system

Abstract

To effectively protect users from fraud and provide a safe and secure calling environment. [Solution] A means for receiving audio data and analyzing the speaker's emotions in real time, A method for detecting suspected fraud when analyzed emotions indicate confusion or anxiety, A means to generate a response strategy when fraud is suspected and to respond appropriately to the other party, A method of generating and speaking dialogue scripts that are used to psychologically unsettle con artists, A means of notifying users of the situation and providing safety advice, We provide a system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a persona chatbot control method performed by at least one processor, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In recent years, fraud through telephone has become increasingly sophisticated, and particularly, there have been many cases of victimization targeting the elderly. With conventional countermeasures, it is difficult to obtain a sufficient effect to prevent fraud in advance, and particularly, real-time response during a call is required. Specifically, there is a current situation where there is a lack of a prompt and appropriate countermeasure method when the recipient feels confused or anxious. In response to this problem, it is required to provide a more effective fraud prevention means. [[ID=3^6]]

Means for Solving the Problems

[0005] This invention provides a means for receiving voice data and analyzing the speaker's emotions in real time, enabling rapid detection of potential fraud when the recipient feels confused or anxious. Furthermore, if a high risk of fraud is determined, it immediately generates a response strategy and creates a dialogue script to psychologically unsettle the fraudster. This dialogue script is then spoken aloud to the fraudster using speech synthesis technology, achieving effective unsettlement. The system also provides users with status notifications and safety advice, and offers further reassurance through reports including call history and the basis for the fraud determination.

[0006] "Audio data" refers to information that represents sound in digital format, and is fundamental data for recording and processing conversations and sounds.

[0007] "Emotion analysis" is a technology that estimates a speaker's emotional state from voice and text data, and is a process that identifies emotions such as relief, confusion, and anxiety.

[0008] "Suspected fraud" refers to a state where signs indicating the possibility of fraudulent activity have been detected in a particular call or action, and it is a signal that you should be vigilant against potential fraudulent activity.

[0009] A "response strategy" refers to a plan or method for determining the optimal action or reaction in a particular situation, and is especially intended to be an effective way to manipulate a con artist during a conversation.

[0010] "Speech synthesis" is a technology that creates artificially generated, natural-sounding voices, and refers to the process of imitating human voices based on text information, for example.

[0011] "User" refers to a customer or user who utilizes this system and is a beneficiary of services aimed at preventing fraudulent calls and providing a safe calling experience.

[0012] A "report" is a document generated by the system after a call, containing a record of information such as call history and the basis for determining fraud. [Brief explanation of the drawing]

[0013] [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]

[0014] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0015] First, the terms used in the following description will be explained.

[0016] In the following embodiments, a processor with a reference numeral (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.

[0017] In the following embodiments, a RAM (Random Access Memory) with a reference numeral is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0018] In the following embodiments, a storage with a reference numeral is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, and the like.

[0019] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0020] 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."

[0021] [First Embodiment]

[0022] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0023] 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.

[0024] 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).

[0025] 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.

[0026] 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.

[0027] 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.

[0028] 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.

[0029] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] 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".

[0034] The system of this invention is designed to analyze incoming calls by users in real time and to take appropriate action based on their content. The server, terminal, and user components each perform their respective functions as follows.

[0035] server

[0036] The server plays a central role in receiving and analyzing audio data during calls. Specifically, the server uses a voice analysis engine to perform sentiment analysis, estimating basic emotional states such as reassurance, confusion, and anxiety from the speaker's voice. To detect typical scam phrases, natural language processing techniques are applied, and the call content is analyzed in real time as text data. If fraud is suspected, the server immediately generates a response strategy and creates a dialogue script to psychologically unsettle the scammer.

[0037] terminal

[0038] The device transmits the audio received by the user to the server and streams the data. It then synthesizes the response instructed by the server and transmits it to the scammer as actual audio. The device also notifies the user of the situation in real time and provides safety advice via voice or screen display.

[0039] User

[0040] Users are beneficiaries of this system, which protects them from fraudulent calls. Users understand that their calls are monitored and that advice will be provided as needed, ensuring a safe calling experience. After the call ends, users receive a report generated by the server via their device, allowing them to review call details and a determination based on the suspicion of fraud.

[0041] Specific example

[0042] As a concrete example, consider a scenario where User B receives a message from a scammer. The server detects phrases suggesting potential fraud through voice communication and senses User B's anxiety through sentiment analysis. The server then generates a question-and-answer dialogue designed to expose inconsistencies in the scammer's statements, and the terminal speaks this script. User B receives instructions from the terminal to perform multiple verifications, preventing the scammer from obtaining information, and after the call ends, can review a detailed report to understand the situation. In this way, the system enables immediate prevention of fraud.

[0043] The following describes the processing flow.

[0044] Step 1:

[0045] As soon as the device initiates a call, it streams the audio data to the server in real time.

[0046] Step 2:

[0047] The server processes the received audio data using a speech analysis engine, analyzing the speaker's tone, pitch, speed, etc., to estimate their emotional state.

[0048] Step 3:

[0049] The server uses a natural language processing module to convert audio data into text and detects typical scam phrases and keywords.

[0050] Step 4:

[0051] The server evaluates the sentiment analysis results and phrase detection results to determine whether there is a suspicion of fraud.

[0052] Step 5:

[0053] If fraud is suspected, the server generates a response strategy and creates a dialogue script to psychologically unsettle the fraudster.

[0054] Step 6:

[0055] The terminal receives a dialogue script from the server, synthesizes it into speech, and speaks it to the scammer.

[0056] Step 7:

[0057] The device notifies the user of the call status and provides safety advice via voice or on-screen display.

[0058] Step 8:

[0059] After the call ends, the server generates a report containing the call history and the basis for the fraud determination, and provides it to the user via the terminal.

[0060] (Example 1)

[0061] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0062] In modern society, telephone fraud is on the rise, and individuals need immediate and effective measures to protect themselves from these scams. However, conventional systems have difficulty detecting fraud during a call and providing appropriate responses in real time. Therefore, there is a need for a system that can detect fraud early during a call and provide users with appropriate responses and assistance information in real time.

[0063] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0064] In this invention, the server includes means for receiving voice information and analyzing the speaker's emotional state, means for detecting suspicion of fraud based on the analyzed emotional state, and means for generating a response plan and dialogue procedures to psychologically unsettle the fraudster. This makes it possible to detect fraud during a call in real time and to quickly provide the user with appropriate responses and safety information.

[0065] "Voice information" refers to the waveform data of sounds contained in phone calls or voice input, and by using this, it becomes possible to analyze the voice spoken by the speaker.

[0066] "Emotional state" refers to the speaker's internal psychological state as analyzed from auditory information, and is categorized into states such as reassurance, confusion, and anxiety.

[0067] "Suspicion of fraud" refers to a situation that may indicate fraudulent activity, as judged from the content of the call and the speaker's emotional state.

[0068] A "response plan" refers to a series of responses and actions taken to prevent fraud when a suspected fraud is detected.

[0069] A "dialogue procedure" refers to a consistent flow of conversation that consists of questions and answers and enticing questions designed to psychologically manipulate the con artist.

[0070] "Speech synthesis" refers to a technology that converts text-based dialogue procedures into speech, creating digital audio data that can be spoken through a speaker.

[0071] "Natural language processing" refers to artificial intelligence technology that enables computers to understand, analyze, and generate natural language used by humans.

[0072] A "report" refers to a document provided to the user that contains a detailed history of the call and the grounds for determining that it was a scam.

[0073] The system of this invention is configured to analyze calls received by users, detect potential fraud in real time, and ensure appropriate responses. It is implemented using three main components: a server, a terminal, and the user.

[0074] Server Embodiment

[0075] The server plays a central role in receiving and analyzing audio information during a call. Equipped with an advanced speech analysis engine (e.g., speech recognition technology), the server estimates the speaker's emotional state from the audio. This information is further converted into text data by a natural language processing engine (e.g., natural language processing technology) and used to detect typical patterns and phrases of fraud. If fraud is suspected, a generative AI model is used to generate dialogue sequences designed to psychologically manipulate the fraudster.

[0076] Terminal embodiment

[0077] The device is responsible for transmitting the audio data of calls received by the user to the server. It streams the audio in real time, enabling secure and rapid communication. Based on the response plan sent from the server, it also uses speech synthesis technology to create voice commands for the conversation and transmit them to the fraudster. The device also notifies the user of the situation and provides safety advice via voice or on-screen display.

[0078] User Embodiment

[0079] Users are subject to the protection of this system and should be provided with a safe calling experience. They understand that calls are monitored and appropriate advice will be provided when necessary. Furthermore, after the call ends, they can receive a report sent from the server via their device, allowing them to review call details and fraud detection results.

[0080] Specific example

[0081] For example, consider a scenario where user B receives a suspicious phone call. The server detects user B's anxiety through voice analysis and identifies characteristic phrases of fraud from the call content. Next, a generative AI model is used to generate a dialogue procedure incorporating effective questions for the scammer, which the terminal then synthesizes and speaks aloud. This prevents malicious information acquisition for user B, and after the call, user B can review the report to ensure their safety.

[0082] An example of a prompt for a generative AI model might be: "Analyze the call received by the user in real time and detect potential fraud. If fraud is suspected, create a dialogue script that suggests specific countermeasures."

[0083] This allows the system to effectively protect users from fraud and provide a safe and secure calling environment.

[0084] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0085] Step 1:

[0086] The device captures the audio of calls received by the user and streams that audio information to the server in real time. The input is the audio data of the call, and the output is the live streaming data to the server. Efficient audio compression technology is used in this process to minimize latency.

[0087] Step 2:

[0088] The server inputs the received audio information into a speech analysis engine to estimate the speaker's emotional state. The input is the audio information, and the output is the analysis result of the speaker's emotions. A specific algorithm is used to analyze the tone and pitch changes of the voice and identify emotions such as reassurance or anxiety.

