system

Abstract

We provide the system. [Solution] A voice input means for collecting ambient sounds, A speech recognition method that converts collected audio into text data, A natural language processing method that analyzes specific keywords and phrases from converted text data, Assessment methods for evaluating fraud risk based on historical data, A warning generation method that generates a warning when it is determined that there is a risk of fraud, An alert transmission means that transmits the generated warning via a communication means, 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 the 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】 Japanese Unexamined Patent Application Publication No. 2022-180282 【Summary of the Invention】 【Problems to be Solved by the Invention】 【0004】 In modern times when cash card fraud is increasing, consumers, especially the elderly, are frequently victimized by fraud. Such fraud methods have become sophisticated, and it is difficult to detect them at an early stage with conventional methods. Therefore, there is a need for a highly reliable system that can detect fraud in real time and issue warnings, because there is a risk that the consumer may expand the damage without noticing the fraud. 【Means for Solving the Problems】 【0005】 This invention provides a system that assesses the risk of fraud by using a voice input means to collect ambient sound and convert that sound into text data, and a natural language processing means to analyze specific keywords and phrases from the converted text data. Based on past data, the system assesses the risk of fraud and provides an alert sending means that automatically generates a warning if it is determined that fraud is possible and sends it to pre-registered contacts via a communication means. This enables consumers to immediately sense when they are exposed to the risk of fraud and to respond quickly. 【0006】 "Voice input means" refers to a device or system that has the function of collecting sound from the surroundings and processing it as digital data. 【0007】 "Speech recognition means" refers to a device or system that has the function of converting collected speech data into text data. 【0008】 A "natural language processing device" is a device or system that has the function of analyzing specific keywords or phrases from text data and extracting information from them. 【0009】 "Evaluation means" refers to a device or system that has the function of determining the risk of fraud by using analysis results based on past data. 【0010】 A "warning generation means" is a device or system that has the function of automatically generating a warning when it determines that there is a possibility of fraud. 【0011】 An "alert transmission means" is a device or system that has the function of sending generated alerts to a specified destination via a network. [Brief explanation of the drawing] 【0012】 [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2]This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, which incorporates an emotion engine. [Figure 14] This is a sequence diagram showing the processing flow of the data processing system in Application Example 2, which combines an emotion engine. [Modes for carrying out the invention] 【0013】 Hereinafter, an example of an embodiment of the system relating to the technology of this disclosure will be described with reference to the attached drawings. 【0014】 First, the terms used in the following description will be explained. 【0015】 In the following embodiments, the numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like. 【0016】 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. 【0017】 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. 【0018】 In the following embodiments, the numbered communication I / F (Interface) is an interface including a communication processor and an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark), etc. 【0019】 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." 【0020】 [First Embodiment] 【0021】 Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment. 【0022】 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. 【0023】 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). 【0024】 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. 【0025】 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. 【0026】 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. 【0027】 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. 【0028】 Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14. 【0029】 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. 【0030】 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. 【0031】 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. 【0032】 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". 【0033】 This invention is implemented as a system that detects cash card fraud in real time and issues a rapid warning. The user holds a terminal equipped with voice input, which constantly monitors surrounding audio data. This allows the user's phone conversations and surrounding audio to be collected in real time. 【0034】 The device converts the collected audio data into text data using speech recognition. This text data is then analyzed using natural language processing to check for the presence of specific keywords or phrases. During this process, an AI model evaluates whether the data is potentially fraudulent based on fraud patterns it has learned in the past. 【0035】 If the device determines that there is a high probability of fraud, the warning generation mechanism will activate and automatically generate warning information. This warning will include the content of the detected suspicious conversation and its evaluation score. Subsequently, the warning information will be sent via an alert sending mechanism to pre-registered contacts such as family members and the police. 【0036】 As a concrete example, consider a scenario where a user is having a phone conversation that includes phrases such as "bank," "hand over cash card," and "PIN." In this case, the device collects the audio and converts it to text using speech recognition. Natural language processing analyzes this text, and an AI model determines that there is a high risk of fraud. As a result, a warning generation system prepares a warning, and an alert sending system sends the warning to the user's family. The family receives the notification and contacts the user for confirmation. 【0037】 In this way, the present invention provides a highly effective means for users to prevent themselves from becoming victims of fraud. 【0038】 The following describes the processing flow. 【0039】 Step 1: 【0040】 The device constantly monitors the sounds around the user and continuously collects audio data through the microphone. 【0041】 Step 2: 【0042】 The device converts the collected audio data into text data using speech recognition technology. During this process, noise is removed, and the content of the conversation is accurately transcribed into text. 【0043】 Step 3: 【0044】 The device passes the transcribed data to a natural language processing system to analyze specific keywords and phrases. The analysis uses an algorithm that detects characteristic words related to fraud. 【0045】 Step 4: 【0046】 The device uses an AI model to evaluate the analysis results and assigns a score to the likelihood of fraud. This score is based on historical data and numerically indicates the degree of likelihood of fraud. 【0047】 Step 5: 【0048】 The device generates a warning using a warning generation mechanism if the fraud score exceeds a threshold. This warning includes details and the score of the suspicious conversation. 【0049】 Step 6: 【0050】 The device sends the generated warning to the server via an alert transmission mechanism. The server receives this warning information and prepares to send notifications to registered family members or the police. 【0051】 Step 7: 【0052】 The server initiates network communication to exchange information to be sent and sends an alert to pre-registered contacts. Family members review the content and contact the user directly if necessary. 【0053】 This process allows users to be warned early if they are exposed to the risk of fraud. 【0054】 (Example 1) 【0055】 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." 【0056】 In recent years, ATM card fraud and other fraudulent activities have become increasingly sophisticated, making it more likely that many individuals will fall victim. Such fraudulent activities need to be detected and warnings issued quickly, but traditional systems struggle to respond in real time. Furthermore, as fraudulent methods are constantly evolving, a flexible system that can adapt to these changes is required. 【0057】 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. 【0058】 In this invention, the server includes data input means for collecting ambient sound, recognition means for converting the collected sound into text information, and natural language processing means for analyzing specific vocabulary and phrases from the converted text information. This makes it possible to assess the risk of fraud in real time using a newly learned generative model, to immediately detect fraudulent activity, and to quickly issue warnings. 【0059】 "Data input means" refers to a function or device for collecting ambient sounds. 【0060】 "Recognition means" refers to a function or device for converting collected audio into textual information. 【0061】 "Natural language processing means" refers to a function or device for analyzing specific vocabulary or phrases from text data. 【0062】 A "generative model" is an algorithm or system that learns from past data and uses that learning to assess the risk of fraud. 【0063】 "Evaluation means" refers to a function or device that evaluates the risk of fraud based on analyzed data. 【0064】 A "warning generation means" is a function or device for generating a warning when it is determined that there is a risk of fraud. 【0065】 "Alert transmission means" refers to a function or device for transmitting generated warnings via communication means. 【0066】 "Network communication means" refers to a communication protocol or infrastructure for transmitting alerts and messages to pre-registered contacts. 【0067】 This invention is implemented as a system that detects cash card fraud in real time and issues a rapid warning. This system uses a user-held terminal equipped with voice input capabilities. The terminal is equipped with a high-performance microphone and dedicated voice recognition software (e.g., a common cloud-based voice recognition API), so it can constantly monitor and collect surrounding sounds. 【0068】 The device uses collected voice data to convert it into text information using speech recognition, and then utilizes natural language processing software (e.g., open-source NLP libraries) to analyze it. At this stage, a generative AI model is used to detect patterns of fraudulent activity based on a historical database and to assess the risk. If the assessment exceeds a certain threshold, a warning is immediately generated and the alert is sent via network communication to pre-registered contacts (e.g., family or police agencies). 【0069】 As a concrete example, consider a situation where a user is having a conversation that includes phrases such as "bank," "hand over cash card," and "PIN." In this case, the device collects the audio and converts it into text using its speech recognition function. Then, a natural language processing system analyzes this text, and a generative AI model determines that the risk of fraud is high. As a result, the device generates a warning and sends it to the user's family using an alert sending system. Through this process, the user can prevent themselves from becoming a victim of fraud. 【0070】 An example of a prompt message is: "Propose a system that assesses the likelihood of fraud in real time based on user statements and generates and sends warnings as needed." This invention enables rapid detection and response to fraudulent activity and provides an effective means of protecting users. 【0071】 The flow of the specific processing in Example 1 will be explained using Figure 11. 【0072】 Step 1: Collect audio data 【0073】 The device constantly monitors surrounding sounds using its voice input function. This function collects the user's phone conversations and surrounding sounds in real time. The input data is raw audio data. No data processing is performed at this stage, and the device proceeds directly to the next step. 【0074】 Step 2: Converting audio to text 【0075】 The terminal uses speech recognition software (e.g., cloud-based speech recognition API) to convert collected speech data into text information. The input is the speech data collected in the previous step, and the output is the converted text information. The speech signal is analyzed digitally and processed into linguistic text. 【0076】 Step 3: Text information analysis and fraud risk assessment 【0077】 The device uses natural language processing software to analyze textual information and detect specific keywords and phrases. The input is the textual information from step 2, and the output is a vocabulary list resulting from the analysis. Based on this list, a generative AI model uses historical data to assess fraud risk and generate a risk score. The data calculation involves scoring using a language model. 【0078】 Step 4: Generate a warning 【0079】 The device activates its warning generation function when the risk score from the evaluation results exceeds a set threshold. The input is the risk evaluation score, and the output is warning information. This output includes the content of conversations that are highly likely to be fraudulent and the evaluation score. Warning generation is performed by an automated message generation script. 【0080】 Step 5: Send an alert 【0081】 The device sends generated warning information to pre-registered contacts via its alert transmission function. The input is the warning information, and the output is a notification that transmission is complete. Communication is carried out using a network communication system, via email or messaging services. 【0082】 (Application Example 1) 【0083】 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." 【0084】 In modern society, fraud using personal information is rampant, and countermeasures against fraud conducted via telephone and everyday conversations are particularly urgent. Many conventional fraud prevention methods only address fraud after it has occurred, making it difficult to prevent damage in the first place. Therefore, there is a need to develop a new system that can detect signs of fraud in real time and provide rapid notification. 【0085】 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. 【0086】 In this invention, the server includes voice input means for collecting ambient sounds, voice recognition means for converting the collected sounds into text data, natural language processing means for analyzing specific keywords and phrases from the converted text data, evaluation means for evaluating the possibility of fraud based on past information, warning generation means for generating a warning when it is determined that there is a possibility of fraud, notification transmission means for sending the generated warning via a communication means, and communication means for informing a third party of the warning content detected by voice along with relevant information. This makes it possible to detect the possibility of fraud in real time and quickly notify relevant organizations and parties. 【0087】 "Voice input means" refers to a device or method for collecting ambient audio data. 【0088】 "Speech recognition means" refers to the process or technology of converting collected speech data into text data. 【0089】 "Natural language processing methods" are technologies that analyze specific keywords and phrases from text data to understand their meaning and intent. 【0090】 An "evaluation tool" is a method or system that has the function of evaluating the possibility of fraud based on past information. 【0091】 A "warning generation method" is a system or method that automatically generates a warning when it is determined that there is a possibility of fraud. 【0092】 "Notification transmission means" refers to a system or method for transmitting generated warnings to relevant parties via communication means. 【0093】 "Communication means" refers to a technology or method for informing a third party of the content of a warning detected by voice, along with related information. 【0094】 In this invention, the user's device plays a central role. The device is equipped with a voice input means for collecting ambient sounds, and continuously acquires voice data through this voice input means. This voice data is converted into text data by a speech recognition means. A speech recognition library such as speech_recognition is used as the software for this purpose. 【0095】 The converted text data is analyzed using natural language processing (NLP) tools. Libraries such as Spacy can be used for NLP. The purpose of the analysis is to extract specific keywords and phrases and detect potential fraud. The analysis results are then evaluated by an evaluation tool to assess fraud risk based on historical data. 【0096】 If a fraudulent activity is deemed highly likely, a warning generation system is activated and generates a warning. This warning is sent to pre-registered contacts via a notification sending system. Services such as Twilio are commonly used for this communication. 【0097】 As a concrete example, consider a scenario where a user is having a conversation related to banking and the phrase "hand over the cash card" is detected in the audio. In this case, the system transcribes the audio into text, detects the relevant keyword, assesses the risk of fraud, and generates and sends a warning. 【0098】 An example of a prompt for a generative AI model might be, "Please tell me how to design a natural language processing system that detects potential fraud from a user's voice conversation and generates a warning." This prompt serves as a guide for the AI ​​in building an effective fraud detection model. 【0099】 The flow of a specific process in Application Example 1 will be explained using Figure 12. 【0100】 Step 1: 【0101】 The device collects ambient sounds using voice input. The input consists of ambient sounds and conversations, which are acquired as analog audio data. The output is digital audio data that becomes input to the speech recognition system. 