating a user's emotional state and the quality of their communication based on audio and video data.

[0698] "Relationship risks" refer to problems and incidents that may arise from harassment or misunderstandings in communication. These can be predicted through analysis and quantified as risks.

[0699] "Visual display" refers to a format in which numerical analysis results or warnings are output as text or graphics on the screen to convey them to the user.

[0700] A "notification" refers to a message that a system sends to a user to quickly convey important information or warnings.

[0701] This invention relates to a system that uses an advanced digital data analysis system to predict harassment risks in real-time communication and provides immediate feedback to users. The following specifically describes embodiments for carrying out this invention.

[0702] First, the user records digital audio and video data using an information processing device (e.g., a smartphone or computer). This uses the built-in microphone and camera. The user activates these functions during the conversation, continuously collecting data.

[0703] Next, the terminal preprocesses the recorded digital data and transmits it to a data processing unit (server) via the internet. This process applies noise reduction algorithms and data compression techniques. The data processing unit analyzes the received data using natural language processing techniques and sentiment analysis algorithms that utilize generative AI models. This extracts risk factors from the data and quantifies the user's emotional state.

[0704] For example, this system could be used when a user is speaking in a workplace meeting. The user's speech is acquired as digital data in real time and immediately analyzed on the server. Based on this analysis, the server evaluates whether the conversation is aggressive or causing discomfort to participants. As a result, an alert is sent to the information processing device if necessary.

[0705] An example of a prompt would be, "Please describe the specific steps taken to analyze audio and video data to determine harassment risk." This prompt provides guidance for understanding in detail how the system processes the data.

[0706] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0707] Step 1:

[0708] The user uses the device to acquire digital audio and video data.

[0709] Specifically, the device launches an application and begins recording and video recording. The input is the user's voice and video, and the output is audio and video data in digital format. This data is stored in the device's storage.

[0710] Step 2:

[0711] The device performs preprocessing on the audio and video data it acquires.

[0712] Specifically, the terminal applies a noise reduction algorithm to compress the data. The input is the unprocessed digital data obtained in step 1, and the output is the noise-reduced and compressed data. The preprocessed data is then ready to be sent to the server.

[0713] Step 3:

[0714] The terminal sends the pre-processed data to the server.

[0715] In terms of specific operations, the terminal uploads data to the server via an internet connection. The input is the data preprocessed in step 2, and the output is the data transferred to the server. The server receives this data and prepares for the next processing step.

[0716] Step 4:

[0717] The server uses natural language processing techniques to convert the received audio data into text.

[0718] In terms of specific operation, the server uses a generative AI model to convert audio data into text. The input is the audio data sent to the server in step 3, and the output is text data. This text forms the basis for the next analysis.

[0719] Step 5:

[0720] The server performs data analysis based on text data, using natural language processing techniques.

[0721] In practice, the server understands specific phrases and contexts and extracts key elements for risk assessment. The input is the text data from step 4, and the output is the risk assessment information resulting from the analysis.

[0722] Step 6:

[0723] The server performs emotion analysis based on video data and quantifies the user's emotional state.

[0724] Specifically, the server analyzes the video, evaluates the user's facial expressions and voice tone, and generates a numerical score. The input is the video data sent to the server in step 3, and the output is the emotion analysis score.

[0725] Step 7:

[0726] The server integrates the results of natural language processing and sentiment analysis to calculate an overall harassment risk score.

[0727] In practice, the server integrates this data and quantifies the risk level. The input is the analysis results obtained in steps 5 and 6, and the output is the overall risk score.

[0728] Step 8:

[0729] The server sends an alert to the user's device based on the risk score.

[0730] Specifically, the server sends alert data to the terminal in real time if it determines the risk to be high. The input is the risk score from step 7, and the output is feedback provided to the user as a warning display on the screen or an audio notification.

[0731] (Application Example 1)

[0732] 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".

[0733] In modern workplaces and educational environments, harassment is a major obstacle to interpersonal relationships, and its prevention is essential. However, conventional methods make it difficult to detect inappropriate behavior during communication in real time and provide immediate feedback. Therefore, a system is needed that can instantly detect risks during conversations and take appropriate action.

[0734] 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.