[0089] Step 3:

[0090] The server inputs audio information into a natural language processing engine, converts it into text data, and detects typical phrases and patterns of fraud. The input consists of audio information and sentiment analysis results, while the output is text data and analysis results indicating suspected fraud. This process analyzes the content extracted from the audio through a language model.

[0091] Step 4:

[0092] If fraud is suspected, the server inputs prompt text into a generating AI model, which then generates a dialogue procedure for the fraudster. The input consists of sentiment analysis results and text analysis results, and the output is the dialogue procedure. Here, the AI ​​model designs a conversation flow that includes questions designed to exploit the fraudster's psychology.

[0093] Step 5:

[0094] The terminal receives dialogue instructions from the server as input and converts them into speech using a speech synthesis engine. The input is the dialogue instructions, and the output is the synthesized speech. In this step, the generated script is output in a voice that resembles the user's voice to create a natural response.

[0095] Step 6:

[0096] The terminal transmits synthesized voice to the caller and notifies the user of the situation. The input is synthesized voice, and the output is actual voice communication and screen display. The terminal also provides warnings and additional instructions to the user.

[0097] Step 7:

[0098] After the call ends, the server generates a report based on the call records and provides it to the user via the terminal. The input is the call data and analysis results, and the output is a detailed report. This report includes the grounds for suspected fraud and a summary of the call.

[0099] This process allows the system to detect suspected fraud in real time and provide advanced security support to users.

[0100] (Application Example 1)

[0101] 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."

[0102] Traditional communication methods make it difficult to detect phone scams and fraudulent activities, potentially threatening user safety and privacy. Furthermore, users lack effective means to immediately defend themselves against potential fraud. Therefore, there is a need to provide safe and secure communication by analyzing call content in real time and notifying users of potential fraud.

[0103] 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.

[0104] In this invention, the server includes means for acquiring voice information and analyzing the speaker's emotions in real time; means for identifying the possibility of fraudulent activity when the analyzed emotions indicate confusion or anxiety; and means for analyzing the content of the call and detecting phrases indicating fraud or changes in the user's emotions, thereby presenting a safety alert to the user in real time during the call. This enables immediate detection of fraudulent activity during a call and notification to the user.

[0105] "Voice information" refers to voice data obtained from phone calls, recordings, and other sources.

[0106] "Speaker" refers to the person providing the audio information.

[0107] "Emotional analysis" is a technology that analyzes audio information to evaluate the speaker's psychological state.

[0108] "Fraudulent activity" refers to communication conducted for fraudulent or illicit purposes.

[0109] "Real-time" refers to processing and analyzing events at the very moment they occur.

[0110] "Identification" is the act of detecting, classifying, or recognizing specific features or patterns.

[0111] "Analysis" is the act of thoroughly examining information and extracting meaning and patterns.

[0112] A "phrase" refers to a combination of words that have a specific meaning.

[0113] "User" refers to an individual or group that utilizes the system.

[0114] An "alert" refers to a notification that draws attention or warning.

[0115] The server acquires voice information in real time from smartphones and other digital devices. The voice analysis engine uses the Google® Speech-to-Text API to convert the voice information into text. Next, this text data is used with natural language processing techniques to identify phrases that indicate the speaker's emotions and specific fraudulent activities. For example, libraries such as spaCy and NLTK can be used to identify potentially fraudulent phrases.

[0116] The device uses a speech synthesis engine (e.g., Amazon Polly) to convert information obtained based on sentiment analysis into speech and provides real-time alerts to the user. The device displays the speech data and analysis results on the screen to help the user understand the current situation. This notification function allows the user to receive immediate warnings of fraudulent activity and take appropriate action.

[0117] After the call ends, users can refer to a detailed report provided by the server to verify the basis for any alleged misconduct during the call. This report can serve as a reference for users to avoid similar risks in the future.

[0118] As a concrete example, consider a scenario where a user is at home when a fraudulent person posing as a delivery service contacts them about a lost package. The server immediately detects phrases indicating fraudulent activity, such as "You need to verify the delivery company," and uses sentiment analysis to identify the user's confusion. In response, the device issues a warning to the user, prompting them to take precautions. This allows the user to understand the risk of fraud and protect their personal information.

[0119] An example of a prompt message might be, "What is the best course of action if I receive a fraudulent phone call?" Based on this message, the generative AI model generates appropriate countermeasures.

[0120] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0121] Step 1:

[0122] The server acquires audio information in real time from smartphones and digital devices. It sends the audio data as input to the Google Speech-to-Text API, which generates text data as output. This text data forms the basis for subsequent analysis processes.

[0123] Step 2:

[0124] The server analyzes the acquired text data using natural language processing libraries (e.g., spaCy or NLTK). It performs specific pattern recognition and sentiment evaluation on the input text data, and identifies relevant fraudulent phrases and emotional states as output. Specifically, it classifies phrases that suggest fraud and the emotions of the user, and evaluates the likelihood of fraud.

[0125] Step 3:

[0126] The server assesses the likelihood of fraud based on sentiment analysis and detection of fraudulent phrases, and generates appropriate dialogue instructions. Using the analysis results as input, it converts the dialogue script for a speech synthesis engine (e.g., Amazon Polly) and prepares an alert to send to the user as output.

[0127] Step 4:

[0128] The terminal converts the dialogue script received from the server into speech and issues warnings to the user in real time. It takes the dialogue script as input and converts it into speech, while outputting notifications and warnings to the user. Specifically, it emits warnings from the terminal's speaker.

[0129] Step 5:

[0130] After the call ends, the server generates a detailed report based on the call history and analysis results, and provides it to the user. The report is created by combining the call content (input) and analysis results (output). This allows the user to gain a deeper understanding of the call details and the possibility of fraud.

[0131] 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.

[0132] This invention is a fraud prevention system that combines an emotion engine, enabling more effective fraud response by analyzing user emotions in real time. Specifically, it provides a mechanism in which a server and a terminal work together to detect suspected fraud early and take appropriate countermeasures.

[0133] server

[0134] The server is equipped with a speech analysis engine and an emotion engine for analyzing audio data received from terminals. The speech analysis engine is responsible for the basic processing of estimating the speaker's emotional state and determining the likelihood of fraud based on this. The emotion engine further recognizes emotions from the user's voice during a call, and if the user is feeling confused or anxious, it immediately assesses the risk and dynamically adjusts the priority of fraud prevention measures. This emotion data is also accumulated and used for long-term pattern analysis.

[0135] terminal

[0136] The device streams voice data to a server, and the server generates a response which is then synthesized and spoken aloud. Based on notifications from the emotion engine, safety advice is provided to the user via on-screen or audio notifications. This helps users anticipate fraud risks and respond appropriately.

[0137] User

[0138] Users are protected from fraud through this system. During a call, the system performs background analysis and takes action as needed. After the call ends, users receive a report generated by the server, allowing them to review details including their emotional state and the basis for the fraud assessment. This provides users with feedback that can help them take future preventative measures.

[0139] Specific example

[0140] As a concrete example, when user C receives an unexpected sales call, the emotion engine detects subtle confusion and anxiety in the voice. Based on this, the server determines the call may be fraudulent and immediately generates an advanced dialogue strategy to unsettle the scammer. The terminal then uses this script to communicate with the scammer. User C is encouraged to end the call before their anxiety escalates, and simultaneously gains reassurance through a post-call report. In this way, the system achieves flexible fraud response that reflects the user's emotions.

[0141] The following describes the processing flow.

[0142] Step 1:

[0143] The device initiates a call and streams audio data to the server in real time. The emotion engine also starts simultaneously and begins collecting audio data.

[0144] Step 2:

[0145] The server uses a speech analysis engine to analyze the tone and speed of the speaker's voice from the audio data and estimate the speaker's emotional state. The user's voice is specifically analyzed by the emotion engine, and its emotional state is evaluated in real time.

[0146] Step 3:

[0147] The server applies natural language processing technology to convert the audio data into text and detects typical phrases and keywords used in scams.

[0148] Step 4:

[0149] The server integrates sentiment analysis and phrase detection results to determine if there is a suspicion of fraud. If the user shows signs of confusion or anxiety, fraud prevention measures are prioritized.

[0150] Step 5:

[0151] If fraud is suspected, the server generates a dialogue script designed to psychologically unsettle the fraudster. If necessary, the script's content is adjusted based on the user's emotions.

[0152] Step 6:

[0153] The terminal receives a dialogue script from the server, converts it using speech synthesis technology, and speaks it to the scammer.

[0154] Step 7:

[0155] The device notifies the user of the call status via voice or on-screen display, along with safety advice.

[0156] Step 8:

[0157] After the call ends, the server generates a report containing the call history and the basis for the fraud detection, and provides it to the user via their device. This allows the user to understand the overall picture of the call, including their emotional state.

[0158] (Example 2)

[0159] 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".

[0160] In recent years, fraudulent activities have diversified, with telephone scams being particularly on the rise. Traditional fraud prevention methods are often limited to detecting simple phrases and issuing delayed warnings, and real-time sentiment analysis and dynamic responses are frequently difficult to implement. As a result, the risk of users becoming victims of fraud is increasing. There is a need for a system that can prevent this and protect users more safely.

[0161] 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.

[0162] In this invention, the server includes means for acquiring voice data and transmitting it to a data processing device via a communication device; means for analyzing the received voice data in detail and estimating the speaker's emotions; and means for evaluating the likelihood of a high level of fraud if the estimated emotions indicate confusion or anxiety. This enables users to reduce the risk of fraud in real time and take quick and appropriate action.

[0163] "Voice data" refers to digital information containing the user's spoken content, which is processed via a communication device.

[0164] "Communication equipment" is a general term for hardware and software used to acquire voice data and transfer it to another device or server.

[0165] A "data processing device" is a computer system that analyzes received audio data and processes the information.

[0166] "Means for estimating emotions" refers to algorithms and analysis engines for identifying a speaker's emotional state from audio data.

[0167] "Means of assessing suspicion of fraud" refers to a process of determining the likelihood of fraud based on sentiment estimation results, in light of numerical values ​​and evaluation criteria.