【0102】 Step 2: 【0103】 The terminal converts digital audio data into text data using speech recognition. In this step, the speech_recognition library is used to extract the linguistic content of the audio as text information. The input is digital audio data, and the output is the converted text data. 【0104】 Step 3: 【0105】 The terminal analyzes the character data converted by a natural language processing (NLP) method. This method utilizes NLP libraries such as spacy to search for specific keywords and phrases and understand the context. The input is character data, and the output is fraud-related phrases and risk information as a result of the analysis. 【0106】 Step 4: 【0107】 The terminal uses an evaluation tool to assess the likelihood of fraud based on the analysis results. In this step, past fraud patterns and statistical data are referenced, and a generative AI model is used to calculate a fraud risk score. The input is the fraud-related analysis results, and the output is the fraud risk evaluation score. 【0108】 Step 5: 【0109】 The device activates a warning generation mechanism and creates warning information when it determines that there is a high risk of fraud. This information includes the basis for the likelihood of fraud and a risk score. The input is the fraud risk assessment score, and the output is the generated warning message. 【0110】 Step 6: 【0111】 The device sends the generated alert to registered contacts via a notification sending method. This step utilizes communication services such as Twilio to deliver the alert via email or SMS. The input is the alert message, and the output is the sent notification result. 【0112】 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. 【0113】 This invention relates to a system that detects cash card fraud in real time by analyzing the user's voice. The system utilizes a terminal equipped with voice input and is designed to constantly monitor surrounding sounds. The collected voice data is converted into text data by a speech recognition system. From this text data, specific keywords and phrases are extracted using a natural language processing system to evaluate the likelihood of fraud. 【0114】 In addition, this system incorporates an emotion engine that recognizes emotions from the user's voice. The emotion engine analyzes the user's stress level and emotional changes, and uses this information to assess the risk of fraud. The emotional data acts as a factor that reinforces the possibility of fraud, and if a high stress level is detected, a warning is immediately generated by the warning generation mechanism. 【0115】 The warning generation system generates warnings based on scored information regarding the risk of fraud. The alert transmission system then sends emotional information and details of the fraud risk to pre-registered contacts, such as family members or the police. The alerts are delivered quickly via network communication, enabling relevant parties to take immediate action. 【0116】 As a concrete example, consider a scenario where a user is having a phone conversation that includes keywords such as "personal information," "PIN," and "card renewal." In this case, the device collects the audio and converts it to text. A natural language processing system analyzes this text, and an emotion engine detects high stress levels from the user's voice tone. If the evaluation system determines that the user is at high risk, a warning generation system is activated, and an alert is quickly sent to family members via an alert transmission system. 【0117】 In this way, the present invention enhances the accuracy of fraud detection and significantly improves the prevention of damage by combining emotion analysis with conventional voice analysis. 【0118】 The following describes the processing flow. 【0119】 Step 1: 【0120】 The device continuously monitors the sounds around the user and collects digital audio data through the microphone. 【0121】 Step 2: 【0122】 The device processes the collected audio data using speech recognition and converts it into text data. This makes the content of the audio analyzable. 【0123】 Step 3: 【0124】 The device analyzes the converted text data using natural language processing techniques to extract specific keywords and phrases related to fraud. This analysis is based on existing fraud patterns. 【0125】 Step 4: 【0126】 The device uses an emotion engine to analyze emotional information from the user's voice. Emotions such as stress and anxiety are detected, and if a high stress level is determined, the emotional data is provided to the evaluation system. 【0127】 Step 5: 【0128】 The device combines data from natural language processing and data from an emotion engine, uses an evaluation method to quantify the risk of fraud, and generates a score. 【0129】 Step 6: 【0130】 The device generates a warning using a warning generation mechanism when the fraud risk score exceeds a threshold. This warning includes the topic of the conversation and the results of sentiment analysis. 【0131】 Step 7: 【0132】 The device sends the generated warning to the server via an alert sending mechanism. Based on the received information, the server sends real-time notifications to pre-registered contacts. 【0133】 Step 8: 【0134】 The user's family and related parties receive alerts from the server on devices such as smartphones and check the content. If necessary, they will contact the user directly and take appropriate action. 【0135】 This process enables highly accurate fraud detection and warnings that take into account both the user's conversation and emotions. 【0136】 (Example 2) 【0137】 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". 【0138】 With the advancement of communication technology in modern society, fraudulent activities targeting personal information are becoming more sophisticated. In this situation, traditional methods make it difficult to detect fraudulent activities in real time, increasing the risk of users becoming victims of fraud. Elderly people, in particular, are often targeted, and a rapid response is required, but current systems are insufficient. Therefore, there is a need to develop a new system that can instantly detect fraudulent activities using voice and emotional data and send warnings to relevant parties. 【0139】 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. 【0140】 In this invention, the server includes sound data acquisition means for collecting ambient sounds, sound data conversion means for converting acquired sounds into text information, language information analysis means for extracting specific words from the converted text information, emotion analysis means for identifying emotions from the user's voice, determination means for evaluating the possibility of fraud, warning information generation means for generating a warning based on the evaluation result, and information transmission means for transmitting the generated warning via communication. This makes it possible to detect the possibility of fraud in real time and quickly send warnings to the relevant parties. 【0141】 "Sound data acquisition means" refers to a device or function for collecting ambient sounds and storing or transmitting them in a usable format. 【0142】 "Sound data conversion means" refers to a process or technology for converting acquired sound into text information, and utilizes speech recognition technology. 【0143】 A "linguistic information analysis tool" is a means for extracting specific words or phrases from converted character information and analyzing the information. 【0144】 "Emotional analysis means" refers to technology that identifies emotions from the user's voice and analyzes that emotional information. 【0145】 The "determination method" refers to a function that evaluates and determines the possibility of fraud based on the results of sentiment analysis and linguistic information analysis. 【0146】 A "warning information generation method" is a means of generating a warning when it is determined that there is a risk of fraud. 【0147】 "Information transmission means" refers to a communication-based mechanism or function for quickly transmitting generated warnings to relevant parties. 【0148】 The system of the present invention is designed to detect the possibility of cash card fraud in real time using voice data spoken around the user. Specific embodiments thereof are described below. 【0149】 First, the terminal is equipped with a means for acquiring sound data, and it constantly monitors ambient sounds using a voice input device (microphone) to collect sound data. On the terminal, the collected sound data is converted into text information using a sound data conversion means. Speech recognition software is used in this process, and specifically, a general speech recognition API plays this role. 【0150】 Next, the server receives the converted text information and uses language information analysis tools to extract specific words and phrases. This process utilizes natural language processing libraries, such as open-source language processing libraries. When specific keywords are detected, sentiment analysis tools are used to evaluate their likelihood and identify emotions from the user's voice. This helps determine whether the user is experiencing high levels of stress or anxiety. 【0151】 The server uses a determination mechanism to comprehensively evaluate the likelihood of fraud based on these analysis results. If the risk of fraud is determined to be high, the warning information generation mechanism is activated and generates a warning. The generated warning is then quickly transmitted to the relevant parties via an information transmission mechanism. Network communication is used for information transmission, with email and messaging protocols used as appropriate. 【0152】 A concrete example is when a user is on the phone and the conversation includes a phrase like, "Tell me your PIN." In such a situation, the device monitors the conversation and converts the audio into text. The server then analyzes this information, detects a high stress level, assesses the likelihood of fraud, and generates a warning. This warning is immediately sent to relevant family and friends. 【0153】 An example of a prompt message that the AI ​​model can provide is, "This audio may be a scam. Please check it immediately." This prompt serves as reference information to help stakeholders take prompt action. 【0154】 The flow of the specific processing in Example 2 will be explained using Figure 13. 【0155】 Step 1: 【0156】 The terminal collects ambient sounds using sound data acquisition means. Specifically, the microphone is always active, capturing user speech and background sounds as digital audio data. The input for this step is an analog audio signal, and the output is digital audio data. 【0157】 Step 2: 【0158】 The device uses speech recognition software to convert collected audio data into text information. In this process, digital audio data is sent to a speech recognition API, which then converts it into a string of characters through phoneme analysis and associative memory. The input for this step is digital audio data, and the output is text data. Specifically, real-time speech recognition is performed using edge computing technology. 【0159】 Step 3: 【0160】 The server utilizes language information analysis tools to extract specific words and phrases related to fraud from the converted text information. It receives text data as input, performs morphological analysis using a natural language processing library, and generates a list of important terms as output. Specifically, it highlights the words that match the keyword dictionary. 【0161】 Step 4: 【0162】 The server uses emotion analysis tools to identify emotions from the user's voice and analyze stress and anxiety levels. The input for this step is voice and extracted phrases, and it generates an emotion score as output using acoustic feature extraction and an emotion model. Specifically, it tracks changes in voice tone and tempo to quantify emotions such as joy, anger, sadness, and happiness. 【0163】 Step 5: 【0164】 The server uses a judgment tool to analyze the results and evaluate the likelihood of fraud. The input for this step is a keyword list and sentiment score, and a risk assessment algorithm is used to calculate a fraud risk score as output. Specifically, it calculates the risk by comparing it with past case data. 【0165】 Step 6: 【0166】 The server generates a warning when the risk of fraud is high using the warning information generation mechanism. The input for this step is a fraud risk score, and the output is a warning message. Specifically, a warning template is created when the risk threshold is exceeded. 【0167】 Step 7: 【0168】 The server transmits the generated warnings to relevant parties via the communication network through the information transmission means. The input for this step is the warning message, and the output is the transmitted notification. Specifically, alerts are sent immediately using email or messaging protocols. 【0169】 (Application Example 2) 【0170】 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". 【0171】 In recent years, fraudulent activities such as cash card fraud have been increasing, and consumers, especially the elderly, are particularly vulnerable to becoming victims, making preventative measures an urgent necessity. However, existing fraud detection systems do not adequately perform risk assessment through emotion analysis from voice, making accurate detection of fraudulent activities difficult. Against this backdrop, there is a need to realize a system that can detect fraudulent activities with higher accuracy by also taking into account changes in the user's emotions. 【0172】 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. 【0173】 In this invention, the server includes voice acquisition means, voice conversion means, language analysis means, evaluation means, warning generation means, notification transmission means, and emotion analysis means. This enables highly accurate detection of fraudulent activity that takes into account emotion evaluation based on the user's voice, as well as rapid warning and response. 【0174】 "Sound acquisition means" refers to devices or technologies that have the function of collecting ambient sounds. 【0175】 "Speech conversion means" refers to technology for converting acquired speech into text information. 【0176】 "Linguistic analysis means" refers to techniques for identifying, extracting, and analyzing specific terms and phrases from converted text information. 【0177】 "Evaluation methods" refer to techniques for assessing the likelihood of fraudulent activity based on past information. 【0178】 A "warning generation method" is a technology for generating information to warn others when it is determined that there is a possibility of fraudulent activity. 【0179】 "Notification transmission means" refers to technology for transmitting generated alerts via communication means. 【0180】 "Emotional analysis methods" refer to technologies that analyze a user's voice and evaluate and extract emotions from it. 【0181】 The present invention is implemented as an integrated system including voice input means, voice conversion means, language analysis means, evaluation means, attention generation means, notification transmission means, and emotion analysis means. 【0182】 The device is equipped with a voice acquisition mechanism to collect ambient sounds in real time. The collected sounds are converted into text information by a speech recognition system (e.g., Google® Cloud Speech-to-Text) as a speech conversion mechanism. This text information is then analyzed by a natural language processing system (e.g., spaCy) using a language analysis mechanism to extract specific keywords and phrases. 【0183】 Simultaneously, the terminal uses an emotion analysis engine (e.g., IBM Watson® Tone Analyzer) as an emotion analysis tool to extract and evaluate emotions from voice data. This data is integrated by an evaluation tool and compared with past information to determine the risk of fraudulent activity. 【0184】 If a high risk is detected, an alert is generated by the alert generation system. The server then quickly sends the alert to pre-registered contacts (e.g., family, police) via the notification transmission system. This allows the user or relevant parties to take prompt action. 【0185】 For example, when a user is having a phone conversation that includes keywords such as "PIN" or "card information," the device detects these and evaluates the emotional changes as high-stress speech. If the result is deemed high-risk, a warning is sent to family members. 【0186】 An example of a prompt for a generative AI model is: "Please tell me about an effective algorithm for a system that detects potential fraud from voice data and performs real-time risk assessment based on changes in emotion." 【0187】 The flow of a specific process in Application Example 2 will be explained using Figure 14. 【0188】 Step 1: 【0189】 The device collects ambient sound in real time. Using a voice input method, it captures environmental sounds with a microphone and records them as digital data. This data becomes the audio data used for subsequent processing. 【0190】 Step 2: 【0191】 The collected audio data is converted into text data within the device by a speech recognition system (e.g., Google Cloud Speech-to-Text). The audio data, as input, is analyzed using a speech conversion method, and text data in sentence format is generated as output. 【0192】 Step 3: 【0193】 The terminal analyzes the generated text data using a natural language processing system (e.g., spaCy). The language analysis means takes the text data as input and processes it to identify and extract keywords and phrases. As a result, it outputs language elements that may be related to fraud. 【0194】 Step 4: 【0195】 In parallel, the terminal uses an emotion analysis engine (e.g., IBM Watson Tone Analyzer) to extract emotional information from the voice data. The emotion analysis tool takes the voice data as input, analyzes its pitch and tone changes, and evaluates the emotional state. The output is emotional data such as the user's stress level and emotional changes. 