[0735] In this invention, the server includes means for acquiring audio and video data on the user's computing device, means for preprocessing the acquired information and transmitting it to a remote computer, means for performing unstructured data analysis and emotional analysis on the remote computer to determine the risk of inappropriate behavior in communication, means for visualizing the determination results and providing them to the user as feedback, and means for monitoring communication in a scene specified by the user, calculating emotional risk in real time, and providing feedback. This makes it possible to prevent harassment during communication and to realize smooth dialogue.

[0736] "Audio data" refers to information recorded in digital format from the user's voice and surrounding sounds.

[0737] "Video data" refers to information that visually captures the user's movements and environment.

[0738] "User's computing device" refers to an electronic device used to acquire and process audio and video data.

[0739] "Preprocessing" refers to the initial stage of processing data to convert it into a format suitable for analysis.

[0740] A "remote computer" is a network-connected electronic device that receives data transmitted from a user's computing device and performs analysis on it.

[0741] "Unstructured data analysis" is a technique for analyzing unstructured data formats such as text and audio using specific methods.

[0742] "Emotional analysis" is the process of determining a person's emotional state based on their tone of voice and facial expressions.

[0743] "Risk of inappropriate behavior" refers to an assessment of the likelihood of causing problems such as harassment in communication.

[0744] "Visualization" refers to a technique that displays analysis results in a format that is easy for users to understand.

[0745] "Feedback" refers to advice or instructions based on information that a system provides to the user.

[0746] The system of this invention consists of a user-operated computing device and a server. It primarily acquires and analyzes audio and video data to assess harassment risk. The user's computing device is a device such as a smartphone or smart glasses, which collects audio and video using a microphone and camera. A speech recognition engine (e.g., Google Speech-to-Text API) is used to convert the audio data into text data. Furthermore, the OpenCV library is used for facial expression analysis.

[0747] The user's computing device can send preprocessed data to a remote server via a secure protocol (e.g., HTTPS). The server performs unstructured data analysis on the received data, including natural language processing libraries (e.g., NLTK, spaCy). For emotion analysis, machine learning techniques (e.g., TensorFlow, PyTorch) are used to evaluate emotional states based on voice tone and facial expression data.

[0748] Based on the evaluation, the server calculates the risk of inappropriate behavior, visualizes the results, and provides feedback to the user. The user is presented with an emotional state score along with specific points to be mindful of during communication.

[0749] For example, a computer might monitor a conversation during a discussion at an educational institution. In this case, if the computer detects that a particular phrase might be offensive based on what is being said and the user's facial expressions, it will immediately provide feedback. This feedback serves as a guide for the user to react appropriately to the situation.

[0750] Examples of prompts for a generative AI model are as follows:

[0751] "Analyze the conversation content and assess the risk of harassment based on the following: phrases used, tone of voice, and participants' facial expressions. Issue an alert if specific factors are deemed high-risk."

[0752] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0753] Step 1:

[0754] The system acquires audio and video data from the user's device. The input is ambient sound and video information, and the output is digital audio and video files. In this step, the device's microphone and camera are used to collect data in real time.

[0755] Step 2:

[0756] The device preprocesses the acquired data and sends it to the server. It converts audio data to text (e.g., Google Speech-to-Text API) and extracts facial expression data from video (e.g., OpenCV). The input is raw audio and video data, and the output is preprocessed text and image data.

[0757] Step 3:

[0758] The server performs unstructured data analysis on text data to understand the meaning and context of the statements. The input is text data, and the output is the analyzed contextual information. At this stage, natural language processing techniques (e.g., NLTK, spaCy) are used to analyze the conversation.

[0759] Step 4:

[0760] The server performs emotion analysis, quantifying the user's emotional state based on their voice tone and facial expressions. The input is voice tone and facial expression data, and the output is an emotion score based on this data. Here, machine learning techniques (e.g., TensorFlow, PyTorch) are used for emotion evaluation.

[0761] Step 5:

[0762] The server calculates the risk of inappropriate behavior based on the analysis results, visualizes this risk, and provides feedback to the user. The input is contextual information and sentiment score, and the output is warning information provided to the user. This feedback is communicated to the user in real time by the terminal through display and audio.

[0763] Step 6:

[0764] Based on the feedback received, users adjust their communication methods and content on the spot. They revise specific actions based on the feedback, reducing the risks of the interaction. The input is feedback information, and the output is the result of the adjusted communication.