[0168] A "generative AI model" is an artificial intelligence technology that generates new outputs using past data and learning algorithms.

[0169] A "dialogue script" is a script that outlines the expected responses to a fraudster, and is played back using speech synthesis.

[0170] "Speech synthesis" is a technology that converts text information into speech information and is used to produce virtual voice responses.

[0171] This fraud prevention system consists of three components: a server, a terminal, and a user. Specific embodiments of each component are shown below.

[0172] The server features a voice analysis engine and an emotion engine, which analyze voice data transmitted from the terminal in real time. The voice analysis engine uses a high-performance processor and database server to analyze the speaker's voice characteristics and estimate their emotional state. This data is further analyzed by the emotion engine, and if negative emotions such as confusion or anxiety are detected, the risk of fraud is immediately assessed. The emotion data is accumulated and used for long-term trend analysis.

[0173] The device is responsible for acquiring the user's voice data and streaming it to the server in real time. The device is equipped with a microphone and speaker and runs an audio streaming application. Upon receiving a response strategy generated by the server, it uses speech synthesis technology to generate and play a response to the scammer. The device also informs the user of the risk of fraud and provides safety advice through screen displays and audio notifications.

[0174] This system protects users from the risk of fraud. The system performs sentiment analysis in the background and takes prompt action if fraud is suspected. After the call ends, users can review the report provided by the server and understand the detected sentiment and the reasons for the fraud assessment, which can help them take future preventative measures.

[0175] As a concrete example, consider a scenario where a user receives an unexpected sales call. In this case, if the emotion engine detects slight confusion and anxiety, the server determines that the sales call may be a scam. Based on this, the server generates sophisticated dialogue strategies to psychologically unsettle the scammer. The terminal synthesizes these strategies into speech and plays them back to the scammer, helping the user end the call before feeling anxious. In this example, the prompt message "Evaluate the likelihood of fraud based on the anxiety level detected from the user's voice and generate sophisticated dialogue strategies" is input into the generating AI model, providing the optimal solution.

[0176] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0177] Step 1:

[0178] The device acquires the user's voice data. A microphone is used to capture the user's voice during a call in digital format. The input is the user's raw voice, and this data is temporarily stored within the device. The output is digitized voice data.

[0179] Step 2:

[0180] The terminal streams the acquired audio data to the server. Wi-Fi or mobile data communication is used for this purpose. The input is digital audio data stored within the terminal, and the output is the process of that data being wirelessly transferred to the server. An efficient data transfer protocol is applied to ensure real-time performance during this process.

[0181] Step 3:

[0182] The server analyzes the received audio data using a speech analysis engine. The analysis extracts characteristics such as intonation, speed, and volume. The input is the digital audio data sent to the server, and the output is a dataset of these characteristics quantified. This dataset is then used for further emotion estimation.

[0183] Step 4:

[0184] The server uses an emotion engine to estimate emotions based on the results of speech analysis. Specifically, it estimates the speaker's emotional state (e.g., confusion, anxiety) using the emotion engine's model based on extracted features. The input is feature data generated by the speech analysis engine, and the output is the estimated emotion label.

[0185] Step 5:

[0186] The server assesses the likelihood of fraud based on estimated sentiment. Here, a generative AI model is used to compare sentiment data with past fraud patterns. The input is the sentiment label estimated by the sentiment engine, and the output is a fraud risk assessment index. If the assessment index is high, the process proceeds to the next step.

[0187] Step 6:

[0188] The server generates a response strategy for the fraudster when it assesses the likelihood of fraud. Specifically, it uses a generative AI model to generate a dialogue script from the prompt text to psychologically influence the fraudster. The input is an assessment of fraud risk, and the output is the generated dialogue script.

[0189] Step 7:

[0190] The terminal plays back the dialogue script received from the server using speech synthesis technology. The audio is played through the speaker. The input is the voice script sent from the server, and the output is a voice message directed at the fraudster. This process employs a psychological approach to the fraudster.

[0191] Step 8:

[0192] The device notifies the user of potential fraud and suggests safe countermeasures. Notifications are delivered via screen displays and additional audio messages. Input is the server's evaluation result, and output is a warning and suggested action to the user. This allows the user to make informed decisions and mitigate the risk of fraud.

[0193] (Application Example 2)

[0194] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0195] With the widespread adoption of electronic payments, the risk of users becoming victims of fraud is increasing. Therefore, there is a need for a system that analyzes user emotions in real time, quickly detects potential fraud, and provides appropriate responses. Furthermore, a mechanism to verify the security of transactions and immediately terminate risky transactions is also necessary.

[0196] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0197] In this invention, the server includes means for receiving voice information and immediately analyzing the speaker's emotions, means for detecting the possibility of fraud if the analyzed emotions express confusion or anxiety, and means for creating countermeasures and providing appropriate responses to the responder when the possibility of fraud is detected. This makes it possible for users to detect the risk of fraud more quickly and for a safer trading environment to be provided.

[0198] "Voice information" refers to digital data used to collect and analyze the content of users' phone calls in real time.

[0199] A "device that instantly analyzes the speaker's emotions" is a technology that quickly estimates the user's emotional state from received audio information.

[0200] A "device for detecting potential fraud" is a function that identifies potential fraud when suspicious signs are found in the analyzed emotions.

[0201] A "device for generating response measures" is a technology for generating specific actions and intervention methods to be taken when fraud is suspected.

[0202] A "device that provides appropriate responses to those involved" is a function that provides appropriate support and guidance to users and other stakeholders based on the policies that have been created.

[0203] "Payment communication" refers to the exchange of data that occurs during the process of conducting financial transactions and payments using electronic means.

[0204] A "device for verifying the security of transactions" is a technology that detects signs of fraudulent activity during settlement communications and performs a risk assessment in response.

[0205] A "mechanism to immediately terminate risky trades" is a process that automatically stops a trade when it is deemed to be high-risk, thereby preventing potential losses.

[0206] This invention relates to a system that uses voice analysis and sentiment analysis to reduce the risk of fraud during electronic payments. This system consists of a smartphone terminal and server software running in the background.

[0207] The server receives audio data and analyzes the audio information in real time. A speech analysis engine, such as Google Cloud Speech-to-Text, is used for the audio analysis. From the analyzed audio information, an emotion analysis engine instantly estimates the speaker's emotions and detects potential fraud based on the results. If fraud is detected, the server creates countermeasures and generates an appropriate response based on those measures.

[0208] The terminal plays this response as an audio message to notify the user. Furthermore, if a risky transaction occurs, the terminal has a function to automatically block the transaction and notify the user immediately to terminate it.

[0209] For example, if anxiety or confusion is detected from voice information when a user attempts to make an electronic payment, the server will immediately issue a warning, and the terminal will provide advice such as, "This transaction may be risky." Another example of a prompt message is, "I feel uneasy about this new online payment method, and I would like to confirm if this transaction is safe. Please let me know if the system has detected anything unusual." This prompt message forms the basis for the generative AI model to provide appropriate support.

[0210] In this way, the server and terminal work together to provide users with consistent support for fraud prevention.

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

[0212] Step 1:

[0213] The server receives audio data from the terminal. This audio data is a recording of the user's phone call and is transmitted in real time. Audio information is provided as input, and basic data for audio analysis is prepared as output.

[0214] Step 2:

[0215] The server converts the received audio data into text using Google Cloud Speech-to-Text. This process transforms the audio information into text data. The input is audio data, and the output is text data.

[0216] Step 3:

[0217] The server passes text data to an emotion analysis engine to estimate the user's emotions. The emotion analysis engine analyzes the text data to identify the user's emotional state (e.g., confusion, anxiety). The input is text data, and the output is emotional state data.

[0218] Step 4:

[0219] The server assesses the likelihood of fraud based on emotional state data. This assessment is performed by an algorithm that generates a fraud warning if confusion or anxiety exceeds a certain threshold. The input is emotional state data, and the output is fraud risk assessment data.

[0220] Step 5:

[0221] If fraud is detected, the server generates countermeasures. These measures include warning messages to the user and orders to suspend transactions. The input is data indicating fraud risk, and the output is data indicating the countermeasures.

[0222] Step 6:

[0223] The terminal receives policy data provided by the server and issues an audio warning to the user. Specifically, it converts text-based warning messages into speech using a speech synthesis system and plays them through the terminal's speaker. The input is policy data, and the output is an audio warning.

[0224] Step 7:

[0225] If a transaction is determined to be risky, the terminal automatically blocks the transaction and notifies the user of the result. This reduces the risk of the user engaging in fraudulent transactions. The input is the fraud risk assessment data, and the output is the transaction blocking command.

[0226] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0227] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0228] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0229] [Second Embodiment]

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

[0231] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0232] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0233] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0234] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0235] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0236] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0237] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0238] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0239] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0240] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0241] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0242] The system of this invention is designed to analyze incoming calls by users in real time and to take appropriate action based on their content. The server, terminal, and user components each perform their respective functions as follows.

[0243] server

[0244] The server plays a central role in receiving and analyzing audio data during calls. Specifically, the server uses a voice analysis engine to perform sentiment analysis, estimating basic emotional states such as reassurance, confusion, and anxiety from the speaker's voice. To detect typical scam phrases, natural language processing techniques are applied, and the call content is analyzed in real time as text data. If fraud is suspected, the server immediately generates a response strategy and creates a dialogue script to psychologically unsettle the scammer.

[0245] terminal

[0246] The device transmits the audio received by the user to the server and streams the data. It then synthesizes the response instructed by the server and transmits it to the scammer as actual audio. The device also notifies the user of the situation in real time and provides safety advice via voice or screen display.

[0247] User

[0248] Users are beneficiaries of this system, which protects them from fraudulent calls. Users understand that their calls are monitored and that advice will be provided as needed, ensuring a safe calling experience. After the call ends, users receive a report generated by the server via their device, allowing them to review call details and a determination based on the suspicion of fraud.