【0196】 Step 5: 【0197】 The server uses an evaluation tool to integrate extracted keywords and sentiment data, and compares them with historical information to assess the risk of fraudulent activity. Based on this input, the evaluation tool analyzes the data consistency and quantifies the likelihood of fraudulent activity. As output, it generates a risk assessment score. 【0198】 Step 6: 【0199】 If the server determines that the risk assessment score is high, it immediately generates warning information using a warning generation mechanism. This process creates a warning message indicating that danger is imminent to the user. 【0200】 Step 7: 【0201】 The generated alert information is sent by the server via a notification system to pre-registered contacts, allowing them to prepare for an emergency. As output, the alert is delivered to family members' or police communication devices. 【0202】 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. 【0203】 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. 【0204】 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. 【0205】 [Second Embodiment] 【0206】 Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment. 【0207】 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. 【0208】 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). 【0209】 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. 【0210】 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. 【0211】 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). 【0212】 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. 【0213】 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. 【0214】 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. 【0215】 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. 【0216】 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. 【0217】 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". 【0218】 This invention is implemented as a system that detects cash card fraud in real time and issues a rapid warning. The user holds a terminal equipped with voice input, which constantly monitors surrounding audio data. This allows the user's phone conversations and surrounding audio to be collected in real time. 【0219】 The device converts the collected audio data into text data using speech recognition. This text data is then analyzed using natural language processing to check for the presence of specific keywords or phrases. During this process, an AI model evaluates whether the data is potentially fraudulent based on fraud patterns it has learned in the past. 【0220】 If the device determines that there is a high probability of fraud, the warning generation mechanism will activate and automatically generate warning information. This warning will include the content of the detected suspicious conversation and its evaluation score. Subsequently, the warning information will be sent via an alert sending mechanism to pre-registered contacts such as family members and the police. 【0221】 As a concrete example, consider a scenario where a user is having a phone conversation that includes phrases such as "bank," "hand over cash card," and "PIN." In this case, the device collects the audio and converts it to text using speech recognition. Natural language processing analyzes this text, and an AI model determines that there is a high risk of fraud. As a result, a warning generation system prepares a warning, and an alert sending system sends the warning to the user's family. The family receives the notification and contacts the user for confirmation. 【0222】 In this way, the present invention provides a highly effective means for users to prevent themselves from becoming victims of fraud. 【0223】 The following describes the processing flow. 【0224】 Step 1: 【0225】 The device constantly monitors the sounds around the user and continuously collects audio data through the microphone. 【0226】 Step 2: 【0227】 The device converts the collected audio data into text data using speech recognition technology. During this process, noise is removed, and the content of the conversation is accurately transcribed into text. 【0228】 Step 3: 【0229】 The device passes the transcribed data to a natural language processing system to analyze specific keywords and phrases. The analysis uses an algorithm that detects characteristic words related to fraud. 【0230】 Step 4: 【0231】 The device uses an AI model to evaluate the analysis results and assigns a score to the likelihood of fraud. This score is based on historical data and numerically indicates the degree of likelihood of fraud. 【0232】 Step 5: 【0233】 The device generates a warning using a warning generation mechanism if the fraud score exceeds a threshold. This warning includes details and the score of the suspicious conversation. 【0234】 Step 6: 【0235】 The device sends the generated warning to the server via an alert transmission mechanism. The server receives this warning information and prepares to send notifications to registered family members or the police. 【0236】 Step 7: 【0237】 The server initiates network communication to exchange information to be sent and sends an alert to pre-registered contacts. Family members review the content and contact the user directly if necessary. 【0238】 This process allows users to be warned early if they are exposed to the risk of fraud. 【0239】 (Example 1) 【0240】 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." 【0241】 In recent years, ATM card fraud and other fraudulent activities have become increasingly sophisticated, making it more likely that many individuals will fall victim. Such fraudulent activities need to be detected and warnings issued quickly, but traditional systems struggle to respond in real time. Furthermore, as fraudulent methods are constantly evolving, a flexible system that can adapt to these changes is required. 【0242】 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. 【0243】 In this invention, the server includes data input means for collecting ambient sound, recognition means for converting the collected sound into text information, and natural language processing means for analyzing specific vocabulary and phrases from the converted text information. This makes it possible to assess the risk of fraud in real time using a newly learned generative model, to immediately detect fraudulent activity, and to quickly issue warnings. 【0244】 "Data input means" refers to a function or device for collecting ambient sounds. 【0245】 "Recognition means" refers to a function or device for converting collected audio into textual information. 【0246】 "Natural language processing means" refers to a function or device for analyzing specific vocabulary or phrases from text data. 【0247】 A "generative model" is an algorithm or system that learns from past data and uses that learning to assess the risk of fraud. 【0248】 "Evaluation means" refers to a function or device that evaluates the risk of fraud based on analyzed data. 【0249】 A "warning generation means" is a function or device for generating a warning when it is determined that there is a risk of fraud. 【0250】 "Alert transmission means" refers to a function or device for transmitting generated warnings via communication means. 【0251】 "Network communication means" refers to a communication protocol or infrastructure for transmitting alerts and messages to pre-registered contacts. 【0252】 This invention is implemented as a system that detects cash card fraud in real time and issues a rapid warning. This system uses a user-held terminal equipped with voice input capabilities. The terminal is equipped with a high-performance microphone and dedicated voice recognition software (e.g., a common cloud-based voice recognition API), so it can constantly monitor and collect surrounding sounds. 【0253】 The device uses collected voice data to convert it into text information using speech recognition, and then utilizes natural language processing software (e.g., open-source NLP libraries) to analyze it. At this stage, a generative AI model is used to detect patterns of fraudulent activity based on a historical database and to assess the risk. If the assessment exceeds a certain threshold, a warning is immediately generated and the alert is sent via network communication to pre-registered contacts (e.g., family or police agencies). 【0254】 As a concrete example, consider a situation where a user is having a conversation that includes phrases such as "bank," "hand over cash card," and "PIN." In this case, the device collects the audio and converts it into text using its speech recognition function. Then, a natural language processing system analyzes this text, and a generative AI model determines that the risk of fraud is high. As a result, the device generates a warning and sends it to the user's family using an alert sending system. Through this process, the user can prevent themselves from becoming a victim of fraud. 【0255】 An example of a prompt message is: "Propose a system that assesses the likelihood of fraud in real time based on user statements and generates and sends warnings as needed." This invention enables rapid detection and response to fraudulent activity and provides an effective means of protecting users. 【0256】 The flow of the specific processing in Example 1 will be explained using Figure 11. 【0257】 Step 1: Collect audio data 【0258】 The device constantly monitors surrounding sounds using its voice input function. This function collects the user's phone conversations and surrounding sounds in real time. The input data is raw audio data. No data processing is performed at this stage, and the device proceeds directly to the next step. 【0259】 Step 2: Converting audio to text 【0260】 The terminal uses speech recognition software (e.g., cloud-based speech recognition API) to convert collected speech data into text information. The input is the speech data collected in the previous step, and the output is the converted text information. The speech signal is analyzed digitally and processed into linguistic text. 【0261】 Step 3: Text information analysis and fraud risk assessment 【0262】 The device uses natural language processing software to analyze textual information and detect specific keywords and phrases. The input is the textual information from step 2, and the output is a vocabulary list resulting from the analysis. Based on this list, a generative AI model uses historical data to assess fraud risk and generate a risk score. The data calculation involves scoring using a language model. 【0263】 Step 4: Generate a warning 【0264】 The device activates its warning generation function when the risk score from the evaluation results exceeds a set threshold. The input is the risk evaluation score, and the output is warning information. This output includes the content of conversations that are highly likely to be fraudulent and the evaluation score. Warning generation is performed by an automated message generation script. 【0265】 Step 5: Send an alert 【0266】 The device sends generated warning information to pre-registered contacts via its alert transmission function. The input is the warning information, and the output is a notification that transmission is complete. Communication is carried out using a network communication system, via email or messaging services. 【0267】 (Application Example 1) 【0268】 Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal." 【0269】 In modern society, fraud using personal information is rampant, and countermeasures against fraud conducted via telephone and everyday conversations are particularly urgent. Many conventional fraud prevention methods only address fraud after it has occurred, making it difficult to prevent damage in the first place. Therefore, there is a need to develop a new system that can detect signs of fraud in real time and provide rapid notification. 【0270】 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. 【0271】 In this invention, the server includes voice input means for collecting ambient sounds, voice recognition means for converting the collected sounds into text data, natural language processing means for analyzing specific keywords and phrases from the converted text data, evaluation means for evaluating the possibility of fraud based on past information, warning generation means for generating a warning when it is determined that there is a possibility of fraud, notification transmission means for sending the generated warning via a communication means, and communication means for informing a third party of the warning content detected by voice along with relevant information. This makes it possible to detect the possibility of fraud in real time and quickly notify relevant organizations and parties. 【0272】 "Voice input means" refers to a device or method for collecting ambient audio data. 【0273】 "Speech recognition means" refers to the process or technology of converting collected speech data into text data. 【0274】 "Natural language processing methods" are technologies that analyze specific keywords and phrases from text data to understand their meaning and intent. 【0275】 An "evaluation tool" is a method or system that has the function of evaluating the possibility of fraud based on past information. 【0276】 A "warning generation method" is a system or method that automatically generates a warning when it is determined that there is a possibility of fraud. 【0277】 "Notification transmission means" refers to a system or method for transmitting generated warnings to relevant parties via communication means. 【0278】 "Communication means" refers to a technology or method for informing a third party of the content of a warning detected by voice, along with related information. 【0279】 In this invention, the user's terminal plays a central role. The terminal is equipped with voice input means for collecting ambient voices, and constantly captures voice data through this voice input means. This voice data is converted into character data by voice recognition means. As software used for this, there is a voice recognition library such as speech_recognition. 【0280】 The converted character data is analyzed by natural language processing means. Libraries such as spacy can be used for natural language processing. The purpose of the analysis is to extract specific keywords and phrases and detect the possibility of fraud. The analysis result is evaluated for the fraud risk based on past data by evaluation means. 【0281】 When it is determined that the possibility of fraud is high, the warning generation means activates and generates a warning. This warning is sent to the pre-registered contacts by the notification sending means. Services such as twilio are generally used for communication. 【0282】 As a specific example, consider the case where the phrase "hand over the cash card" is detected in the voice when the user is having a conversation related to the bank. In this case, the system converts the voice into text, detects the corresponding keyword, evaluates the fraud risk, and generates and sends a warning. 【0283】 As an example of the prompt sentence for the generation AI model, "Please teach me about the design of a natural language processing system that detects the possibility of fraud from the user's voice conversation and generates a warning." can be considered. This prompt sentence serves as a guide when the AI constructs an effective fraud detection model. 【0284】 The flow of the specific process in Application Example 1 will be described using FIG. 12. 【0285】 Step 1: 【0286】 The terminal collects ambient voices using voice input means. The input is environmental sounds or conversations, which are acquired as analog voice data. The output is digital voice data that serves as input to the voice recognition means. 【0287】 Step 2: 【0288】 The terminal converts digital voice data into character data using voice recognition means. In this step, the speech_recognition library is used to extract the linguistic content of the voice as character information. The input is digital voice data, and the output is the converted character data. 【0289】 Step 3: 【0290】 The terminal analyzes the character data converted by the natural language processing means. In this means, natural language processing libraries such as spacy are utilized to search for specific keywords and phrases and understand the context. The input is character data, and the output is fraud-related phrases and their risk information as the analysis result. 【0291】 Step 4: 【0292】 The terminal evaluates the possibility of fraud from the analysis result using evaluation means. In this step, past fraud patterns and statistical data are referred to, and a generative AI model is used to calculate a fraud risk score. The input is the fraud-related analysis result, and the output is the evaluation score of the fraud risk. 【0293】 Step 5: 【0294】 When the terminal determines that the risk of fraud is high, it activates the warning generation means to create warning information. This information includes the basis for the possibility of fraud and the risk score. The input is the evaluation score of the fraud risk, and the output is the generated warning message. 【0295】 Step 6: 【0296】 The device sends the generated alert to registered contacts via a notification sending method. This step utilizes communication services such as Twilio to deliver the alert via email or SMS. The input is the alert message, and the output is the sent notification result. 【0297】 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. 【0298】 This invention relates to a system that detects cash card fraud in real time by analyzing the user's voice. The system utilizes a terminal equipped with voice input and is designed to constantly monitor surrounding sounds. The collected voice data is converted into text data by a speech recognition system. From this text data, specific keywords and phrases are extracted using a natural language processing system to evaluate the likelihood of fraud. 【0299】 In addition, this system incorporates an emotion engine that recognizes emotions from the user's voice. The emotion engine analyzes the user's stress level and emotional changes, and uses this information to assess the risk of fraud. The emotional data acts as a factor that reinforces the possibility of fraud, and if a high stress level is detected, a warning is immediately generated by the warning generation mechanism. 【0300】 The warning generation system generates warnings based on scored information regarding the risk of fraud. The alert transmission system then sends emotional information and details of the fraud risk to pre-registered contacts, such as family members or the police. The alerts are delivered quickly via network communication, enabling relevant parties to take immediate action. 