[0765] 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.

[0766] This invention is a system for determining the risk of harassment by utilizing user voice and video data. Furthermore, by combining it with an emotion engine, it achieves more accurate emotion recognition, enabling a detailed evaluation of the user's emotional state and the provision of appropriate feedback.

[0767] The user launches an application installed on their device to collect voice and facial expression data. The device records this data in real time, preprocesses it, and then sends it to a server. The server analyzes the data using natural language processing and emotion recognition technologies.

[0768] As an example, let's consider a workplace meeting. In this scenario, the user's statements and reactions are recorded in real time, and an emotion engine evaluates their current emotional state and stress level. Based on this analysis, the server determines how the user's statements are affecting others and generates a risk alert if inappropriate language is found.

[0769] Furthermore, the emotion engine accumulates past emotional history and learns how users react in specific situations. This allows the server to predict the user's emotional state and, if necessary, proactively suggest stress management and relaxation methods.

[0770] Users receive feedback sent from the server to their device, gaining clues to improve their communication. This feedback is provided as visual feedback through text and diagrams, or as audio guidance. In this way, users can achieve more flexible and adaptive communication in their interactions with others.

[0771] This invention can be used in a wide range of settings, including workplaces, educational institutions, and public spaces, and provides support for people to interact with each other in a safe and secure environment.

[0772] The following describes the processing flow.

[0773] Step 1:

[0774] The user launches a dedicated application on their device and begins capturing audio and video. The device's microphone and camera record the conversation and the user's facial expressions in real time.

[0775] Step 2:

[0776] The device preprocesses the recorded audio and video data. The audio data is de-noised and converted into text data using speech recognition. For the video data, facial recognition technology is used to extract features for sentiment analysis.

[0777] Step 3:

[0778] The device sends pre-processed speech-to-text data and sentiment features to the server. The transmission uses an encrypted communication protocol to ensure data security.

[0779] Step 4:

[0780] The server uses a natural language processing (NLP) model to analyze the received data, extracting the context and meaning of the conversation from the text data.

[0781] Step 5:

[0782] The server activates an emotion engine to evaluate the user's emotional state based on voice intonation and facial expression data. This evaluation also includes the user's stress and tension levels.

[0783] Step 6:

[0784] The server integrates NLP analysis and emotion engine results to score the harassment risk of a statement. If necessary, it generates alerts for statements that require improvement.

[0785] Step 7:

[0786] The server sends the generated risk assessment and feedback to the user's device. The feedback is presented as specific advice and visualized data.

[0787] Step 8:

[0788] The user's device receives feedback from the server and notifies the user visually and audibly. Based on the feedback received, the user can adjust their actions on the spot to improve communication.

[0789] (Example 2)

[0790] 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".

[0791] In recent years, harassment and inappropriate communication in interpersonal interactions in the workplace and educational settings have become a significant social problem. Such behavior negatively impacts individuals' mental and physical health and reduces overall organizational performance; therefore, systems are needed to prevent it. Furthermore, there is a need for technology that can provide appropriate feedback in diverse interaction settings and alert individuals to risky behaviors in real time.

[0792] 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.

[0793] In this invention, the server includes means for acquiring voice and image data with an information processing device, means for organizing the acquired data for data analysis and transmitting it to a remote processing device, and means for performing automatic processing and sentiment evaluation with the remote processing device to determine the risk of interpersonal behavior. This makes it possible to identify harassment risks in complex interpersonal interactions in real time and provide information users with guidelines for appropriate behavior.

[0794] "Audio and image data" refers to audio and visual information obtained from the user.

[0795] An "information processing device" is an electronic device used to acquire data and perform initial data organization.

[0796] "Data analysis" refers to the process of processing acquired data and transforming it into a useful form of information.

[0797] A "remote processing device" is a server or cloud system that receives acquired data and performs further advanced analysis on it.

[0798] "Automated processing" refers to computational processes that analyze and evaluate data without human intervention.

[0799] "Emotional assessment" is a process for determining a user's emotional state based on data.

[0800] "Interpersonal behavior risk" refers to the possibility that a user's behavior may be inappropriate or harmful to others.

[0801] "Information users" refer to users who receive the system's results and feedback.

[0802] "Visualization" refers to the process of displaying the results of data analysis in a visually easy-to-understand format.