[0249] Specific example

[0250] As a concrete example, consider a scenario where User B receives a message from a scammer. The server detects phrases suggesting potential fraud through voice communication and senses User B's anxiety through sentiment analysis. The server then generates a question-and-answer dialogue designed to expose inconsistencies in the scammer's statements, and the terminal speaks this script. User B receives instructions from the terminal to perform multiple verifications, preventing the scammer from obtaining information, and after the call ends, can review a detailed report to understand the situation. In this way, the system enables immediate prevention of fraud.

[0251] The following describes the processing flow.

[0252] Step 1:

[0253] As soon as the device initiates a call, it streams the audio data to the server in real time.

[0254] Step 2:

[0255] The server processes the received audio data using a speech analysis engine, analyzing the speaker's tone, pitch, speed, etc., to estimate their emotional state.

[0256] Step 3:

[0257] The server uses a natural language processing module to convert audio data into text and detects typical scam phrases and keywords.

[0258] Step 4:

[0259] The server evaluates the sentiment analysis results and phrase detection results to determine whether there is a suspicion of fraud.

[0260] Step 5:

[0261] If fraud is suspected, the server generates a response strategy and creates a dialogue script to psychologically unsettle the fraudster.

[0262] Step 6:

[0263] The terminal receives a dialogue script from the server, synthesizes it into speech, and speaks it to the scammer.

[0264] Step 7:

[0265] The device notifies the user of the call status and provides safety advice via voice or on-screen display.

[0266] Step 8:

[0267] After the call ends, the server generates a report containing the call history and the basis for the fraud determination, and provides it to the user via the terminal.

[0268] (Example 1)

[0269] 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."

[0270] In modern society, telephone fraud is on the rise, and individuals need immediate and effective measures to protect themselves from these scams. However, conventional systems have difficulty detecting fraud during a call and providing appropriate responses in real time. Therefore, there is a need for a system that can detect fraud early during a call and provide users with appropriate responses and assistance information in real time.

[0271] 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.

[0272] In this invention, the server includes means for receiving voice information and analyzing the speaker's emotional state, means for detecting suspicion of fraud based on the analyzed emotional state, and means for generating a response plan and dialogue procedures to psychologically unsettle the fraudster. This makes it possible to detect fraud during a call in real time and to quickly provide the user with appropriate responses and safety information.

[0273] "Voice information" refers to the waveform data of sounds contained in phone calls or voice input, and by using this, it becomes possible to analyze the voice spoken by the speaker.

[0274] "Emotional state" refers to the speaker's internal psychological state as analyzed from auditory information, and is categorized into states such as reassurance, confusion, and anxiety.

[0275] "Suspicion of fraud" refers to a situation that may indicate fraudulent activity, as judged from the content of the call and the speaker's emotional state.

[0276] A "response plan" refers to a series of responses and actions taken to prevent fraud when a suspected fraud is detected.

[0277] A "dialogue procedure" refers to a consistent flow of conversation that consists of questions and answers and enticing questions designed to psychologically manipulate the con artist.

[0278] "Speech synthesis" refers to a technology that converts text-based dialogue procedures into speech, creating digital audio data that can be spoken through a speaker.

[0279] "Natural language processing" refers to artificial intelligence technology that enables computers to understand, analyze, and generate natural language used by humans.

[0280] A "report" refers to a document provided to the user that contains a detailed history of the call and the grounds for determining that it was a scam.

[0281] The system of this invention is configured to analyze calls received by users, detect potential fraud in real time, and ensure appropriate responses. It is implemented using three main components: a server, a terminal, and the user.

[0282] Server Embodiment

[0283] The server plays a central role in receiving the voice information during a call and analyzing the data. The server is equipped with an advanced voice analysis engine (such as voice recognition technology) to estimate the emotional state of the speaker from the voice. This information is further converted into text data by a natural language processing engine (such as natural language analysis technology) and used to detect typical patterns and phrases of fraud. If there is suspicion of fraud, a generative AI model is utilized to generate a dialogue procedure for psychologically disturbing the fraudster.

[0284] Embodiment of the terminal

[0285] The terminal is responsible for transmitting the voice data of the call received by the user to the server. It performs real-time streaming to enable safe and rapid communication. Also, based on the response plan transmitted from the server, it vocalizes the dialogue procedure using voice synthesis technology and transmits it to the fraudster. The terminal also notifies the user of the situation and provides advice on safety via voice or screen display.

[0286] Embodiment of the user

[0287] The user is the object to be protected by this system and should be provided with a safe call experience. The user understands that the call is monitored and appropriate advice is provided when necessary. Furthermore, after the call ends, the user can receive the report transmitted from the server via the terminal and check the details of the call and the fraud determination result.

[0288] Specific example

[0289] For example, consider the case where user B receives a suspicious call. The server senses user B's uneasiness through voice analysis and detects characteristic phrases of fraud from the call content. Next, a generative AI model is used to generate a dialogue procedure incorporating effective questions for the fraudster, and the terminal synthesizes and vocalizes it. As a result, user B prevents malicious information acquisition and checks the report after the call to ensure safety.

[0290] An example of a prompt for a generative AI model might be: "Analyze the call received by the user in real time and detect potential fraud. If fraud is suspected, create a dialogue script that suggests specific countermeasures."

[0291] This allows the system to effectively protect users from fraud and provide a safe and secure calling environment.

[0292] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0293] Step 1:

[0294] The device captures the audio of calls received by the user and streams that audio information to the server in real time. The input is the audio data of the call, and the output is the live streaming data to the server. Efficient audio compression technology is used in this process to minimize latency.

[0295] Step 2:

[0296] The server inputs the received audio information into a speech analysis engine to estimate the speaker's emotional state. The input is the audio information, and the output is the analysis result of the speaker's emotions. A specific algorithm is used to analyze the tone and pitch changes of the voice and identify emotions such as reassurance or anxiety.

[0297] Step 3:

[0298] The server inputs audio information into a natural language processing engine, converts it into text data, and detects typical phrases and patterns of fraud. The input consists of audio information and sentiment analysis results, while the output is text data and analysis results indicating suspected fraud. This process analyzes the content extracted from the audio through a language model.

[0299] Step 4:

[0300] When there is suspicion of fraud, the server inputs a prompt text into the generative AI model to generate the dialogue procedure for dealing with the fraudster. The input is the sentiment analysis result and the text analysis result, and the output is the dialogue procedure. Here, the AI model designs the flow of the conversation including questions that pierce through the psychology of the fraudster.

[0301] Step 5:

[0302] The terminal inputs the dialogue procedure received from the server and converts it into voice using the speech synthesis engine. The input is the dialogue procedure, and the output is the synthesized voice. In this step, a natural response is created by outputting the generated script in a voice similar to the user's voice.

[0303] Step 6:

[0304] The terminal dials the synthesized voice to the call partner and notifies the user of the situation. The input is the synthesized voice, and the output is the actual voice communication and screen display. Here, the terminal also provides warnings and additional instructions to the user.

[0305] Step 7:

[0306] After the call ends, the server generates a report based on the call record and provides it to the user through the terminal. The input is the call data and the analysis result, and the output is a detailed report. This report includes the basis for suspicion of fraud and a summary of the call.

[0307] Through this process, the system can detect suspicion of fraud in real time and provide high-level security support to the user.

[0308] (Application Example 1)

[0309] Next, Application Example 1 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart glasses 214 are referred to as the "terminal".

[0310] Traditional communication methods make it difficult to detect phone scams and fraudulent activities, potentially threatening user safety and privacy. Furthermore, users lack effective means to immediately defend themselves against potential fraud. Therefore, there is a need to provide safe and secure communication by analyzing call content in real time and notifying users of potential fraud.

[0311] 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.

[0312] In this invention, the server includes means for acquiring voice information and analyzing the speaker's emotions in real time; means for identifying the possibility of fraudulent activity when the analyzed emotions indicate confusion or anxiety; and means for analyzing the content of the call and detecting phrases indicating fraud or changes in the user's emotions, thereby presenting a safety alert to the user in real time during the call. This enables immediate detection of fraudulent activity during a call and notification to the user.

[0313] "Voice information" refers to voice data obtained from phone calls, recordings, and other sources.

[0314] "Speaker" refers to the person providing the audio information.

[0315] "Emotional analysis" is a technology that analyzes audio information to evaluate the speaker's psychological state.

[0316] "Fraudulent activity" refers to communication conducted for fraudulent or illicit purposes.

[0317] "Real-time" refers to processing and analyzing events at the very moment they occur.

[0318] "Identification" is the act of detecting, classifying, or recognizing specific features or patterns.

[0319] "Analysis" is the act of thoroughly examining information and extracting meaning and patterns.

[0320] A "phrase" refers to a combination of words that have a specific meaning.

[0321] "User" refers to an individual or group that utilizes the system.

[0322] An "alert" refers to a notification that draws attention or warning.

[0323] The server acquires audio information in real time from smartphones and other digital devices. The Google Speech-to-Text API is used as the speech analysis engine to convert the audio information into text. Next, natural language processing techniques are used with this text data to identify phrases that indicate the speaker's emotions and specific fraudulent activities. For example, libraries such as spaCy and NLTK can be used to identify potentially fraudulent phrases.

[0324] The device uses a speech synthesis engine (e.g., Amazon Polly) to convert information obtained based on sentiment analysis into speech and provides real-time alerts to the user. The device displays the speech data and analysis results on the screen to help the user understand the current situation. This notification function allows the user to receive immediate warnings of fraudulent activity and take appropriate action.

[0325] After the call ends, users can refer to a detailed report provided by the server to verify the basis for any alleged misconduct during the call. This report can serve as a reference for users to avoid similar risks in the future.

[0326] As a concrete example, consider a scenario where a user is at home when a fraudulent person posing as a delivery service contacts them about a lost package. The server immediately detects phrases indicating fraudulent activity, such as "You need to verify the delivery company," and uses sentiment analysis to identify the user's confusion. In response, the device issues a warning to the user, prompting them to take precautions. This allows the user to understand the risk of fraud and protect their personal information.