【0301】 As a specific example, consider a case where a user is having a conversation on the phone that includes keywords such as "personal information", "password", and "card update". At this time, the terminal collects the voice and converts it into text. The natural language processing means analyzes this, and further, the emotion engine detects high stress from the user's voice tone. As a result of the evaluation means determining it as a high risk, the warning generation means activates, and a warning is quickly sent to the family via the alert transmission means. 【0302】 In this way, in addition to normal voice analysis, the present invention combines emotion analysis to improve the detection accuracy of fraud and significantly enhance the prevention of damage. 【0303】 The following describes the processing flow. 【0304】 Step 1: 【0305】 The terminal continuously monitors the voices around the user and collects digital voice data through the microphone. 【0306】 Step 2: [[ID=二十一]] 【0307】 [[ID=二十四]]The terminal processes the collected voice data with voice recognition means and converts it into text data. As a result, the content of the voice is made into an analyzable form. 【0308】 Step 3: 【0309】 The terminal analyzes the converted text data with natural language processing means and extracts specific keywords and phrases related to fraud. This analysis is performed based on existing fraud patterns. 【0310】 Step 4: 【0311】 The terminal analyzes emotion information from the user's voice using the emotion engine. If emotions such as stress and anxiety are detected and it is determined that there is a high stress level, the emotion data is provided to the evaluation means. 【0312】 Step 5: 【0313】 The device combines data from natural language processing and data from an emotion engine, uses an evaluation method to quantify the risk of fraud, and generates a score. 【0314】 Step 6: 【0315】 The device generates a warning using a warning generation mechanism when the fraud risk score exceeds a threshold. This warning includes the topic of the conversation and the results of sentiment analysis. 【0316】 Step 7: 【0317】 The device sends the generated warning to the server via an alert sending mechanism. Based on the received information, the server sends real-time notifications to pre-registered contacts. 【0318】 Step 8: 【0319】 The user's family and related parties receive alerts from the server on devices such as smartphones and check the content. If necessary, they will contact the user directly and take appropriate action. 【0320】 This process enables highly accurate fraud detection and warnings that take into account both the user's conversation and emotions. 【0321】 (Example 2) 【0322】 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". 【0323】 With the advancement of communication technology in modern society, fraudulent activities targeting personal information are becoming more sophisticated. In this situation, traditional methods make it difficult to detect fraudulent activities in real time, increasing the risk of users becoming victims of fraud. Elderly people, in particular, are often targeted, and a rapid response is required, but current systems are insufficient. Therefore, there is a need to develop a new system that can instantly detect fraudulent activities using voice and emotional data and send warnings to relevant parties. 【0324】 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. 【0325】 In this invention, the server includes sound data acquisition means for collecting ambient sounds, sound data conversion means for converting acquired sounds into text information, language information analysis means for extracting specific words from the converted text information, emotion analysis means for identifying emotions from the user's voice, determination means for evaluating the possibility of fraud, warning information generation means for generating a warning based on the evaluation result, and information transmission means for transmitting the generated warning via communication. This makes it possible to detect the possibility of fraud in real time and quickly send warnings to the relevant parties. 【0326】 "Sound data acquisition means" refers to a device or function for collecting ambient sounds and storing or transmitting them in a usable format. 【0327】 "Sound data conversion means" refers to a process or technology for converting acquired sound into text information, and utilizes speech recognition technology. 【0328】 A "linguistic information analysis tool" is a means for extracting specific words or phrases from converted character information and analyzing the information. 【0329】 "Emotional analysis means" refers to technology that identifies emotions from the user's voice and analyzes that emotional information. 【0330】 The "determination method" refers to a function that evaluates and determines the possibility of fraud based on the results of sentiment analysis and linguistic information analysis. 【0331】 A "warning information generation method" is a means of generating a warning when it is determined that there is a risk of fraud. 【0332】 "Information transmission means" refers to a communication-based mechanism or function for quickly transmitting generated warnings to relevant parties. 【0333】 The system of the present invention is designed to detect the possibility of cash card fraud in real time using voice data spoken around the user. Specific embodiments thereof are described below. 【0334】 First, the terminal is equipped with a means for acquiring sound data, and it constantly monitors ambient sounds using a voice input device (microphone) to collect sound data. On the terminal, the collected sound data is converted into text information using a sound data conversion means. Speech recognition software is used in this process, and specifically, a general speech recognition API plays this role. 【0335】 Next, the server receives the converted text information and uses language information analysis tools to extract specific words and phrases. This process utilizes natural language processing libraries, such as open-source language processing libraries. When specific keywords are detected, sentiment analysis tools are used to evaluate their likelihood and identify emotions from the user's voice. This helps determine whether the user is experiencing high levels of stress or anxiety. 【0336】 The server uses a determination mechanism to comprehensively evaluate the likelihood of fraud based on these analysis results. If the risk of fraud is determined to be high, the warning information generation mechanism is activated and generates a warning. The generated warning is then quickly transmitted to the relevant parties via an information transmission mechanism. Network communication is used for information transmission, with email and messaging protocols used as appropriate. 【0337】 A concrete example is when a user is on the phone and the conversation includes a phrase like, "Tell me your PIN." In such a situation, the device monitors the conversation and converts the audio into text. The server then analyzes this information, detects a high stress level, assesses the likelihood of fraud, and generates a warning. This warning is immediately sent to relevant family and friends. 【0338】 An example of a prompt message that the AI ​​model can provide is, "This audio may be a scam. Please check it immediately." This prompt serves as reference information to help stakeholders take prompt action. 【0339】 The flow of the specific processing in Example 2 will be explained using Figure 13. 【0340】 Step 1: 【0341】 The terminal collects ambient sounds using sound data acquisition means. Specifically, the microphone is always active, capturing user speech and background sounds as digital audio data. The input for this step is an analog audio signal, and the output is digital audio data. 【0342】 Step 2: 【0343】 The device uses speech recognition software to convert collected audio data into text information. In this process, digital audio data is sent to a speech recognition API, which then converts it into a string of characters through phoneme analysis and associative memory. The input for this step is digital audio data, and the output is text data. Specifically, real-time speech recognition is performed using edge computing technology. 【0344】 Step 3: 【0345】 The server utilizes language information analysis tools to extract specific words and phrases related to fraud from the converted text information. It receives text data as input, performs morphological analysis using a natural language processing library, and generates a list of important terms as output. Specifically, it highlights the words that match the keyword dictionary. 【0346】 Step 4: 【0347】 The server uses emotion analysis tools to identify emotions from the user's voice and analyze stress and anxiety levels. The input for this step is voice and extracted phrases, and it generates an emotion score as output using acoustic feature extraction and an emotion model. Specifically, it tracks changes in voice tone and tempo to quantify emotions such as joy, anger, sadness, and happiness. 【0348】 Step 5: 【0349】 The server uses a judgment tool to analyze the results and evaluate the likelihood of fraud. The input for this step is a keyword list and sentiment score, and a risk assessment algorithm is used to calculate a fraud risk score as output. Specifically, it calculates the risk by comparing it with past case data. 【0350】 Step 6: 【0351】 The server generates a warning when the risk of fraud is high using the warning information generation mechanism. The input for this step is a fraud risk score, and the output is a warning message. Specifically, a warning template is created when the risk threshold is exceeded. 【0352】 Step 7: 【0353】 The server transmits the generated warnings to relevant parties via the communication network through the information transmission means. The input for this step is the warning message, and the output is the transmitted notification. Specifically, alerts are sent immediately using email or messaging protocols. 【0354】 (Application Example 2) 【0355】 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." 【0356】 In recent years, fraudulent activities such as cash card fraud have been increasing, and consumers, especially the elderly, are particularly vulnerable to becoming victims, making preventative measures an urgent necessity. However, existing fraud detection systems do not adequately perform risk assessment through emotion analysis from voice, making accurate detection of fraudulent activities difficult. Against this backdrop, there is a need to realize a system that can detect fraudulent activities with higher accuracy by also taking into account changes in the user's emotions. 【0357】 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. 【0358】 In this invention, the server includes voice acquisition means, voice conversion means, language analysis means, evaluation means, warning generation means, notification transmission means, and emotion analysis means. This enables highly accurate detection of fraudulent activity that takes into account emotion evaluation based on the user's voice, as well as rapid warning and response. 【0359】 "Sound acquisition means" refers to devices or technologies that have the function of collecting ambient sounds. 【0360】 "Speech conversion means" refers to technology for converting acquired speech into text information. 【0361】 "Linguistic analysis means" refers to techniques for identifying, extracting, and analyzing specific terms and phrases from converted text information. 【0362】 "Evaluation methods" refer to techniques for assessing the likelihood of fraudulent activity based on past information. 【0363】 A "warning generation method" is a technology for generating information to warn others when it is determined that there is a possibility of fraudulent activity. 【0364】 "Notification transmission means" refers to technology for transmitting generated alerts via communication means. 【0365】 "Emotional analysis methods" refer to technologies that analyze a user's voice and evaluate and extract emotions from it. 【0366】 The present invention is implemented as an integrated system including voice input means, voice conversion means, language analysis means, evaluation means, attention generation means, notification transmission means, and emotion analysis means. 【0367】 The device is equipped with a voice acquisition mechanism to collect ambient sounds in real time. The collected sounds are converted into text information by a speech recognition system (e.g., Google Cloud Speech-to-Text) as a speech conversion mechanism. This text information is then analyzed by a natural language processing system (e.g., spaCy) using a language analysis mechanism to extract specific keywords and phrases. 【0368】 Simultaneously, the terminal uses an emotion analysis engine (e.g., IBM Watson Tone Analyzer) as an emotion analysis tool to extract and evaluate emotions from voice data. This data is integrated by an evaluation tool and compared with past information to determine the risk of fraudulent activity. 【0369】 If a high risk is detected, an alert is generated by the alert generation system. The server then quickly sends the alert to pre-registered contacts (e.g., family, police) via the notification transmission system. This allows the user or relevant parties to take prompt action. 【0370】 For example, when a user is having a phone conversation that includes keywords such as "PIN" or "card information," the device detects these and evaluates the emotional changes as high-stress speech. If the result is deemed high-risk, a warning is sent to family members. 【0371】 An example of a prompt for a generative AI model is: "Please tell me about an effective algorithm for a system that detects potential fraud from voice data and performs real-time risk assessment based on changes in emotion." 【0372】 The flow of a specific process in Application Example 2 will be explained using Figure 14. 【0373】 Step 1: 【0374】 The device collects ambient sound in real time. Using a voice input method, it captures environmental sounds with a microphone and records them as digital data. This data becomes the audio data used for subsequent processing. 【0375】 Step 2: 【0376】 The collected audio data is converted into text data within the device by a speech recognition system (e.g., Google Cloud Speech-to-Text). The audio data, as input, is analyzed using a speech conversion method, and text data in sentence format is generated as output. 【0377】 Step 3: 【0378】 The terminal analyzes the generated text data using a natural language processing system (e.g., spaCy). The language analysis means takes the text data as input and processes it to identify and extract keywords and phrases. As a result, it outputs language elements that may be related to fraud. 【0379】 Step 4: 【0380】 In parallel, the terminal uses an emotion analysis engine (e.g., IBM Watson Tone Analyzer) to extract emotional information from the voice data. The emotion analysis tool takes the voice data as input, analyzes its pitch and tone changes, and evaluates the emotional state. The output is emotional data such as the user's stress level and emotional changes. 【0381】 Step 5: 【0382】 The server uses an evaluation tool to integrate extracted keywords and sentiment data, and compares them with historical information to assess the risk of fraudulent activity. Based on this input, the evaluation tool analyzes the data consistency and quantifies the likelihood of fraudulent activity. As output, it generates a risk assessment score. 【0383】 Step 6: 【0384】 If the server determines that the risk assessment score is high, it immediately generates warning information using a warning generation mechanism. This process creates a warning message indicating that danger is imminent to the user. 【0385】 Step 7: 【0386】 The generated alert information is sent by the server via a notification system to pre-registered contacts, allowing them to prepare for an emergency. As output, the alert is delivered to family members' or police communication devices. 【0387】 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. 【0388】 Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). An 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. 【0389】 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. 【0390】 [Third Embodiment] 【0391】 Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment. 【0392】 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. 【0393】 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). 【0394】 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. 【0395】 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. 【0396】 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). 【0397】 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. 【0398】 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. 【0399】 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. 【0400】 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. 【0401】 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. 【0402】 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". 【0403】 This invention is implemented as a system that detects cash card fraud in real time and issues a rapid warning. The user holds a terminal equipped with voice input, which constantly monitors surrounding audio data. This allows the user's phone conversations and surrounding audio to be collected in real time. 【0404】 The device converts the collected audio data into text data using speech recognition. This text data is then analyzed using natural language processing to check for the presence of specific keywords or phrases. During this process, an AI model evaluates whether the data is potentially fraudulent based on fraud patterns it has learned in the past. 【0405】 If the device determines that there is a high probability of fraud, the warning generation mechanism will activate and automatically generate warning information. This warning will include the content of the detected suspicious conversation and its evaluation score. Subsequently, the warning information will be sent via an alert sending mechanism to pre-registered contacts such as family members and the police. 