[0803] "Evaluation value" refers to an indicator that quantifies a specific emotional state or risk.

[0804] A "warning" is the act of providing users with information that draws attention to a specific risk or issue.

[0805] "Real-time" refers to a process where data analysis and feedback are performed immediately on the spot.

[0806] This invention relates to a system that performs sentiment analysis and interpersonal behavior risk assessment using a user's voice and image data. This system consists of an information processing device, a remote processing device, and software modules that work in conjunction with them.

[0807] The user collects their own voice and facial expressions through an information processing device. This device consists of a typical mobile device or computer, and uses a camera and microphone as needed. This allows the user to acquire their own communication data. For example, it would record voice and facial expressions in real time during meetings or conversations.

[0808] The terminal processes the collected audio and image data, performing tasks such as noise reduction and facial recognition. Existing software is used for noise reduction, and libraries such as the OpenCV are used for facial recognition. The processed data is transmitted to the remote processing unit via encrypted communication.

[0809] The server acts as a remote processing unit, analyzing the received data. The server implements advanced automated processing technologies for natural language processing and sentiment evaluation. It performs speech recognition and evaluates emotional states based on the extracted text data. Specifically, it can utilize natural language processing APIs from common cloud services.

[0810] Based on the analysis results, the server determines the impact and risks of the user's statements on others and generates warnings as needed. Furthermore, it can learn from past data and predict the user's response in specific scenarios. This information is provided as real-time feedback to the device. Users can use the received feedback to try to improve their communication. The feedback is provided in text and visual formats, taking usability into consideration.

[0811] As a concrete example, in a workplace meeting, if a user speaks in an inappropriate tone, a warning will appear on their device immediately afterward. This warning includes information suggesting ways to improve their speech and explaining the potential future benefits.

[0812] Using a generative AI model, an example of a prompt message directed at a user might be, "Please suggest ways for the user to relax when they feel stressed in a specific situation."

[0813] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0814] Step 1:

[0815] The user launches an application on the information processing device. Audio and image data are acquired as input via the device's microphone and camera. This data is primary data recording the user's speech and facial expressions. Specifically, the information processing device begins recording and video, capturing the user's communication in real time.

[0816] Step 2:

[0817] The device performs noise reduction on the acquired audio data and applies a face recognition algorithm to the image data. This removes ambient noise from the audio and extracts only the user's facial expressions from the image. The output consists of pre-processed audio waveform data and clipped image data of the face region. The specific operation involves executing an audio processing algorithm and face recognition using OpenCV.

[0818] Step 3:

[0819] The terminal sends pre-processed audio and image data to the server. The data is encrypted using a secure protocol and transmitted via the server's API endpoint. The input is an encrypted data packet, and the output is a notification to the server indicating that data reception is complete. Specifically, the terminal establishes an HTTPS connection and sends the data to the server.

[0820] Step 4:

[0821] The server passes the received audio data to a natural language processing engine for conversion into text. Next, it passes the image data to an emotion evaluation engine to extract emotion information based on facial expressions. Inputs include audio waveforms and facial images, while outputs generate text data and emotion scores. Specific operations include calling speech recognition APIs and executing emotion analysis algorithms.

[0822] Step 5:

[0823] Based on the analysis results, the server evaluates the impact of a user's statements on others and generates risk alerts as needed. Furthermore, it predicts the user's emotional state by combining this with historical sentiment data. The inputs are text data and sentiment scores, and the outputs are a risk assessment report and predictive data. The specific operation involves applying statistical algorithms and predictive models.

[0824] Step 6:

[0825] The server generates and sends feedback to the user's terminal. The feedback is provided in a visual and audio instructional format and includes specific improvement suggestions. Inputs include risk assessment reports and predictive data, while output is the generation of feedback messages. Specific actions include creating the feedback format and sending the messages.

[0826] Step 7:

[0827] The user receives feedback displayed on their device and uses it to improve communication. The input is feedback messages from the server, and the output is the formation of a concrete action plan for future interactions. These actions include reviewing the feedback and formulating individual countermeasures.

[0828] (Application Example 2)

[0829] 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".