[0327] An example of a prompt message might be, "What is the best course of action if I receive a fraudulent phone call?" Based on this message, the generative AI model generates appropriate countermeasures.

[0328] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0329] Step 1:

[0330] The server acquires audio information in real time from smartphones and digital devices. It sends the audio data as input to the Google Speech-to-Text API, which generates text data as output. This text data forms the basis for subsequent analysis processes.

[0331] Step 2:

[0332] The server analyzes the acquired text data using natural language processing libraries (e.g., spaCy or NLTK). It performs specific pattern recognition and sentiment evaluation on the input text data, and identifies relevant fraudulent phrases and emotional states as output. Specifically, it classifies phrases that suggest fraud and the emotions of the user, and evaluates the likelihood of fraud.

[0333] Step 3:

[0334] The server assesses the likelihood of fraud based on sentiment analysis and detection of fraudulent phrases, and generates appropriate dialogue instructions. Using the analysis results as input, it converts the dialogue script for a speech synthesis engine (e.g., Amazon Polly) and prepares an alert to send to the user as output.

[0335] Step 4:

[0336] The terminal converts the dialogue script received from the server into speech and issues warnings to the user in real time. It takes the dialogue script as input and converts it into speech, while outputting notifications and warnings to the user. Specifically, it emits warnings from the terminal's speaker.

[0337] Step 5:

[0338] After the call ends, the server generates a detailed report based on the call history and analysis results, and provides it to the user. The report is created by combining the call content (input) and analysis results (output). This allows the user to gain a deeper understanding of the call details and the possibility of fraud.

[0339] 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.

[0340] This invention is a fraud prevention system that combines an emotion engine, enabling more effective fraud response by analyzing user emotions in real time. Specifically, it provides a mechanism in which a server and a terminal work together to detect suspected fraud early and take appropriate countermeasures.

[0341] server

[0342] The server is equipped with a speech analysis engine and an emotion engine for analyzing audio data received from terminals. The speech analysis engine is responsible for the basic processing of estimating the speaker's emotional state and determining the likelihood of fraud based on this. The emotion engine further recognizes emotions from the user's voice during a call, and if the user is feeling confused or anxious, it immediately assesses the risk and dynamically adjusts the priority of fraud prevention measures. This emotion data is also accumulated and used for long-term pattern analysis.

[0343] terminal

[0344] The device streams voice data to a server, and the server generates a response which is then synthesized and spoken aloud. Based on notifications from the emotion engine, safety advice is provided to the user via on-screen or audio notifications. This helps users anticipate fraud risks and respond appropriately.

[0345] User

[0346] Users are protected from fraud through this system. During a call, the system performs background analysis and takes action as needed. After the call ends, users receive a report generated by the server, allowing them to review details including their emotional state and the basis for the fraud assessment. This provides users with feedback that can help them take future preventative measures.

[0347] Specific example

[0348] As a concrete example, when user C receives an unexpected sales call, the emotion engine detects subtle confusion and anxiety in the voice. Based on this, the server determines the call may be fraudulent and immediately generates an advanced dialogue strategy to unsettle the scammer. The terminal then uses this script to communicate with the scammer. User C is encouraged to end the call before their anxiety escalates, and simultaneously gains reassurance through a post-call report. In this way, the system achieves flexible fraud response that reflects the user's emotions.

[0349] The following describes the processing flow.

[0350] Step 1:

[0351] The device initiates a call and streams audio data to the server in real time. The emotion engine also starts simultaneously and begins collecting audio data.

[0352] Step 2:

[0353] The server uses a speech analysis engine to analyze the tone and speed of the speaker's voice from the audio data and estimate the speaker's emotional state. The user's voice is specifically analyzed by the emotion engine, and its emotional state is evaluated in real time.

[0354] Step 3:

[0355] The server applies natural language processing technology to convert the audio data into text and detects typical phrases and keywords used in scams.

[0356] Step 4:

[0357] The server integrates sentiment analysis and phrase detection results to determine if there is a suspicion of fraud. If the user shows signs of confusion or anxiety, fraud prevention measures are prioritized.

[0358] Step 5:

[0359] If fraud is suspected, the server generates a dialogue script designed to psychologically unsettle the fraudster. If necessary, the script's content is adjusted based on the user's emotions.

[0360] Step 6:

[0361] The terminal receives a dialogue script from the server, converts it using speech synthesis technology, and speaks it to the scammer.

[0362] Step 7:

[0363] The device notifies the user of the call status via voice or on-screen display, along with safety advice.

[0364] Step 8:

[0365] After the call ends, the server generates a report containing the call history and the basis for the fraud detection, and provides it to the user via their device. This allows the user to understand the overall picture of the call, including their emotional state.

[0366] (Example 2)

[0367] 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".

[0368] In recent years, fraudulent activities have diversified, with telephone scams being particularly on the rise. Traditional fraud prevention methods are often limited to detecting simple phrases and issuing delayed warnings, and real-time sentiment analysis and dynamic responses are frequently difficult to implement. As a result, the risk of users becoming victims of fraud is increasing. There is a need for a system that can prevent this and protect users more safely.

[0369] 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.

[0370] In this invention, the server includes means for acquiring voice data and transmitting it to a data processing device via a communication device; means for analyzing the received voice data in detail and estimating the speaker's emotions; and means for evaluating the likelihood of a high level of fraud if the estimated emotions indicate confusion or anxiety. This enables users to reduce the risk of fraud in real time and take quick and appropriate action.

[0371] "Voice data" refers to digital information containing the user's spoken content, which is processed via a communication device.

[0372] "Communication equipment" is a general term for hardware and software used to acquire voice data and transfer it to another device or server.

[0373] A "data processing device" is a computer system that analyzes received audio data and processes the information.

[0374] "Means for estimating emotions" refers to algorithms and analysis engines for identifying a speaker's emotional state from audio data.

[0375] "Means of assessing suspicion of fraud" refers to a process of determining the likelihood of fraud based on sentiment estimation results, in light of numerical values ​​and evaluation criteria.

[0376] A "generative AI model" is an artificial intelligence technology that generates new outputs using past data and learning algorithms.

[0377] A "dialogue script" is a script that outlines the expected responses to a fraudster, and is played back using speech synthesis.

[0378] "Speech synthesis" is a technology that converts text information into speech information and is used to produce virtual voice responses.

[0379] This fraud prevention system consists of three components: a server, a terminal, and a user. Specific embodiments of each component are shown below.

[0380] The server features a voice analysis engine and an emotion engine, which analyze voice data transmitted from the terminal in real time. The voice analysis engine uses a high-performance processor and database server to analyze the speaker's voice characteristics and estimate their emotional state. This data is further analyzed by the emotion engine, and if negative emotions such as confusion or anxiety are detected, the risk of fraud is immediately assessed. The emotion data is accumulated and used for long-term trend analysis.

[0381] The device is responsible for acquiring the user's voice data and streaming it to the server in real time. The device is equipped with a microphone and speaker and runs an audio streaming application. Upon receiving a response strategy generated by the server, it uses speech synthesis technology to generate and play a response to the scammer. The device also informs the user of the risk of fraud and provides safety advice through screen displays and audio notifications.

[0382] This system protects users from the risk of fraud. The system performs sentiment analysis in the background and takes prompt action if fraud is suspected. After the call ends, users can review the report provided by the server and understand the detected sentiment and the reasons for the fraud assessment, which can help them take future preventative measures.

[0383] As a concrete example, consider a scenario where a user receives an unexpected sales call. In this case, if the emotion engine detects slight confusion and anxiety, the server determines that the sales call may be a scam. Based on this, the server generates sophisticated dialogue strategies to psychologically unsettle the scammer. The terminal synthesizes these strategies into speech and plays them back to the scammer, helping the user end the call before feeling anxious. In this example, the prompt message "Evaluate the likelihood of fraud based on the anxiety level detected from the user's voice and generate sophisticated dialogue strategies" is input into the generating AI model, providing the optimal solution.

[0384] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0385] Step 1:

[0386] The device acquires the user's voice data. A microphone is used to capture the user's voice during a call in digital format. The input is the user's raw voice, and this data is temporarily stored within the device. The output is digitized voice data.

[0387] Step 2:

[0388] The terminal streams the acquired audio data to the server. Wi-Fi or mobile data communication is used for this purpose. The input is digital audio data stored within the terminal, and the output is the process of that data being wirelessly transferred to the server. An efficient data transfer protocol is applied to ensure real-time performance during this process.

[0389] Step 3:

[0390] The server analyzes the received audio data using a speech analysis engine. The analysis extracts characteristics such as intonation, speed, and volume. The input is the digital audio data sent to the server, and the output is a dataset of these characteristics quantified. This dataset is then used for further emotion estimation.

[0391] Step 4:

[0392] The server uses an emotion engine to estimate emotions based on the results of speech analysis. Specifically, it estimates the speaker's emotional state (e.g., confusion, anxiety) using the emotion engine's model based on extracted features. The input is feature data generated by the speech analysis engine, and the output is the estimated emotion label.

[0393] Step 5:

[0394] The server assesses the likelihood of fraud based on estimated sentiment. Here, a generative AI model is used to compare sentiment data with past fraud patterns. The input is the sentiment label estimated by the sentiment engine, and the output is a fraud risk assessment index. If the assessment index is high, the process proceeds to the next step.

[0395] Step 6:

[0396] The server generates a response strategy for the fraudster when it assesses the likelihood of fraud. Specifically, it uses a generative AI model to generate a dialogue script from the prompt text to psychologically influence the fraudster. The input is an assessment of fraud risk, and the output is the generated dialogue script.

[0397] Step 7:

[0398] The terminal plays back the dialogue script received from the server using speech synthesis technology. The audio is played through the speaker. The input is the voice script sent from the server, and the output is a voice message directed at the fraudster. This process employs a psychological approach to the fraudster.

[0399] Step 8:

[0400] The device notifies the user of potential fraud and suggests safe countermeasures. Notifications are delivered via screen displays and additional audio messages. Input is the server's evaluation result, and output is a warning and suggested action to the user. This allows the user to make informed decisions and mitigate the risk of fraud.