【0406】 As a concrete example, consider a scenario where a user is having a phone conversation that includes phrases such as "bank," "hand over cash card," and "PIN." In this case, the device collects the audio and converts it to text using speech recognition. Natural language processing analyzes this text, and an AI model determines that there is a high risk of fraud. As a result, a warning generation system prepares a warning, and an alert sending system sends the warning to the user's family. The family receives the notification and contacts the user for confirmation. 【0407】 In this way, the present invention provides a highly effective means for users to prevent themselves from becoming victims of fraud. 【0408】 The following describes the processing flow. 【0409】 Step 1: 【0410】 The device constantly monitors the sounds around the user and continuously collects audio data through the microphone. 【0411】 Step 2: 【0412】 The device converts the collected audio data into text data using speech recognition technology. During this process, noise is removed, and the content of the conversation is accurately transcribed into text. 【0413】 Step 3: 【0414】 The device passes the transcribed data to a natural language processing system to analyze specific keywords and phrases. The analysis uses an algorithm that detects characteristic words related to fraud. 【0415】 Step 4: 【0416】 The device uses an AI model to evaluate the analysis results and assigns a score to the likelihood of fraud. This score is based on historical data and numerically indicates the degree of likelihood of fraud. 【0417】 Step 5: 【0418】 The device generates a warning using a warning generation mechanism if the fraud score exceeds a threshold. This warning includes details and the score of the suspicious conversation. 【0419】 Step 6: 【0420】 The device sends the generated warning to the server via an alert transmission mechanism. The server receives this warning information and prepares to send notifications to registered family members or the police. 【0421】 Step 7: 【0422】 The server initiates network communication to exchange information to be sent and sends an alert to pre-registered contacts. Family members review the content and contact the user directly if necessary. 【0423】 This process allows users to be warned early if they are exposed to the risk of fraud. 【0424】 (Example 1) 【0425】 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." 【0426】 In recent years, ATM card fraud and other fraudulent activities have become increasingly sophisticated, making it more likely that many individuals will fall victim. Such fraudulent activities need to be detected and warnings issued quickly, but traditional systems struggle to respond in real time. Furthermore, as fraudulent methods are constantly evolving, a flexible system that can adapt to these changes is required. 【0427】 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. 【0428】 In this invention, the server includes data input means for collecting ambient sound, recognition means for converting the collected sound into text information, and natural language processing means for analyzing specific vocabulary and phrases from the converted text information. This makes it possible to assess the risk of fraud in real time using a newly learned generative model, to immediately detect fraudulent activity, and to quickly issue warnings. 【0429】 "Data input means" refers to a function or device for collecting ambient sounds. 【0430】 "Recognition means" refers to a function or device for converting collected audio into textual information. 【0431】 "Natural language processing means" refers to a function or device for analyzing specific vocabulary or phrases from text data. 【0432】 A "generative model" is an algorithm or system that learns from past data and uses that learning to assess the risk of fraud. 【0433】 "Evaluation means" refers to a function or device that evaluates the risk of fraud based on analyzed data. 【0434】 A "warning generation means" is a function or device for generating a warning when it is determined that there is a risk of fraud. 【0435】 "Alert transmission means" refers to a function or device for transmitting generated warnings via communication means. 【0436】 "Network communication means" refers to a communication protocol or infrastructure for transmitting alerts and messages to pre-registered contacts. 【0437】 This invention is implemented as a system that detects cash card fraud in real time and issues a rapid warning. This system uses a user-held terminal equipped with voice input capabilities. The terminal is equipped with a high-performance microphone and dedicated voice recognition software (e.g., a common cloud-based voice recognition API), so it can constantly monitor and collect surrounding sounds. 【0438】 The device uses collected voice data to convert it into text information using speech recognition, and then utilizes natural language processing software (e.g., open-source NLP libraries) to analyze it. At this stage, a generative AI model is used to detect patterns of fraudulent activity based on a historical database and to assess the risk. If the assessment exceeds a certain threshold, a warning is immediately generated and the alert is sent via network communication to pre-registered contacts (e.g., family or police agencies). 【0439】 As a concrete example, consider a situation where a user is having a conversation that includes phrases such as "bank," "hand over cash card," and "PIN." In this case, the device collects the audio and converts it into text using its speech recognition function. Then, a natural language processing system analyzes this text, and a generative AI model determines that the risk of fraud is high. As a result, the device generates a warning and sends it to the user's family using an alert sending system. Through this process, the user can prevent themselves from becoming a victim of fraud. 【0440】 An example of a prompt message is: "Propose a system that assesses the likelihood of fraud in real time based on user statements and generates and sends warnings as needed." This invention enables rapid detection and response to fraudulent activity and provides an effective means of protecting users. 【0441】 The flow of the specific processing in Example 1 will be explained using Figure 11. 【0442】 Step 1: Collect audio data 【0443】 The device constantly monitors surrounding sounds using its voice input function. This function collects the user's phone conversations and surrounding sounds in real time. The input data is raw audio data. No data processing is performed at this stage, and the device proceeds directly to the next step. 【0444】 Step 2: Converting audio to text 【0445】 The terminal uses speech recognition software (e.g., cloud-based speech recognition API) to convert collected speech data into text information. The input is the speech data collected in the previous step, and the output is the converted text information. The speech signal is analyzed digitally and processed into linguistic text. 【0446】 Step 3: Text information analysis and fraud risk assessment 【0447】 The device uses natural language processing software to analyze textual information and detect specific keywords and phrases. The input is the textual information from step 2, and the output is a vocabulary list resulting from the analysis. Based on this list, a generative AI model uses historical data to assess fraud risk and generate a risk score. The data calculation involves scoring using a language model. 【0448】 Step 4: Generate a warning 【0449】 The device activates its warning generation function when the risk score from the evaluation results exceeds a set threshold. The input is the risk evaluation score, and the output is warning information. This output includes the content of conversations that are highly likely to be fraudulent and the evaluation score. Warning generation is performed by an automated message generation script. 【0450】 Step 5: Send an alert 【0451】 The device sends generated warning information to pre-registered contacts via its alert transmission function. The input is the warning information, and the output is a notification that transmission is complete. Communication is carried out using a network communication system, via email or messaging services. 【0452】 (Application Example 1) 【0453】 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." 【0454】 In modern society, fraud using personal information is rampant, and countermeasures against fraud conducted via telephone and everyday conversations are particularly urgent. Many conventional fraud prevention methods only address fraud after it has occurred, making it difficult to prevent damage in the first place. Therefore, there is a need to develop a new system that can detect signs of fraud in real time and provide rapid notification. 【0455】 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. 【0456】 In this invention, the server includes voice input means for collecting ambient sounds, voice recognition means for converting the collected sounds into text data, natural language processing means for analyzing specific keywords and phrases from the converted text data, evaluation means for evaluating the possibility of fraud based on past information, warning generation means for generating a warning when it is determined that there is a possibility of fraud, notification transmission means for sending the generated warning via a communication means, and communication means for informing a third party of the warning content detected by voice along with relevant information. This makes it possible to detect the possibility of fraud in real time and quickly notify relevant organizations and parties. 【0457】 "Voice input means" refers to a device or method for collecting ambient audio data. 【0458】 "Speech recognition means" refers to the process or technology of converting collected speech data into text data. 【0459】 "Natural language processing methods" are technologies that analyze specific keywords and phrases from text data to understand their meaning and intent. 【0460】 An "evaluation tool" is a method or system that has the function of evaluating the possibility of fraud based on past information. 【0461】 A "warning generation method" is a system or method that automatically generates a warning when it is determined that there is a possibility of fraud. 【0462】 "Notification transmission means" refers to a system or method for transmitting generated warnings to relevant parties via communication means. 【0463】 "Communication means" refers to a technology or method for informing a third party of the content of a warning detected by voice, along with related information. 【0464】 In this invention, the user's device plays a central role. The device is equipped with a voice input means for collecting ambient sounds, and continuously acquires voice data through this voice input means. This voice data is converted into text data by a speech recognition means. A speech recognition library such as speech_recognition is used as the software for this purpose. 【0465】 The converted text data is analyzed using natural language processing (NLP) tools. Libraries such as Spacy can be used for NLP. The purpose of the analysis is to extract specific keywords and phrases and detect potential fraud. The analysis results are then evaluated by an evaluation tool to assess fraud risk based on historical data. 【0466】 If a fraudulent activity is deemed highly likely, a warning generation system is activated and generates a warning. This warning is sent to pre-registered contacts via a notification sending system. Services such as Twilio are commonly used for this communication. 【0467】 As a concrete example, consider a scenario where a user is having a conversation related to banking and the phrase "hand over the cash card" is detected in the audio. In this case, the system transcribes the audio into text, detects the relevant keyword, assesses the risk of fraud, and generates and sends a warning. 【0468】 An example of a prompt for a generative AI model might be, "Please tell me how to design a natural language processing system that detects potential fraud from a user's voice conversation and generates a warning." This prompt serves as a guide for the AI ​​in building an effective fraud detection model. 【0469】 The flow of a specific process in Application Example 1 will be explained using Figure 12. 【0470】 Step 1: 【0471】 The device collects ambient sounds using voice input. The input consists of ambient sounds and conversations, which are acquired as analog audio data. The output is digital audio data that becomes input to the speech recognition system. 【0472】 Step 2: 【0473】 The terminal converts digital audio data into text data using speech recognition. In this step, the speech_recognition library is used to extract the linguistic content of the audio as text information. The input is digital audio data, and the output is the converted text data. 【0474】 Step 3: 【0475】 The terminal analyzes the character data converted by a natural language processing (NLP) method. This method utilizes NLP libraries such as spacy to search for specific keywords and phrases and understand the context. The input is character data, and the output is fraud-related phrases and risk information as a result of the analysis. 【0476】 Step 4: 【0477】 The terminal uses an evaluation tool to assess the likelihood of fraud based on the analysis results. In this step, past fraud patterns and statistical data are referenced, and a generative AI model is used to calculate a fraud risk score. The input is the fraud-related analysis results, and the output is the fraud risk evaluation score. 【0478】 Step 5: 【0479】 The device activates a warning generation mechanism and creates warning information when it determines that there is a high risk of fraud. This information includes the basis for the likelihood of fraud and a risk score. The input is the fraud risk assessment score, and the output is the generated warning message. 【0480】 Step 6: 【0481】 The device sends the generated alert to registered contacts via a notification sending method. This step utilizes communication services such as Twilio to deliver the alert via email or SMS. The input is the alert message, and the output is the sent notification result. 【0482】 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. 【0483】 This invention relates to a system that detects cash card fraud in real time by analyzing the user's voice. The system utilizes a terminal equipped with voice input and is designed to constantly monitor surrounding sounds. The collected voice data is converted into text data by a speech recognition system. From this text data, specific keywords and phrases are extracted using a natural language processing system to evaluate the likelihood of fraud. 【0484】 In addition, this system incorporates an emotion engine that recognizes emotions from the user's voice. The emotion engine analyzes the user's stress level and emotional changes, and uses this information to assess the risk of fraud. The emotional data acts as a factor that reinforces the possibility of fraud, and if a high stress level is detected, a warning is immediately generated by the warning generation mechanism. 【0485】 The warning generation system generates warnings based on scored information regarding the risk of fraud. The alert transmission system then sends emotional information and details of the fraud risk to pre-registered contacts, such as family members or the police. The alerts are delivered quickly via network communication, enabling relevant parties to take immediate action. 【0486】 As a concrete example, consider a scenario where a user is having a phone conversation that includes keywords such as "personal information," "PIN," and "card renewal." In this case, the device collects the audio and converts it to text. A natural language processing system analyzes this text, and an emotion engine detects high stress levels from the user's voice tone. If the evaluation system determines that the user is at high risk, a warning generation system is activated, and an alert is quickly sent to family members via an alert transmission system. 【0487】 In this way, the present invention enhances the accuracy of fraud detection and significantly improves the prevention of damage by combining emotion analysis with conventional voice analysis. 【0488】 The following describes the processing flow. 【0489】 Step 1: 【0490】 The device continuously monitors the sounds around the user and collects digital audio data through the microphone. 【0491】 Step 2: 【0492】 The device processes the collected audio data using speech recognition and converts it into text data. This makes the content of the audio analyzable. 【0493】 Step 3: 【0494】 The device analyzes the converted text data using natural language processing techniques to extract specific keywords and phrases related to fraud. This analysis is based on existing fraud patterns. 【0495】 Step 4: 【0496】 The device uses an emotion engine to analyze emotional information from the user's voice. Emotions such as stress and anxiety are detected, and if a high stress level is determined, the emotional data is provided to the evaluation system. 【0497】 Step 5: 【0498】 The device combines data from natural language processing and data from an emotion engine, uses an evaluation method to quantify the risk of fraud, and generates a score. 【0499】 Step 6: 【0500】 The device generates a warning using a warning generation mechanism when the fraud risk score exceeds a threshold. This warning includes the topic of the conversation and the results of sentiment analysis. 【0501】 Step 7: 【0502】 The device sends the generated warning to the server via an alert sending mechanism. Based on the received information, the server sends real-time notifications to pre-registered contacts. 【0503】 Step 8: 【0504】 The user's family and related parties receive alerts from the server on devices such as smartphones and check the content. If necessary, they will contact the user directly and take appropriate action. 【0505】 This process enables highly accurate fraud detection and warnings that take into account both the user's conversation and emotions. 【0506】 (Example 2) 【0507】 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." 