[0830] In brick-and-mortar stores, staff often struggle to accurately assess customers' emotions and concerns, resulting in a failure to prevent customer dissatisfaction and harassment. Traditional methods often rely on the individual experience and judgment of staff, lacking the ability to adjust customer service in response to changes in emotions. Therefore, technical solutions are needed to improve the quality of customer service in brick-and-mortar stores.

[0831] 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.

[0832] In this invention, the server includes means for acquiring audio and video data at a user terminal, means for initial processing the acquired information and transferring it to an information processing device, and means for performing contextual understanding and emotion evaluation on the information processing device and determining risk. This makes it possible to immediately analyze customer emotions and provide customer service support information that allows staff to respond appropriately.

[0833] "Audio data" refers to information about speech and related sounds, including user speech and other audio signals.

[0834] "Video data" refers to visual information acquired by cameras and other devices, including data such as the user's facial expressions, gestures, and other visual elements.

[0835] "User terminal" refers to a computer or smart device used to acquire and process data, specifically a device operated by the user.

[0836] An "information processing device" refers to a computer system that analyzes and processes data, and is a device that processes information using a specific algorithm.

[0837] "Contextual understanding" is the process of analyzing the meaning and intent of audio and text data to grasp linguistic and situational implications.

[0838] "Emotional assessment" is a technical process that identifies emotions from a user's voice and video data and evaluates that emotional state quantitatively or qualitatively.

[0839] "Assessing risk" is the process of evaluating potential risks and problems based on analyzed data, and determining their presence and degree.

[0840] "Customer service support information" refers to information provided to support smooth communication with customers, specifically real-time response strategies and guidance.

[0841] To implement this invention, users (staff) in a physical store wear a terminal such as smart glasses to acquire conversation data with customers. The terminal collects acoustic data using a microphone and records video data using a camera. After initial processing within the terminal, this data is transferred to an information processing device (server) in the cloud.

[0842] The server uses natural language processing software to understand the context of the data. It analyzes the linguistic features of audio data using Google Cloud's "Cloud Natural Language API." Furthermore, for video data, it estimates the emotional state from the user's facial expressions using an emotion assessment algorithm. Machine learning frameworks such as TensorFlow are utilized in this process.

[0843] Subsequently, based on the analyzed emotional and contextual information, the server determines the level of risk and sends the result as real-time feedback to the user's device. For example, if a customer expresses dissatisfaction with the service, customer service support information such as "Try to remain calm in this situation" will be displayed on the device's screen. This allows staff to respond instantly to the situation.

[0844] An example of a prompt for a generative AI model is, "Suggest how to handle a situation where a customer is angry." This is used by the server to provide appropriate advice to staff in real time.

[0845] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0846] Step 1:

[0847] The user wears a smart device and begins interacting with customers in the store. Input consists of audio data collected by the device's microphone and video data recorded by the camera. Output is this data in a format prepared for initial processing.

[0848] Step 2:

[0849] The terminal performs initial processing of the collected audio and video data. Specifically, it performs processes such as noise reduction and identifying areas of interest (faces) within the video using face detection algorithms. The output is pre-processed data that is converted into a format that can be sent to the server.

[0850] Step 3:

[0851] The server receives pre-processed data sent from the terminal. The input consists of pre-processed audio and video data. The server uses Google Cloud's "Cloud Natural Language API" to perform natural language processing on the audio data and analyze its linguistic features. For the video data, it performs sentiment evaluation using TensorFlow to estimate the user's emotional state. The output is a set of contextual information and sentiment information.

[0852] Step 4:

[0853] The server determines the risk level based on contextual and sentiment information. It uses a machine learning model to compare the results with historical data. The output is the determined risk level and the corresponding real-time countermeasures.

[0854] Step 5:

[0855] The server sends the generated risk assessment results and customer service support information to the user's terminal. The input consists of the assessment results and customer service support information. The output is customer service support information displayed on the terminal's screen, which specifically includes advice such as "If the customer expresses dissatisfaction, respond calmly."

[0856] Step 6:

[0857] Users optimize their customer service based on the support information provided. In practice, they are required to provide customer-friendly service while following the instructions on the terminal display.

[0858] 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.

[0859] 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.

[0860] 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.

[0861] 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.

[0862] 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.

[0863] 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.

[0864] 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.

[0865] 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.

[0866] 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."

[0867] 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.

[0868] 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.

[0869] 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.

[0870] 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.