[0401] (Application Example 2)

[0402] 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."

[0403] With the widespread adoption of electronic payments, the risk of users becoming victims of fraud is increasing. Therefore, there is a need for a system that analyzes user emotions in real time, quickly detects potential fraud, and provides appropriate responses. Furthermore, a mechanism to verify the security of transactions and immediately terminate risky transactions is also necessary.

[0404] 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.

[0405] In this invention, the server includes means for receiving voice information and immediately analyzing the speaker's emotions, means for detecting the possibility of fraud if the analyzed emotions express confusion or anxiety, and means for creating countermeasures and providing appropriate responses to the responder when the possibility of fraud is detected. This makes it possible for users to detect the risk of fraud more quickly and for a safer trading environment to be provided.

[0406] "Voice information" refers to digital data used to collect and analyze the content of users' phone calls in real time.

[0407] A "device that instantly analyzes the speaker's emotions" is a technology that quickly estimates the user's emotional state from received audio information.

[0408] A "device for detecting potential fraud" is a function that identifies potential fraud when suspicious signs are found in the analyzed emotions.

[0409] A "device for generating response measures" is a technology for generating specific actions and intervention methods to be taken when fraud is suspected.

[0410] A "device that provides appropriate responses to those involved" is a function that provides appropriate support and guidance to users and other stakeholders based on the policies that have been created.

[0411] "Payment communication" refers to the exchange of data that occurs during the process of conducting financial transactions and payments using electronic means.

[0412] A "device for verifying the security of transactions" is a technology that detects signs of fraudulent activity during settlement communications and performs a risk assessment in response.

[0413] A "mechanism to immediately terminate risky trades" is a process that automatically stops a trade when it is deemed to be high-risk, thereby preventing potential losses.

[0414] This invention relates to a system that uses voice analysis and sentiment analysis to reduce the risk of fraud during electronic payments. This system consists of a smartphone terminal and server software running in the background.

[0415] The server receives audio data and analyzes the audio information in real time. A speech analysis engine, such as Google Cloud Speech-to-Text, is used for the audio analysis. From the analyzed audio information, an emotion analysis engine instantly estimates the speaker's emotions and detects potential fraud based on the results. If fraud is detected, the server creates countermeasures and generates an appropriate response based on those measures.

[0416] The terminal plays this response as an audio message to notify the user. Furthermore, if a risky transaction occurs, the terminal has a function to automatically block the transaction and notify the user immediately to terminate it.

[0417] For example, if anxiety or confusion is detected from voice information when a user attempts to make an electronic payment, the server will immediately issue a warning, and the terminal will provide advice such as, "This transaction may be risky." Another example of a prompt message is, "I feel uneasy about this new online payment method, and I would like to confirm if this transaction is safe. Please let me know if the system has detected anything unusual." This prompt message forms the basis for the generative AI model to provide appropriate support.

[0418] In this way, the server and terminal work together to provide users with consistent support for fraud prevention.

[0419] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0420] Step 1:

[0421] The server receives audio data from the terminal. This audio data is a recording of the user's phone call and is transmitted in real time. Audio information is provided as input, and basic data for audio analysis is prepared as output.

[0422] Step 2:

[0423] The server converts the received audio data into text using Google Cloud Speech-to-Text. This process transforms the audio information into text data. The input is audio data, and the output is text data.

[0424] Step 3:

[0425] The server passes text data to an emotion analysis engine to estimate the user's emotions. The emotion analysis engine analyzes the text data to identify the user's emotional state (e.g., confusion, anxiety). The input is text data, and the output is emotional state data.

[0426] Step 4:

[0427] The server assesses the likelihood of fraud based on emotional state data. This assessment is performed by an algorithm that generates a fraud warning if confusion or anxiety exceeds a certain threshold. The input is emotional state data, and the output is fraud risk assessment data.

[0428] Step 5:

[0429] If fraud is detected, the server generates countermeasures. These measures include warning messages to the user and orders to suspend transactions. The input is data indicating fraud risk, and the output is data indicating the countermeasures.

[0430] Step 6:

[0431] The terminal receives policy data provided by the server and issues an audio warning to the user. Specifically, it converts text-based warning messages into speech using a speech synthesis system and plays them through the terminal's speaker. The input is policy data, and the output is an audio warning.

[0432] Step 7:

[0433] If a transaction is determined to be risky, the terminal automatically blocks the transaction and notifies the user of the result. This reduces the risk of the user engaging in fraudulent transactions. The input is the fraud risk assessment data, and the output is the transaction blocking command.

[0434] 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.

[0435] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0436] 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.

[0437] [Third Embodiment]

[0438] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0439] 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.

[0440] 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).

[0441] 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.

[0442] 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.

[0443] 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).

[0444] 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.

[0445] 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.

[0446] 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.

[0447] 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.

[0448] 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.

[0449] 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".

[0450] The system of this invention is designed to analyze incoming calls by users in real time and to take appropriate action based on their content. The server, terminal, and user components each perform their respective functions as follows.

[0451] server

[0452] The server plays a central role in receiving and analyzing audio data during calls. Specifically, the server uses a voice analysis engine to perform sentiment analysis, estimating basic emotional states such as reassurance, confusion, and anxiety from the speaker's voice. To detect typical scam phrases, natural language processing techniques are applied, and the call content is analyzed in real time as text data. If fraud is suspected, the server immediately generates a response strategy and creates a dialogue script to psychologically unsettle the scammer.

[0453] terminal

[0454] The device transmits the audio received by the user to the server and streams the data. It then synthesizes the response instructed by the server and transmits it to the scammer as actual audio. The device also notifies the user of the situation in real time and provides safety advice via voice or screen display.

[0455] User

[0456] Users are beneficiaries of this system, which protects them from fraudulent calls. Users understand that their calls are monitored and that advice will be provided as needed, ensuring a safe calling experience. After the call ends, users receive a report generated by the server via their device, allowing them to review call details and a determination based on the suspicion of fraud.

[0457] Specific example

[0458] As a concrete example, consider a scenario where User B receives a message from a scammer. The server detects phrases suggesting potential fraud through voice communication and senses User B's anxiety through sentiment analysis. The server then generates a question-and-answer dialogue designed to expose inconsistencies in the scammer's statements, and the terminal speaks this script. User B receives instructions from the terminal to perform multiple verifications, preventing the scammer from obtaining information, and after the call ends, can review a detailed report to understand the situation. In this way, the system enables immediate prevention of fraud.

[0459] The following describes the processing flow.

[0460] Step 1:

[0461] As soon as the device initiates a call, it streams the audio data to the server in real time.

[0462] Step 2:

[0463] The server processes the received audio data using a speech analysis engine, analyzing the speaker's tone, pitch, speed, etc., to estimate their emotional state.

[0464] Step 3:

[0465] The server uses a natural language processing module to convert audio data into text and detects typical scam phrases and keywords.

[0466] Step 4:

[0467] The server evaluates the sentiment analysis results and phrase detection results to determine whether there is a suspicion of fraud.

[0468] Step 5:

[0469] If fraud is suspected, the server generates a response strategy and creates a dialogue script to psychologically unsettle the fraudster.

[0470] Step 6:

[0471] The terminal receives a dialogue script from the server, synthesizes it into speech, and speaks it to the scammer.

[0472] Step 7:

[0473] The device notifies the user of the call status and provides safety advice via voice or on-screen display.

[0474] Step 8:

[0475] After the call ends, the server generates a report containing the call history and the basis for the fraud determination, and provides it to the user via the terminal.

[0476] (Example 1)

[0477] 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."

[0478] In modern society, telephone fraud is on the rise, and individuals need immediate and effective measures to protect themselves from these scams. However, conventional systems have difficulty detecting fraud during a call and providing appropriate responses in real time. Therefore, there is a need for a system that can detect fraud early during a call and provide users with appropriate responses and assistance information in real time.

[0479] 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.

[0480] In this invention, the server includes means for receiving voice information and analyzing the speaker's emotional state, means for detecting suspicion of fraud based on the analyzed emotional state, and means for generating a response plan and dialogue procedures to psychologically unsettle the fraudster. This makes it possible to detect fraud during a call in real time and to quickly provide the user with appropriate responses and safety information.

[0481] "Voice information" refers to the waveform data of sounds contained in phone calls or voice input, and by using this, it becomes possible to analyze the voice spoken by the speaker.

[0482] "Emotional state" refers to the speaker's internal psychological state as analyzed from auditory information, and is categorized into states such as reassurance, confusion, and anxiety.

[0483] "Suspicion of fraud" refers to a situation that may indicate fraudulent activity, as judged from the content of the call and the speaker's emotional state.

[0484] A "response plan" refers to a series of responses and actions taken to prevent fraud when a suspected fraud is detected.

[0485] A "dialogue procedure" refers to a consistent flow of conversation that consists of questions and answers and enticing questions designed to psychologically manipulate the con artist.

[0486] "Speech synthesis" refers to a technology that converts text-based dialogue procedures into speech, creating digital audio data that can be spoken through a speaker.

[0487] "Natural language processing" refers to artificial intelligence technology that enables computers to understand, analyze, and generate natural language used by humans.

[0488] A "report" refers to a document provided to the user that contains a detailed history of the call and the grounds for determining that it was a scam.

[0489] The system of this invention is configured to analyze calls received by users, detect potential fraud in real time, and ensure appropriate responses. It is implemented using three main components: a server, a terminal, and the user.

[0490] Server Embodiment

[0491] The server plays a central role in receiving and analyzing audio information during a call. Equipped with an advanced speech analysis engine (e.g., speech recognition technology), the server estimates the speaker's emotional state from the audio. This information is further converted into text data by a natural language processing engine (e.g., natural language processing technology) and used to detect typical patterns and phrases of fraud. If fraud is suspected, a generative AI model is used to generate dialogue sequences designed to psychologically manipulate the fraudster.