【0508】 With the advancement of communication technology in modern society, fraudulent activities targeting personal information are becoming more sophisticated. In this situation, traditional methods make it difficult to detect fraudulent activities in real time, increasing the risk of users becoming victims of fraud. Elderly people, in particular, are often targeted, and a rapid response is required, but current systems are insufficient. Therefore, there is a need to develop a new system that can instantly detect fraudulent activities using voice and emotional data and send warnings to relevant parties. 【0509】 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. 【0510】 In this invention, the server includes sound data acquisition means for collecting ambient sounds, sound data conversion means for converting acquired sounds into text information, language information analysis means for extracting specific words from the converted text information, emotion analysis means for identifying emotions from the user's voice, determination means for evaluating the possibility of fraud, warning information generation means for generating a warning based on the evaluation result, and information transmission means for transmitting the generated warning via communication. This makes it possible to detect the possibility of fraud in real time and quickly send warnings to the relevant parties. 【0511】 "Sound data acquisition means" refers to a device or function for collecting ambient sounds and storing or transmitting them in a usable format. 【0512】 "Sound data conversion means" refers to a process or technology for converting acquired sound into text information, and utilizes speech recognition technology. 【0513】 A "linguistic information analysis tool" is a means for extracting specific words or phrases from converted character information and analyzing the information. 【0514】 "Emotional analysis means" refers to technology that identifies emotions from the user's voice and analyzes that emotional information. 【0515】 The "determination method" refers to a function that evaluates and determines the possibility of fraud based on the results of sentiment analysis and linguistic information analysis. 【0516】 A "warning information generation method" is a means of generating a warning when it is determined that there is a risk of fraud. 【0517】 "Information transmission means" refers to a communication-based mechanism or function for quickly transmitting generated warnings to relevant parties. 【0518】 The system of the present invention is designed to detect the possibility of cash card fraud in real time using voice data spoken around the user. Specific embodiments thereof are described below. 【0519】 First, the terminal is equipped with a means for acquiring sound data, and it constantly monitors ambient sounds using a voice input device (microphone) to collect sound data. On the terminal, the collected sound data is converted into text information using a sound data conversion means. Speech recognition software is used in this process, and specifically, a general speech recognition API plays this role. 【0520】 Next, the server receives the converted text information and uses language information analysis tools to extract specific words and phrases. This process utilizes natural language processing libraries, such as open-source language processing libraries. When specific keywords are detected, sentiment analysis tools are used to evaluate their likelihood and identify emotions from the user's voice. This helps determine whether the user is experiencing high levels of stress or anxiety. 【0521】 The server uses a determination mechanism to comprehensively evaluate the likelihood of fraud based on these analysis results. If the risk of fraud is determined to be high, the warning information generation mechanism is activated and generates a warning. The generated warning is then quickly transmitted to the relevant parties via an information transmission mechanism. Network communication is used for information transmission, with email and messaging protocols used as appropriate. 【0522】 A concrete example is when a user is on the phone and the conversation includes a phrase like, "Tell me your PIN." In such a situation, the device monitors the conversation and converts the audio into text. The server then analyzes this information, detects a high stress level, assesses the likelihood of fraud, and generates a warning. This warning is immediately sent to relevant family and friends. 【0523】 An example of a prompt message that the AI ​​model can provide is, "This audio may be a scam. Please check it immediately." This prompt serves as reference information to help stakeholders take prompt action. 【0524】 The flow of the specific processing in Example 2 will be explained using Figure 13. 【0525】 Step 1: 【0526】 The terminal collects ambient sounds using sound data acquisition means. Specifically, the microphone is always active, capturing user speech and background sounds as digital audio data. The input for this step is an analog audio signal, and the output is digital audio data. 【0527】 Step 2: 【0528】 The device uses speech recognition software to convert collected audio data into text information. In this process, digital audio data is sent to a speech recognition API, which then converts it into a string of characters through phoneme analysis and associative memory. The input for this step is digital audio data, and the output is text data. Specifically, real-time speech recognition is performed using edge computing technology. 【0529】 Step 3: 【0530】 The server utilizes language information analysis tools to extract specific words and phrases related to fraud from the converted text information. It receives text data as input, performs morphological analysis using a natural language processing library, and generates a list of important terms as output. Specifically, it highlights the words that match the keyword dictionary. 【0531】 Step 4: 【0532】 The server uses emotion analysis tools to identify emotions from the user's voice and analyze stress and anxiety levels. The input for this step is voice and extracted phrases, and it generates an emotion score as output using acoustic feature extraction and an emotion model. Specifically, it tracks changes in voice tone and tempo to quantify emotions such as joy, anger, sadness, and happiness. 【0533】 Step 5: 【0534】 The server uses a judgment tool to analyze the results and evaluate the likelihood of fraud. The input for this step is a keyword list and sentiment score, and a risk assessment algorithm is used to calculate a fraud risk score as output. Specifically, it calculates the risk by comparing it with past case data. 【0535】 Step 6: 【0536】 The server generates a warning when the risk of fraud is high using the warning information generation mechanism. The input for this step is a fraud risk score, and the output is a warning message. Specifically, a warning template is created when the risk threshold is exceeded. 【0537】 Step 7: 【0538】 The server transmits the generated warnings to relevant parties via the communication network through the information transmission means. The input for this step is the warning message, and the output is the transmitted notification. Specifically, alerts are sent immediately using email or messaging protocols. 【0539】 (Application Example 2) 【0540】 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." 【0541】 In recent years, fraudulent activities such as cash card fraud have been increasing, and consumers, especially the elderly, are particularly vulnerable to becoming victims, making preventative measures an urgent necessity. However, existing fraud detection systems do not adequately perform risk assessment through emotion analysis from voice, making accurate detection of fraudulent activities difficult. Against this backdrop, there is a need to realize a system that can detect fraudulent activities with higher accuracy by also taking into account changes in the user's emotions. 【0542】 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. 【0543】 In this invention, the server includes voice acquisition means, voice conversion means, language analysis means, evaluation means, warning generation means, notification transmission means, and emotion analysis means. This enables highly accurate detection of fraudulent activity that takes into account emotion evaluation based on the user's voice, as well as rapid warning and response. 【0544】 "Sound acquisition means" refers to devices or technologies that have the function of collecting ambient sounds. 【0545】 "Speech conversion means" refers to technology for converting acquired speech into text information. 【0546】 "Linguistic analysis means" refers to techniques for identifying, extracting, and analyzing specific terms and phrases from converted text information. 【0547】 "Evaluation methods" refer to techniques for assessing the likelihood of fraudulent activity based on past information. 【0548】 A "warning generation method" is a technology for generating information to warn others when it is determined that there is a possibility of fraudulent activity. 【0549】 "Notification transmission means" refers to technology for transmitting generated alerts via communication means. 【0550】 "Emotional analysis methods" refer to technologies that analyze a user's voice and evaluate and extract emotions from it. 【0551】 The present invention is implemented as an integrated system including voice input means, voice conversion means, language analysis means, evaluation means, attention generation means, notification transmission means, and emotion analysis means. 【0552】 The device is equipped with a voice acquisition mechanism to collect ambient sounds in real time. The collected sounds are converted into text information by a speech recognition system (e.g., Google Cloud Speech-to-Text) as a speech conversion mechanism. This text information is then analyzed by a natural language processing system (e.g., spaCy) using a language analysis mechanism to extract specific keywords and phrases. 【0553】 Simultaneously, the terminal uses an emotion analysis engine (e.g., IBM Watson Tone Analyzer) as an emotion analysis tool to extract and evaluate emotions from voice data. This data is integrated by an evaluation tool and compared with past information to determine the risk of fraudulent activity. 【0554】 If a high risk is detected, an alert is generated by the alert generation system. The server then quickly sends the alert to pre-registered contacts (e.g., family, police) via the notification transmission system. This allows the user or relevant parties to take prompt action. 【0555】 For example, when a user is having a phone conversation that includes keywords such as "PIN" or "card information," the device detects these and evaluates the emotional changes as high-stress speech. If the result is deemed high-risk, a warning is sent to family members. 【0556】 An example of a prompt for a generative AI model is: "Please tell me about an effective algorithm for a system that detects potential fraud from voice data and performs real-time risk assessment based on changes in emotion." 【0557】 The flow of a specific process in Application Example 2 will be explained using Figure 14. 【0558】 Step 1: 【0559】 The device collects ambient sound in real time. Using a voice input method, it captures environmental sounds with a microphone and records them as digital data. This data becomes the audio data used for subsequent processing. 【0560】 Step 2: 【0561】 The collected audio data is converted into text data within the device by a speech recognition system (e.g., Google Cloud Speech-to-Text). The audio data, as input, is analyzed using a speech conversion method, and text data in sentence format is generated as output. 【0562】 Step 3: 【0563】 The terminal analyzes the generated text data using a natural language processing system (e.g., spaCy). The language analysis means takes the text data as input and processes it to identify and extract keywords and phrases. As a result, it outputs language elements that may be related to fraud. 【0564】 Step 4: 【0565】 In parallel, the terminal uses an emotion analysis engine (e.g., IBM Watson Tone Analyzer) to extract emotional information from the voice data. The emotion analysis tool takes the voice data as input, analyzes its pitch and tone changes, and evaluates the emotional state. The output is emotional data such as the user's stress level and emotional changes. 【0566】 Step 5: 【0567】 The server uses an evaluation tool to integrate extracted keywords and sentiment data, and compares them with historical information to assess the risk of fraudulent activity. Based on this input, the evaluation tool analyzes the data consistency and quantifies the likelihood of fraudulent activity. As output, it generates a risk assessment score. 【0568】 Step 6: 【0569】 If the server determines that the risk assessment score is high, it immediately generates warning information using a warning generation mechanism. This process creates a warning message indicating that danger is imminent to the user. 【0570】 Step 7: 【0571】 The generated alert information is sent by the server via a notification system to pre-registered contacts, allowing them to prepare for an emergency. As output, the alert is delivered to family members' or police communication devices. 【0572】 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. 【0573】 Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). An 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. 【0574】 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. 【0575】 [Fourth Embodiment] 【0576】 Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment. 【0577】 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. 【0578】 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). 【0579】 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. 【0580】 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. 【0581】 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). 【0582】 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. 【0583】 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. 【0584】 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. 【0585】 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. 【0586】 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. 【0587】 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. 【0588】 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". 【0589】 This invention is implemented as a system that detects cash card fraud in real time and issues a rapid warning. The user holds a terminal equipped with voice input, which constantly monitors surrounding audio data. This allows the user's phone conversations and surrounding audio to be collected in real time. 【0590】 The device converts the collected audio data into text data using speech recognition. This text data is then analyzed using natural language processing to check for the presence of specific keywords or phrases. During this process, an AI model evaluates whether the data is potentially fraudulent based on fraud patterns it has learned in the past. 【0591】 If the device determines that there is a high probability of fraud, the warning generation mechanism will activate and automatically generate warning information. This warning will include the content of the detected suspicious conversation and its evaluation score. Subsequently, the warning information will be sent via an alert sending mechanism to pre-registered contacts such as family members and the police. 【0592】 As a concrete example, consider a scenario where a user is having a phone conversation that includes phrases such as "bank," "hand over cash card," and "PIN." In this case, the device collects the audio and converts it to text using speech recognition. Natural language processing analyzes this text, and an AI model determines that there is a high risk of fraud. As a result, a warning generation system prepares a warning, and an alert sending system sends the warning to the user's family. The family receives the notification and contacts the user for confirmation. 【0593】 In this way, the present invention provides a highly effective means for users to prevent themselves from becoming victims of fraud. 【0594】 The following describes the processing flow. 【0595】 Step 1: 【0596】 The device constantly monitors the sounds around the user and continuously collects audio data through the microphone. 【0597】 Step 2: 【0598】 The device converts the collected audio data into text data using speech recognition technology. During this process, noise is removed, and the content of the conversation is accurately transcribed into text. 【0599】 Step 3: 【0600】 The device passes the transcribed data to a natural language processing system to analyze specific keywords and phrases. The analysis uses an algorithm that detects characteristic words related to fraud. 【0601】 Step 4: 【0602】 The device uses an AI model to evaluate the analysis results and assigns a score to the likelihood of fraud. This score is based on historical data and numerically indicates the degree of likelihood of fraud. 【0603】 Step 5: 【0604】 The device generates a warning using a warning generation mechanism if the fraud score exceeds a threshold. This warning includes details and the score of the suspicious conversation. 【0605】 Step 6: 【0606】 The device sends the generated warning to the server via an alert transmission mechanism. The server receives this warning information and prepares to send notifications to registered family members or the police. 【0607】 Step 7: 【0608】 The server initiates network communication to exchange information to be sent and sends an alert to pre-registered contacts. Family members review the content and contact the user directly if necessary. 【0609】 This process allows users to be warned early if they are exposed to the risk of fraud. 【0610】 (Example 1) 【0611】 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". 【0612】 In recent years, ATM card fraud and other fraudulent activities have become increasingly sophisticated, making it more likely that many individuals will fall victim. Such fraudulent activities need to be detected and warnings issued quickly, but traditional systems struggle to respond in real time. Furthermore, as fraudulent methods are constantly evolving, a flexible system that can adapt to these changes is required. 【0613】 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. 