[0871] 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.

[0872] 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.

[0873] 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.

[0874] 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.

[0875] 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.

[0876] 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.

[0877] 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.

[0878] 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.

[0879] The following is further disclosed regarding the embodiments described above.

[0880] (Claim 1)

[0881] A means of acquiring audio and video data on the user's terminal,

[0882] A means of preprocessing the acquired data and sending it to the server,

[0883] A method for determining the risk of harassment by performing natural language processing and sentiment analysis on a server,

[0884] A means of visualizing the judgment results and providing them to the user as feedback,

[0885] A system that includes this.

[0886] (Claim 2)

[0887] The system according to claim 1, which generates scores for the degree of discomfort and anxiety based on the results of emotional analysis.

[0888] (Claim 3)

[0889] The system according to claim 1, which sends alerts to users in real time.

[0890] "Example 1"

[0891] (Claim 1)

[0892] A means for recording audio and video digital data using an information processing device,

[0893] A means for pre-processing recorded digital data and transmitting it to a data processing device,

[0894] A data processing device performs information processing technology and state analysis to provide a means for evaluating the risks in human relationships.

[0895] A means for notifying the information processing device of the evaluation results as a visual display,

[0896] A system that includes this.

[0897] (Claim 2)

[0898] The system according to claim 1, which generates an evaluation index score based on the results of state analysis.

[0899] (Claim 3)

[0900] The system according to claim 1, which transmits notifications to an information processing device in real time.

[0901] "Application Example 1"

[0902] (Claim 1)

[0903] A means for acquiring audio and video data on the user's computing device,

[0904] A means for preprocessing acquired information and transmitting it to a remote computer,

[0905] A means of determining the risk of inappropriate behavior in communication by performing unstructured data analysis and emotional analysis using a remote computer,

[0906] A means of visualizing the judgment results and providing them to the user as feedback,

[0907] A means of monitoring communication in user-specified scenarios, calculating emotional risk in real time, and providing feedback,

[0908] A system that includes this.

[0909] (Claim 2)

[0910] The system according to claim 1, which generates indicators of psychological discomfort and anxiety based on the results of emotional analysis.

[0911] (Claim 3)

[0912] The system according to claim 1, which immediately issues an alarm to the user.

[0913] "Example 2 of combining an emotion engine"

[0914] (Claim 1)

[0915] A means for acquiring audio and image data using an information processing device,

[0916] A means for organizing the acquired data for data analysis and transmitting it to a remote processing device,

[0917] A means of determining the risk of interpersonal behavior by performing automated processing and emotional evaluation using a remote processing device,

[0918] A means of visualizing the judgment results based on the information and providing them to the information user as results,

[0919] A means of accumulating past emotional history and learning responses in specific situations,

[0920] A means of predicting one's own emotional state and suggesting behavioral management and calming methods in advance,

[0921] A system that includes this.

[0922] (Claim 2)

[0923] The system according to claim 1, which generates evaluation values ​​for the degree of discomfort and the degree of anxiety based on the results of an emotional evaluation.

[0924] (Claim 3)

[0925] The system according to claim 1, which issues a warning to the information user in real time.

[0926] "Application example 2 when combining with an emotional engine"

[0927] (Claim 1)

[0928] A means of acquiring audio and video data on the user's terminal,

[0929] A means for initial processing the acquired information and transferring it to an information processing device,

[0930] A means for determining risk by performing contextual understanding and sentiment evaluation using an information processing device,

[0931] A means of visually representing the judgment results and providing them to the user as feedback,

[0932] A system that includes means for providing customer service support information based on analysis results.

[0933] (Claim 2)

[0934] The system according to claim 1, which generates indicators of impression and mental stress based on the results of emotional evaluation.

[0935] (Claim 3)

[0936] The system according to claim 1, which immediately issues a warning to the user. [Explanation of symbols]

[0937] 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 of acquiring audio and video data on the user's terminal, A means of preprocessing the acquired data and sending it to the server, A method for determining the risk of harassment by performing natural language processing and sentiment analysis on a server, A means of visualizing the judgment results and providing them to the user as feedback, A system that includes this.

2. The system according to claim 1, which generates scores for the degree of discomfort and anxiety based on the results of emotional analysis.

3. The system according to claim 1, which sends alerts to users in real time.

Images