[0492] Terminal embodiment

[0493] The device is responsible for transmitting the audio data of calls received by the user to the server. It streams the audio in real time, enabling secure and rapid communication. Based on the response plan sent from the server, it also uses speech synthesis technology to create voice commands for the conversation and transmit them to the fraudster. The device also notifies the user of the situation and provides safety advice via voice or on-screen display.

[0494] User Embodiment

[0495] Users are subject to the protection of this system and should be provided with a safe calling experience. They understand that calls are monitored and appropriate advice will be provided when necessary. Furthermore, after the call ends, they can receive a report sent from the server via their device, allowing them to review call details and fraud detection results.

[0496] Specific example

[0497] For example, consider a scenario where user B receives a suspicious phone call. The server detects user B's anxiety through voice analysis and identifies characteristic phrases of fraud from the call content. Next, a generative AI model is used to generate a dialogue procedure incorporating effective questions for the scammer, which the terminal then synthesizes and speaks aloud. This prevents malicious information acquisition for user B, and after the call, user B can review the report to ensure their safety.

[0498] An example of a prompt for a generative AI model might be: "Analyze the call received by the user in real time and detect potential fraud. If fraud is suspected, create a dialogue script that suggests specific countermeasures."

[0499] This allows the system to effectively protect users from fraud and provide a safe and secure calling environment.

[0500] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0501] Step 1:

[0502] The device captures the audio of calls received by the user and streams that audio information to the server in real time. The input is the audio data of the call, and the output is the live streaming data to the server. Efficient audio compression technology is used in this process to minimize latency.

[0503] Step 2:

[0504] The server inputs the received audio information into a speech analysis engine to estimate the speaker's emotional state. The input is the audio information, and the output is the analysis result of the speaker's emotions. A specific algorithm is used to analyze the tone and pitch changes of the voice and identify emotions such as reassurance or anxiety.

[0505] Step 3:

[0506] The server inputs audio information into a natural language processing engine, converts it into text data, and detects typical phrases and patterns of fraud. The input consists of audio information and sentiment analysis results, while the output is text data and analysis results indicating suspected fraud. This process analyzes the content extracted from the audio through a language model.

[0507] Step 4:

[0508] If fraud is suspected, the server inputs prompt text into a generating AI model, which then generates a dialogue procedure for the fraudster. The input consists of sentiment analysis results and text analysis results, and the output is the dialogue procedure. Here, the AI ​​model designs a conversation flow that includes questions designed to exploit the fraudster's psychology.

[0509] Step 5:

[0510] The terminal receives dialogue instructions from the server as input and converts them into speech using a speech synthesis engine. The input is the dialogue instructions, and the output is the synthesized speech. In this step, the generated script is output in a voice that resembles the user's voice to create a natural response.

[0511] Step 6:

[0512] The terminal transmits synthesized voice to the caller and notifies the user of the situation. The input is synthesized voice, and the output is actual voice communication and screen display. The terminal also provides warnings and additional instructions to the user.

[0513] Step 7:

[0514] After the call ends, the server generates a report based on the call records and provides it to the user via the terminal. The input is the call data and analysis results, and the output is a detailed report. This report includes the grounds for suspected fraud and a summary of the call.

[0515] This process allows the system to detect suspected fraud in real time and provide advanced security support to users.

[0516] (Application Example 1)

[0517] 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."

[0518] Traditional communication methods make it difficult to detect phone scams and fraudulent activities, potentially threatening user safety and privacy. Furthermore, users lack effective means to immediately defend themselves against potential fraud. Therefore, there is a need to provide safe and secure communication by analyzing call content in real time and notifying users of potential fraud.

[0519] 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.

[0520] In this invention, the server includes means for acquiring voice information and analyzing the speaker's emotions in real time; means for identifying the possibility of fraudulent activity when the analyzed emotions indicate confusion or anxiety; and means for analyzing the content of the call and detecting phrases indicating fraud or changes in the user's emotions, thereby presenting a safety alert to the user in real time during the call. This enables immediate detection of fraudulent activity during a call and notification to the user.

[0521] "Voice information" refers to voice data obtained from phone calls, recordings, and other sources.

[0522] "Speaker" refers to the person providing the audio information.

[0523] "Emotional analysis" is a technology that analyzes audio information to evaluate the speaker's psychological state.

[0524] "Fraudulent activity" refers to communication conducted for fraudulent or illicit purposes.

[0525] "Real-time" refers to processing and analyzing events at the very moment they occur.

[0526] "Identification" is the act of detecting, classifying, or recognizing specific features or patterns.

[0527] "Analysis" is the act of thoroughly examining information and extracting meaning and patterns.

[0528] A "phrase" refers to a combination of words that have a specific meaning.

[0529] "User" refers to an individual or group that utilizes the system.

[0530] An "alert" refers to a notification that draws attention or warning.

[0531] The server acquires audio information in real time from smartphones and other digital devices. The Google Speech-to-Text API is used as the speech analysis engine to convert the audio information into text. Next, natural language processing techniques are used with this text data to identify phrases that indicate the speaker's emotions and specific fraudulent activities. For example, libraries such as spaCy and NLTK can be used to identify potentially fraudulent phrases.

[0532] The device uses a speech synthesis engine (e.g., Amazon Polly) to convert information obtained based on sentiment analysis into speech and provides real-time alerts to the user. The device displays the speech data and analysis results on the screen to help the user understand the current situation. This notification function allows the user to receive immediate warnings of fraudulent activity and take appropriate action.

[0533] After the call ends, users can refer to a detailed report provided by the server to verify the basis for any alleged misconduct during the call. This report can serve as a reference for users to avoid similar risks in the future.

[0534] As a concrete example, consider a scenario where a user is at home when a fraudulent person posing as a delivery service contacts them about a lost package. The server immediately detects phrases indicating fraudulent activity, such as "You need to verify the delivery company," and uses sentiment analysis to identify the user's confusion. In response, the device issues a warning to the user, prompting them to take precautions. This allows the user to understand the risk of fraud and protect their personal information.

[0535] An example of a prompt message might be, "What is the best course of action if I receive a fraudulent phone call?" Based on this message, the generative AI model generates appropriate countermeasures.

[0536] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0537] Step 1:

[0538] The server acquires audio information in real time from smartphones and digital devices. It sends the audio data as input to the Google Speech-to-Text API, which generates text data as output. This text data forms the basis for subsequent analysis processes.

[0539] Step 2:

[0540] The server analyzes the acquired text data using natural language processing libraries (e.g., spaCy or NLTK). It performs specific pattern recognition and sentiment evaluation on the input text data, and identifies relevant fraudulent phrases and emotional states as output. Specifically, it classifies phrases that suggest fraud and the emotions of the user, and evaluates the likelihood of fraud.

[0541] Step 3:

[0542] The server assesses the likelihood of fraud based on sentiment analysis and detection of fraudulent phrases, and generates appropriate dialogue instructions. Using the analysis results as input, it converts the dialogue script for a speech synthesis engine (e.g., Amazon Polly) and prepares an alert to send to the user as output.

[0543] Step 4:

[0544] The terminal converts the dialogue script received from the server into speech and issues warnings to the user in real time. It takes the dialogue script as input and converts it into speech, while outputting notifications and warnings to the user. Specifically, it emits warnings from the terminal's speaker.

[0545] Step 5:

[0546] After the call ends, the server generates a detailed report based on the call history and analysis results, and provides it to the user. The report is created by combining the call content (input) and analysis results (output). This allows the user to gain a deeper understanding of the call details and the possibility of fraud.

[0547] 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.

[0548] This invention is a fraud prevention system that combines an emotion engine, enabling more effective fraud response by analyzing user emotions in real time. Specifically, it provides a mechanism in which a server and a terminal work together to detect suspected fraud early and take appropriate countermeasures.

[0549] server

[0550] The server is equipped with a speech analysis engine and an emotion engine for analyzing audio data received from terminals. The speech analysis engine is responsible for the basic processing of estimating the speaker's emotional state and determining the likelihood of fraud based on this. The emotion engine further recognizes emotions from the user's voice during a call, and if the user is feeling confused or anxious, it immediately assesses the risk and dynamically adjusts the priority of fraud prevention measures. This emotion data is also accumulated and used for long-term pattern analysis.

[0551] terminal

[0552] The device streams voice data to a server, and the server generates a response which is then synthesized and spoken aloud. Based on notifications from the emotion engine, safety advice is provided to the user via on-screen or audio notifications. This helps users anticipate fraud risks and respond appropriately.

[0553] User

[0554] Users are protected from fraud through this system. During a call, the system performs background analysis and takes action as needed. After the call ends, users receive a report generated by the server, allowing them to review details including their emotional state and the basis for the fraud assessment. This provides users with feedback that can help them take future preventative measures.

[0555] Specific example

[0556] As a concrete example, when user C receives an unexpected sales call, the emotion engine detects subtle confusion and anxiety in the voice. Based on this, the server determines the call may be fraudulent and immediately generates an advanced dialogue strategy to unsettle the scammer. The terminal then uses this script to communicate with the scammer. User C is encouraged to end the call before their anxiety escalates, and simultaneously gains reassurance through a post-call report. In this way, the system achieves flexible fraud response that reflects the user's emotions.

[0557] The following describes the processing flow.

[0558] Step 1:

[0559] The device initiates a call and streams audio data to the server in real time. The emotion engine also starts simultaneously and begins collecting audio data.

[0560] Step 2:

[0561] The server uses a speech analysis engine to analyze the tone and speed of the speaker's voice from the audio data and estimate the speaker's emotional state. The user's voice is specifically analyzed by the emotion engine, and its emotional state is evaluated in real time.

[0562] Step 3:

[0563] The server applies natural language processing technology to convert the audio data into text and detects typical phrases and keywords used in scams.

[0564] Step 4:

[0565] The server integrates sentiment analysis and phrase detection results to determine if there is a suspicion of fraud. If the user shows signs of confusion or anxiety, fraud prevention measures are prioritized.

[0566] Step 5:

[0567] If fraud is suspected, the server generates a dialogue script designed to psychologically unsettle the fraudster. If necessary, the script's content is adjusted based on the user's emotions.