【0614】 In this invention, the server includes data input means for collecting ambient sound, recognition means for converting the collected sound into text information, and natural language processing means for analyzing specific vocabulary and phrases from the converted text information. This makes it possible to assess the risk of fraud in real time using a newly learned generative model, to immediately detect fraudulent activity, and to quickly issue warnings. 【0615】 "Data input means" refers to a function or device for collecting ambient sounds. 【0616】 "Recognition means" refers to a function or device for converting collected audio into textual information. 【0617】 "Natural language processing means" refers to a function or device for analyzing specific vocabulary or phrases from text data. 【0618】 A "generative model" is an algorithm or system that learns from past data and uses that learning to assess the risk of fraud. 【0619】 "Evaluation means" refers to a function or device that evaluates the risk of fraud based on analyzed data. 【0620】 A "warning generation means" is a function or device for generating a warning when it is determined that there is a risk of fraud. 【0621】 "Alert transmission means" refers to a function or device for transmitting generated warnings via communication means. 【0622】 "Network communication means" refers to a communication protocol or infrastructure for transmitting alerts and messages to pre-registered contacts. 【0623】 This invention is implemented as a system that detects cash card fraud in real time and issues a rapid warning. This system uses a user-held terminal equipped with voice input capabilities. The terminal is equipped with a high-performance microphone and dedicated voice recognition software (e.g., a common cloud-based voice recognition API), so it can constantly monitor and collect surrounding sounds. 【0624】 The device uses collected voice data to convert it into text information using speech recognition, and then utilizes natural language processing software (e.g., open-source NLP libraries) to analyze it. At this stage, a generative AI model is used to detect patterns of fraudulent activity based on a historical database and to assess the risk. If the assessment exceeds a certain threshold, a warning is immediately generated and the alert is sent via network communication to pre-registered contacts (e.g., family or police agencies). 【0625】 As a concrete example, consider a situation where a user is having a conversation that includes phrases such as "bank," "hand over cash card," and "PIN." In this case, the device collects the audio and converts it into text using its speech recognition function. Then, a natural language processing system analyzes this text, and a generative AI model determines that the risk of fraud is high. As a result, the device generates a warning and sends it to the user's family using an alert sending system. Through this process, the user can prevent themselves from becoming a victim of fraud. 【0626】 An example of a prompt message is: "Propose a system that assesses the likelihood of fraud in real time based on user statements and generates and sends warnings as needed." This invention enables rapid detection and response to fraudulent activity and provides an effective means of protecting users. 【0627】 The flow of the specific processing in Example 1 will be explained using Figure 11. 【0628】 Step 1: Collect audio data 【0629】 The device constantly monitors surrounding sounds using its voice input function. This function collects the user's phone conversations and surrounding sounds in real time. The input data is raw audio data. No data processing is performed at this stage, and the device proceeds directly to the next step. 【0630】 Step 2: Converting audio to text 【0631】 The terminal uses speech recognition software (e.g., cloud-based speech recognition API) to convert collected speech data into text information. The input is the speech data collected in the previous step, and the output is the converted text information. The speech signal is analyzed digitally and processed into linguistic text. 【0632】 Step 3: Text information analysis and fraud risk assessment 【0633】 The device uses natural language processing software to analyze textual information and detect specific keywords and phrases. The input is the textual information from step 2, and the output is a vocabulary list resulting from the analysis. Based on this list, a generative AI model uses historical data to assess fraud risk and generate a risk score. The data calculation involves scoring using a language model. 【0634】 Step 4: Generate a warning 【0635】 The device activates its warning generation function when the risk score from the evaluation results exceeds a set threshold. The input is the risk evaluation score, and the output is warning information. This output includes the content of conversations that are highly likely to be fraudulent and the evaluation score. Warning generation is performed by an automated message generation script. 【0636】 Step 5: Send an alert 【0637】 The device sends generated warning information to pre-registered contacts via its alert transmission function. The input is the warning information, and the output is a notification that transmission is complete. Communication is carried out using a network communication system, via email or messaging services. 【0638】 (Application Example 1) 【0639】 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". 【0640】 In modern society, fraud using personal information is rampant, and countermeasures against fraud conducted via telephone and everyday conversations are particularly urgent. Many conventional fraud prevention methods only address fraud after it has occurred, making it difficult to prevent damage in the first place. Therefore, there is a need to develop a new system that can detect signs of fraud in real time and provide rapid notification. 【0641】 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. 【0642】 In this invention, the server includes voice input means for collecting ambient sounds, voice recognition means for converting the collected sounds into text data, natural language processing means for analyzing specific keywords and phrases from the converted text data, evaluation means for evaluating the possibility of fraud based on past information, warning generation means for generating a warning when it is determined that there is a possibility of fraud, notification transmission means for sending the generated warning via a communication means, and communication means for informing a third party of the warning content detected by voice along with relevant information. This makes it possible to detect the possibility of fraud in real time and quickly notify relevant organizations and parties. 【0643】 "Voice input means" refers to a device or method for collecting ambient audio data. 【0644】 "Speech recognition means" refers to the process or technology of converting collected speech data into text data. 【0645】 "Natural language processing methods" are technologies that analyze specific keywords and phrases from text data to understand their meaning and intent. 【0646】 An "evaluation tool" is a method or system that has the function of evaluating the possibility of fraud based on past information. 【0647】 A "warning generation method" is a system or method that automatically generates a warning when it is determined that there is a possibility of fraud. 【0648】 "Notification transmission means" refers to a system or method for transmitting generated warnings to relevant parties via communication means. 【0649】 "Communication means" refers to a technology or method for informing a third party of the content of a warning detected by voice, along with related information. 【0650】 In this invention, the user's device plays a central role. The device is equipped with a voice input means for collecting ambient sounds, and continuously acquires voice data through this voice input means. This voice data is converted into text data by a speech recognition means. A speech recognition library such as speech_recognition is used as the software for this purpose. 【0651】 The converted text data is analyzed using natural language processing (NLP) tools. Libraries such as Spacy can be used for NLP. The purpose of the analysis is to extract specific keywords and phrases and detect potential fraud. The analysis results are then evaluated by an evaluation tool to assess fraud risk based on historical data. 【0652】 If a fraudulent activity is deemed highly likely, a warning generation system is activated and generates a warning. This warning is sent to pre-registered contacts via a notification sending system. Services such as Twilio are commonly used for this communication. 【0653】 As a concrete example, consider a scenario where a user is having a conversation related to banking and the phrase "hand over the cash card" is detected in the audio. In this case, the system transcribes the audio into text, detects the relevant keyword, assesses the risk of fraud, and generates and sends a warning. 【0654】 An example of a prompt for a generative AI model might be, "Please tell me how to design a natural language processing system that detects potential fraud from a user's voice conversation and generates a warning." This prompt serves as a guide for the AI ​​in building an effective fraud detection model. 【0655】 The flow of a specific process in Application Example 1 will be explained using Figure 12. 【0656】 Step 1: 【0657】 The device collects ambient sounds using voice input. The input consists of ambient sounds and conversations, which are acquired as analog audio data. The output is digital audio data that becomes input to the speech recognition system. 【0658】 Step 2: 【0659】 The terminal converts digital audio data into text data using speech recognition. In this step, the speech_recognition library is used to extract the linguistic content of the audio as text information. The input is digital audio data, and the output is the converted text data. 【0660】 Step 3: 【0661】 The terminal analyzes the character data converted by a natural language processing (NLP) method. This method utilizes NLP libraries such as spacy to search for specific keywords and phrases and understand the context. The input is character data, and the output is fraud-related phrases and risk information as a result of the analysis. 【0662】 Step 4: 【0663】 The terminal uses an evaluation tool to assess the likelihood of fraud based on the analysis results. In this step, past fraud patterns and statistical data are referenced, and a generative AI model is used to calculate a fraud risk score. The input is the fraud-related analysis results, and the output is the fraud risk evaluation score. 【0664】 Step 5: 【0665】 The device activates a warning generation mechanism and creates warning information when it determines that there is a high risk of fraud. This information includes the basis for the likelihood of fraud and a risk score. The input is the fraud risk assessment score, and the output is the generated warning message. 【0666】 Step 6: 【0667】 The device sends the generated alert to registered contacts via a notification sending method. This step utilizes communication services such as Twilio to deliver the alert via email or SMS. The input is the alert message, and the output is the sent notification result. 【0668】 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. 【0669】 This invention relates to a system that detects cash card fraud in real time by analyzing the user's voice. The system utilizes a terminal equipped with voice input and is designed to constantly monitor surrounding sounds. The collected voice data is converted into text data by a speech recognition system. From this text data, specific keywords and phrases are extracted using a natural language processing system to evaluate the likelihood of fraud. 【0670】 In addition, this system incorporates an emotion engine that recognizes emotions from the user's voice. The emotion engine analyzes the user's stress level and emotional changes, and uses this information to assess the risk of fraud. The emotional data acts as a factor that reinforces the possibility of fraud, and if a high stress level is detected, a warning is immediately generated by the warning generation mechanism. 【0671】 The warning generation system generates warnings based on scored information regarding the risk of fraud. The alert transmission system then sends emotional information and details of the fraud risk to pre-registered contacts, such as family members or the police. The alerts are delivered quickly via network communication, enabling relevant parties to take immediate action. 【0672】 As a concrete example, consider a scenario where a user is having a phone conversation that includes keywords such as "personal information," "PIN," and "card renewal." In this case, the device collects the audio and converts it to text. A natural language processing system analyzes this text, and an emotion engine detects high stress levels from the user's voice tone. If the evaluation system determines that the user is at high risk, a warning generation system is activated, and an alert is quickly sent to family members via an alert transmission system. 【0673】 In this way, the present invention enhances the accuracy of fraud detection and significantly improves the prevention of damage by combining emotion analysis with conventional voice analysis. 【0674】 The following describes the processing flow. 【0675】 Step 1: 【0676】 The device continuously monitors the sounds around the user and collects digital audio data through the microphone. 【0677】 Step 2: 【0678】 The device processes the collected audio data using speech recognition and converts it into text data. This makes the content of the audio analyzable. 【0679】 Step 3: 【0680】 The device analyzes the converted text data using natural language processing techniques to extract specific keywords and phrases related to fraud. This analysis is based on existing fraud patterns. 【0681】 Step 4: 【0682】 The device uses an emotion engine to analyze emotional information from the user's voice. Emotions such as stress and anxiety are detected, and if a high stress level is determined, the emotional data is provided to the evaluation system. 【0683】 Step 5: 【0684】 The device combines data from natural language processing and data from an emotion engine, uses an evaluation method to quantify the risk of fraud, and generates a score. 【0685】 Step 6: 【0686】 The device generates a warning using a warning generation mechanism when the fraud risk score exceeds a threshold. This warning includes the topic of the conversation and the results of sentiment analysis. 【0687】 Step 7: 【0688】 The device sends the generated warning to the server via an alert sending mechanism. Based on the received information, the server sends real-time notifications to pre-registered contacts. 【0689】 Step 8: 【0690】 The user's family and related parties receive alerts from the server on devices such as smartphones and check the content. If necessary, they will contact the user directly and take appropriate action. 【0691】 This process enables highly accurate fraud detection and warnings that take into account both the user's conversation and emotions. 【0692】 (Example 2) 【0693】 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". 【0694】 With the advancement of communication technology in modern society, fraudulent activities targeting personal information are becoming more sophisticated. In this situation, traditional methods make it difficult to detect fraudulent activities in real time, increasing the risk of users becoming victims of fraud. Elderly people, in particular, are often targeted, and a rapid response is required, but current systems are insufficient. Therefore, there is a need to develop a new system that can instantly detect fraudulent activities using voice and emotional data and send warnings to relevant parties. 【0695】 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. 【0696】 In this invention, the server includes sound data acquisition means for collecting ambient sounds, sound data conversion means for converting acquired sounds into text information, language information analysis means for extracting specific words from the converted text information, emotion analysis means for identifying emotions from the user's voice, determination means for evaluating the possibility of fraud, warning information generation means for generating a warning based on the evaluation result, and information transmission means for transmitting the generated warning via communication. This makes it possible to detect the possibility of fraud in real time and quickly send warnings to the relevant parties. 【0697】 "Sound data acquisition means" refers to a device or function for collecting ambient sounds and storing or transmitting them in a usable format. 【0698】 "Sound data conversion means" refers to a process or technology for converting acquired sound into text information, and utilizes speech recognition technology. 【0699】 A "linguistic information analysis tool" is a means for extracting specific words or phrases from converted character information and analyzing the information. 【0700】 "Emotional analysis means" refers to technology that identifies emotions from the user's voice and analyzes that emotional information. 【0701】 The "determination method" refers to a function that evaluates and determines the possibility of fraud based on the results of sentiment analysis and linguistic information analysis. 【0702】 A "warning information generation method" is a means of generating a warning when it is determined that there is a risk of fraud. 【0703】 "Information transmission means" refers to a communication-based mechanism or function for quickly transmitting generated warnings to relevant parties. 【0704】 The system of the present invention is designed to detect the possibility of cash card fraud in real time using voice data spoken around the user. Specific embodiments thereof are described below. 【0705】 First, the terminal is equipped with a means for acquiring sound data, and it constantly monitors ambient sounds using a voice input device (microphone) to collect sound data. On the terminal, the collected sound data is converted into text information using a sound data conversion means. Speech recognition software is used in this process, and specifically, a general speech recognition API plays this role. 【0706】 Next, the server receives the converted text information and uses language information analysis tools to extract specific words and phrases. This process utilizes natural language processing libraries, such as open-source language processing libraries. When specific keywords are detected, sentiment analysis tools are used to evaluate their likelihood and identify emotions from the user's voice. This helps determine whether the user is experiencing high levels of stress or anxiety. 【0707】 The server uses a determination mechanism to comprehensively evaluate the likelihood of fraud based on these analysis results. If the risk of fraud is determined to be high, the warning information generation mechanism is activated and generates a warning. The generated warning is then quickly transmitted to the relevant parties via an information transmission mechanism. Network communication is used for information transmission, with email and messaging protocols used as appropriate. 【0708】 A concrete example is when a user is on the phone and the conversation includes a phrase like, "Tell me your PIN." In such a situation, the device monitors the conversation and converts the audio into text. The server then analyzes this information, detects a high stress level, assesses the likelihood of fraud, and generates a warning. This warning is immediately sent to relevant family and friends. 【0709】 An example of a prompt message that the AI ​​model can provide is, "This audio may be a scam. Please check it immediately." This prompt serves as reference information to help stakeholders take prompt action. 【0710】 The flow of the specific processing in Example 2 will be explained using Figure 13. 【0711】 Step 1: 【0712】 The terminal collects ambient sounds using sound data acquisition means. Specifically, the microphone is always active, capturing user speech and background sounds as digital audio data. The input for this step is an analog audio signal, and the output is digital audio data. 【0713】 Step 2: 【0714】 The device uses speech recognition software to convert collected audio data into text information. In this process, digital audio data is sent to a speech recognition API, which then converts it into a string of characters through phoneme analysis and associative memory. The input for this step is digital audio data, and the output is text data. Specifically, real-time speech recognition is performed using edge computing technology. 【0715】 Step 3: 【0716】 The server utilizes language information analysis tools to extract specific words and phrases related to fraud from the converted text information. It receives text data as input, performs morphological analysis using a natural language processing library, and generates a list of important terms as output. Specifically, it highlights the words that match the keyword dictionary. 【0717】 Step 4: 【0718】 The server uses emotion analysis tools to identify emotions from the user's voice and analyze stress and anxiety levels. The input for this step is voice and extracted phrases, and it generates an emotion score as output using acoustic feature extraction and an emotion model. Specifically, it tracks changes in voice tone and tempo to quantify emotions such as joy, anger, sadness, and happiness. 【0719】 Step 5: 【0720】 The server uses a judgment tool to analyze the results and evaluate the likelihood of fraud. The input for this step is a keyword list and sentiment score, and a risk assessment algorithm is used to calculate a fraud risk score as output. Specifically, it calculates the risk by comparing it with past case data. 【0721】 Step 6: 【0722】 The server generates a warning when the risk of fraud is high using the warning information generation mechanism. The input for this step is a fraud risk score, and the output is a warning message. Specifically, a warning template is created when the risk threshold is exceeded. 【0723】 Step 7: 【0724】 The server transmits the generated warnings to relevant parties via the communication network through the information transmission means. The input for this step is the warning message, and the output is the transmitted notification. Specifically, alerts are sent immediately using email or messaging protocols. 【0725】 (Application Example 2) 【0726】 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". 【0727】 In recent years, fraudulent activities such as cash card fraud have been increasing, and consumers, especially the elderly, are particularly vulnerable to becoming victims, making preventative measures an urgent necessity. However, existing fraud detection systems do not adequately perform risk assessment through emotion analysis from voice, making accurate detection of fraudulent activities difficult. Against this backdrop, there is a need to realize a system that can detect fraudulent activities with higher accuracy by also taking into account changes in the user's emotions. 【0728】 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. 【0729】 In this invention, the server includes voice acquisition means, voice conversion means, language analysis means, evaluation means, warning generation means, notification transmission means, and emotion analysis means. This enables highly accurate detection of fraudulent activity that takes into account emotion evaluation based on the user's voice, as well as rapid warning and response. 【0730】 "Sound acquisition means" refers to devices or technologies that have the function of collecting ambient sounds. 【0731】 "Speech conversion means" refers to technology for converting acquired speech into text information. 【0732】 "Linguistic analysis means" refers to techniques for identifying, extracting, and analyzing specific terms and phrases from converted text information. 【0733】 "Evaluation methods" refer to techniques for assessing the likelihood of fraudulent activity based on past information. 【0734】 A "warning generation method" is a technology for generating information to warn others when it is determined that there is a possibility of fraudulent activity. 【0735】 "Notification transmission means" refers to technology for transmitting generated alerts via communication means. 【0736】 "Emotional analysis methods" refer to technologies that analyze a user's voice and evaluate and extract emotions from it. 【0737】 The present invention is implemented as an integrated system including voice input means, voice conversion means, language analysis means, evaluation means, attention generation means, notification transmission means, and emotion analysis means. 【0738】 The device is equipped with a voice acquisition mechanism to collect ambient sounds in real time. The collected sounds are converted into text information by a speech recognition system (e.g., Google Cloud Speech-to-Text) as a speech conversion mechanism. This text information is then analyzed by a natural language processing system (e.g., spaCy) using a language analysis mechanism to extract specific keywords and phrases. 【0739】 Simultaneously, the terminal uses an emotion analysis engine (e.g., IBM Watson Tone Analyzer) as an emotion analysis tool to extract and evaluate emotions from voice data. This data is integrated by an evaluation tool and compared with past information to determine the risk of fraudulent activity. 【0740】 If a high risk is detected, an alert is generated by the alert generation system. The server then quickly sends the alert to pre-registered contacts (e.g., family, police) via the notification transmission system. This allows the user or relevant parties to take prompt action. 【0741】 For example, when a user is having a phone conversation that includes keywords such as "PIN" or "card information," the device detects these and evaluates the emotional changes as high-stress speech. If the result is deemed high-risk, a warning is sent to family members. 【0742】 An example of a prompt for a generative AI model is: "Please tell me about an effective algorithm for a system that detects potential fraud from voice data and performs real-time risk assessment based on changes in emotion." 【0743】 The flow of a specific process in Application Example 2 will be explained using Figure 14. 【0744】 Step 1: 【0745】 The device collects ambient sound in real time. Using a voice input method, it captures environmental sounds with a microphone and records them as digital data. This data becomes the audio data used for subsequent processing. 【0746】 Step 2: 【0747】 The collected audio data is converted into text data within the device by a speech recognition system (e.g., Google Cloud Speech-to-Text). The audio data, as input, is analyzed using a speech conversion method, and text data in sentence format is generated as output. 【0748】 Step 3: 【0749】 The terminal analyzes the generated text data using a natural language processing system (e.g., spaCy). The language analysis means takes the text data as input and processes it to identify and extract keywords and phrases. As a result, it outputs language elements that may be related to fraud. 【0750】 Step 4: 【0751】 In parallel, the terminal uses an emotion analysis engine (e.g., IBM Watson Tone Analyzer) to extract emotional information from the voice data. The emotion analysis tool takes the voice data as input, analyzes its pitch and tone changes, and evaluates the emotional state. The output is emotional data such as the user's stress level and emotional changes. 【0752】 Step 5: 【0753】 The server uses an evaluation tool to integrate extracted keywords and sentiment data, and compares them with historical information to assess the risk of fraudulent activity. Based on this input, the evaluation tool analyzes the data consistency and quantifies the likelihood of fraudulent activity. As output, it generates a risk assessment score. 【0754】 Step 6: 【0755】 If the server determines that the risk assessment score is high, it immediately generates warning information using a warning generation mechanism. This process creates a warning message indicating that danger is imminent to the user. 【0756】 Step 7: 【0757】 The generated alert information is sent by the server via a notification system to pre-registered contacts, allowing them to prepare for an emergency. As output, the alert is delivered to family members' or police communication devices. 【0758】 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. 【0759】 Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). An 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. 【0760】 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. 【0761】 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. 【0762】 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. 【0763】 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. 【0764】 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. 【0765】 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. 【0766】 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." 【0767】 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. 【0768】 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. 【0769】 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. 【0770】 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. 【0771】 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. 【0772】 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. 【0773】 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. 【0774】 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. 【0775】 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. 【0776】 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. 【0777】 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. 【0778】 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. 【0779】 The following is further disclosed regarding the embodiments described above. 【0780】 (Claim 1) 【0781】 A voice input means for collecting ambient sounds, 【0782】 A speech recognition method that converts collected audio into text data, 【0783】 A natural language processing method that analyzes specific keywords and phrases from converted text data, 【0784】 Assessment methods for evaluating fraud risk based on historical data, 【0785】 A warning generation method that generates a warning when it is determined that there is a risk of fraud, 【0786】 An alert transmission means that transmits the generated warning via a communication means, 【0787】 A system that includes this. 【0788】 (Claim 2) 【0789】 The system according to claim 1, in which the natural language processing means updates fraud techniques in real time and can respond to newly emerging fraud techniques. 【0790】 (Claim 3) 【0791】 The system according to claim 1, wherein the alert transmission means utilizes a network communication means for transmitting alerts to pre-registered contacts. 【0792】 "Example 1" 【0793】 (Claim 1) 【0794】 A data input means for collecting ambient sounds, 【0795】 A recognition means for converting collected audio into text information, 【0796】 A natural language processing method that analyzes specific vocabulary or phrases from converted character information, 【0797】 An evaluation method for assessing fraud risk using a newly learned generative model, 【0798】 A warning generation means that generates a warning when it is determined that there is a risk of fraud, 【0799】 An alert transmission means that transmits the generated warning via a communication means, 【0800】 A system that includes this. 【0801】 (Claim 2) 【0802】 The system according to claim 1, in which the natural language processing means updates fraud techniques in real time and can respond to newly emerging fraud techniques. 【0803】 (Claim 3) 【0804】 The system according to claim 1, wherein the alert transmission means utilizes a network communication means for transmitting alerts to pre-registered contacts. 【0805】 "Application Example 1" 【0806】 (Claim 1) 【0807】 A voice input means for collecting ambient sounds, 【0808】 A speech recognition method that converts collected audio into text data, 【0809】 A natural language processing method that analyzes specific keywords or phrases from converted character data, 【0810】 An evaluation method for assessing the likelihood of fraud based on past information, 【0811】 A warning generation method that generates a warning when it is determined that there is a possibility of fraud, 【0812】 A notification transmission means that transmits the generated warning via a communication means, 【0813】 A communication method that notifies a third party of the warning content detected by voice, along with related information, 【0814】 A system that includes this. 【0815】 (Claim 2) 【0816】 The system according to claim 1, in which the natural language processing means updates fraud techniques in real time and is capable of responding to newly emerging fraud techniques. 【0817】 (Claim 3) 【0818】 The system according to claim 1, wherein the notification transmission means utilizes a communication network means for transmitting notifications to pre-registered contacts. 【0819】 "Example 2 of combining an emotion engine" 【0820】 (Claim 1) 【0821】 A means for acquiring sound data to collect ambient sounds, 【0822】 A sound data conversion means that converts acquired sound into text information, 【0823】 A language information analysis means for extracting specific words or phrases from converted character information, 【0824】 An emotion analysis method that identifies emotions from the user's voice, 【0825】 A means of determining the possibility of fraud, 【0826】 A warning information generation means that generates a warning based on the evaluation results, 【0827】 An information transmission means for transmitting the generated warnings via communication, 【0828】 A system that includes this. 【0829】 (Claim 2) 【0830】 The system according to claim 1, wherein the language information analysis means can update fraudulent methods in real time and respond to newly emerging fraudulent methods. 【0831】 (Claim 3) 【0832】 The system according to claim 1, wherein the information transmission means uses a communication network utilization means to transmit a warning to a pre-set communication destination. 【0833】 "Application example 2 when combining with an emotional engine" 【0834】 (Claim 1) 【0835】 A means for collecting ambient sounds, 【0836】 A speech conversion method that converts collected audio into text information, 【0837】 A language analysis tool that analyzes specific terms and phrases from converted text information, 【0838】 An evaluation method for assessing the possibility of misconduct based on past information, 【0839】 A warning generation means that generates a warning when it is determined that there is a possibility of fraudulent activity, 【0840】 A notification transmission means that transmits the generated alert via a communication means, 【0841】 An emotion analysis method that analyzes the user's voice and evaluates their emotions, 【0842】 An evaluation method that uses emotional information to assess fraud risk, 【0843】 A system that includes this. 【0844】 (Claim 2) 【0845】 The system according to claim 1, in which the language analysis means can immediately update the methods of fraudulent activity and respond to newly emerging methods of fraudulent activity. 【0846】 (Claim 3) 【0847】 The system according to claim 1, wherein the notification transmission means utilizes a communication network means for transmitting notifications to pre-registered contacts. [Explanation of Symbols] 【0848】 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

[Claim 1] A voice input means for collecting ambient sounds, A speech recognition method that converts collected audio into text data, A natural language processing method that analyzes specific keywords and phrases from converted text data, Assessment methods for evaluating fraud risk based on historical data, A warning generation method that generates a warning when it is determined that there is a risk of fraud, An alert transmission means that transmits the generated warning via a communication means, A system that includes this. [Claim 2] The system according to claim 1, in which the natural language processing means can update fraudulent methods in real time and respond to newly emerging fraudulent methods. [Claim 3] The system according to claim 1, wherein the alert transmission means utilizes a network communication means to transmit alerts to pre-registered contacts.

Images