[0568] Step 6:

[0569] The terminal receives a dialogue script from the server, converts it using speech synthesis technology, and speaks it to the scammer.

[0570] Step 7:

[0571] The device notifies the user of the call status via voice or on-screen display, along with safety advice.

[0572] Step 8:

[0573] After the call ends, the server generates a report containing the call history and the basis for the fraud detection, and provides it to the user via their device. This allows the user to understand the overall picture of the call, including their emotional state.

[0574] (Example 2)

[0575] 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."

[0576] In recent years, fraudulent activities have diversified, with telephone scams being particularly on the rise. Traditional fraud prevention methods are often limited to detecting simple phrases and issuing delayed warnings, and real-time sentiment analysis and dynamic responses are frequently difficult to implement. As a result, the risk of users becoming victims of fraud is increasing. There is a need for a system that can prevent this and protect users more safely.

[0577] 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.

[0578] In this invention, the server includes means for acquiring voice data and transmitting it to a data processing device via a communication device; means for analyzing the received voice data in detail and estimating the speaker's emotions; and means for evaluating the likelihood of a high level of fraud if the estimated emotions indicate confusion or anxiety. This enables users to reduce the risk of fraud in real time and take quick and appropriate action.

[0579] "Voice data" refers to digital information containing the user's spoken content, which is processed via a communication device.

[0580] "Communication equipment" is a general term for hardware and software used to acquire voice data and transfer it to another device or server.

[0581] A "data processing device" is a computer system that analyzes received audio data and processes the information.

[0582] "Means for estimating emotions" refers to algorithms and analysis engines for identifying a speaker's emotional state from audio data.

[0583] "Means of assessing suspicion of fraud" refers to a process of determining the likelihood of fraud based on sentiment estimation results, in light of numerical values ​​and evaluation criteria.

[0584] A "generative AI model" is an artificial intelligence technology that generates new outputs using past data and learning algorithms.

[0585] A "dialogue script" is a script that outlines the expected responses to a fraudster, and is played back using speech synthesis.

[0586] "Speech synthesis" is a technology that converts text information into speech information and is used to produce virtual voice responses.

[0587] This fraud prevention system consists of three components: a server, a terminal, and a user. Specific embodiments of each component are shown below.

[0588] The server features a voice analysis engine and an emotion engine, which analyze voice data transmitted from the terminal in real time. The voice analysis engine uses a high-performance processor and database server to analyze the speaker's voice characteristics and estimate their emotional state. This data is further analyzed by the emotion engine, and if negative emotions such as confusion or anxiety are detected, the risk of fraud is immediately assessed. The emotion data is accumulated and used for long-term trend analysis.

[0589] The device is responsible for acquiring the user's voice data and streaming it to the server in real time. The device is equipped with a microphone and speaker and runs an audio streaming application. Upon receiving a response strategy generated by the server, it uses speech synthesis technology to generate and play a response to the scammer. The device also informs the user of the risk of fraud and provides safety advice through screen displays and audio notifications.

[0590] This system protects users from the risk of fraud. The system performs sentiment analysis in the background and takes prompt action if fraud is suspected. After the call ends, users can review the report provided by the server and understand the detected sentiment and the reasons for the fraud assessment, which can help them take future preventative measures.

[0591] As a concrete example, consider a scenario where a user receives an unexpected sales call. In this case, if the emotion engine detects slight confusion and anxiety, the server determines that the sales call may be a scam. Based on this, the server generates sophisticated dialogue strategies to psychologically unsettle the scammer. The terminal synthesizes these strategies into speech and plays them back to the scammer, helping the user end the call before feeling anxious. In this example, the prompt message "Evaluate the likelihood of fraud based on the anxiety level detected from the user's voice and generate sophisticated dialogue strategies" is input into the generating AI model, providing the optimal solution.

[0592] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0593] Step 1:

[0594] The device acquires the user's voice data. A microphone is used to capture the user's voice during a call in digital format. The input is the user's raw voice, and this data is temporarily stored within the device. The output is digitized voice data.

[0595] Step 2:

[0596] The terminal streams the acquired audio data to the server. Wi-Fi or mobile data communication is used for this purpose. The input is digital audio data stored within the terminal, and the output is the process of that data being wirelessly transferred to the server. An efficient data transfer protocol is applied to ensure real-time performance during this process.

[0597] Step 3:

[0598] The server analyzes the received audio data using a speech analysis engine. The analysis extracts characteristics such as intonation, speed, and volume. The input is the digital audio data sent to the server, and the output is a dataset of these characteristics quantified. This dataset is then used for further emotion estimation.

[0599] Step 4:

[0600] The server uses an emotion engine to estimate emotions based on the results of speech analysis. Specifically, it estimates the speaker's emotional state (e.g., confusion, anxiety) using the emotion engine's model based on extracted features. The input is feature data generated by the speech analysis engine, and the output is the estimated emotion label.

[0601] Step 5:

[0602] The server assesses the likelihood of fraud based on estimated sentiment. Here, a generative AI model is used to compare sentiment data with past fraud patterns. The input is the sentiment label estimated by the sentiment engine, and the output is a fraud risk assessment index. If the assessment index is high, the process proceeds to the next step.

[0603] Step 6:

[0604] The server generates a response strategy for the fraudster when it assesses the likelihood of fraud. Specifically, it uses a generative AI model to generate a dialogue script from the prompt text to psychologically influence the fraudster. The input is an assessment of fraud risk, and the output is the generated dialogue script.

[0605] Step 7:

[0606] The terminal plays back the dialogue script received from the server using speech synthesis technology. The audio is played through the speaker. The input is the voice script sent from the server, and the output is a voice message directed at the fraudster. This process employs a psychological approach to the fraudster.

[0607] Step 8:

[0608] The device notifies the user of potential fraud and suggests safe countermeasures. Notifications are delivered via screen displays and additional audio messages. Input is the server's evaluation result, and output is a warning and suggested action to the user. This allows the user to make informed decisions and mitigate the risk of fraud.

[0609] (Application Example 2)

[0610] 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."

[0611] With the widespread adoption of electronic payments, the risk of users becoming victims of fraud is increasing. Therefore, there is a need for a system that analyzes user emotions in real time, quickly detects potential fraud, and provides appropriate responses. Furthermore, a mechanism to verify the security of transactions and immediately terminate risky transactions is also necessary.

[0612] 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.

[0613] In this invention, the server includes means for receiving voice information and immediately analyzing the speaker's emotions, means for detecting the possibility of fraud if the analyzed emotions express confusion or anxiety, and means for creating countermeasures and providing appropriate responses to the responder when the possibility of fraud is detected. This makes it possible for users to detect the risk of fraud more quickly and for a safer trading environment to be provided.

[0614] "Voice information" refers to digital data used to collect and analyze the content of users' phone calls in real time.

[0615] A "device that instantly analyzes the speaker's emotions" is a technology that quickly estimates the user's emotional state from received audio information.

[0616] A "device for detecting potential fraud" is a function that identifies potential fraud when suspicious signs are found in the analyzed emotions.

[0617] A "device for generating response measures" is a technology for generating specific actions and intervention methods to be taken when fraud is suspected.

[0618] A "device that provides appropriate responses to those involved" is a function that provides appropriate support and guidance to users and other stakeholders based on the policies that have been created.

[0619] "Payment communication" refers to the exchange of data that occurs during the process of conducting financial transactions and payments using electronic means.

[0620] A "device for verifying the security of transactions" is a technology that detects signs of fraudulent activity during settlement communications and performs a risk assessment in response.

[0621] A "mechanism to immediately terminate risky trades" is a process that automatically stops a trade when it is deemed to be high-risk, thereby preventing potential losses.

[0622] This invention relates to a system that uses voice analysis and sentiment analysis to reduce the risk of fraud during electronic payments. This system consists of a smartphone terminal and server software running in the background.

[0623] The server receives audio data and analyzes the audio information in real time. A speech analysis engine, such as Google Cloud Speech-to-Text, is used for the audio analysis. From the analyzed audio information, an emotion analysis engine instantly estimates the speaker's emotions and detects potential fraud based on the results. If fraud is detected, the server creates countermeasures and generates an appropriate response based on those measures.

[0624] The terminal plays this response as an audio message to notify the user. Furthermore, if a risky transaction occurs, the terminal has a function to automatically block the transaction and notify the user immediately to terminate it.

[0625] For example, if anxiety or confusion is detected from voice information when a user attempts to make an electronic payment, the server will immediately issue a warning, and the terminal will provide advice such as, "This transaction may be risky." Another example of a prompt message is, "I feel uneasy about this new online payment method, and I would like to confirm if this transaction is safe. Please let me know if the system has detected anything unusual." This prompt message forms the basis for the generative AI model to provide appropriate support.

[0626] In this way, the server and terminal work together to provide users with consistent support for fraud prevention.

[0627] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0628] Step 1:

[0629] The server receives audio data from the terminal. This audio data is a recording of the user's phone call and is transmitted in real time. Audio information is provided as input, and basic data for audio analysis is prepared as output.

[0630] Step 2:

[0631] The server converts the received audio data into text using Google Cloud Speech-to-Text. This process transforms the audio information into text data. The input is audio data, and the output is text data.

[0632] Step 3:

[0633] The server passes text data to an emotion analysis engine to estimate the user's emotions. The emotion analysis engine analyzes the text data to identify the user's emotional state (e.g., confusion, anxiety). The input is text data, and the output is emotional state data.

[0634] Step 4:

[0635] The server assesses the likelihood of fraud based on emotional state data. This assessment is performed by an algorithm that generates a fraud warning if confusion or anxiety exceeds a certain threshold. The input is emotional state data, and the output is fraud risk assessment data.

[0636] Step 5:

[0637] If fraud is detected, the server generates countermeasures. These measures include warning messages to the user and orders to suspend transactions. The input is data indicating fraud risk, and the output is data indicating the countermeasures.

[0638] Step 6:

[0639] The terminal receives policy data provided by the server and issues an audio warning to the user. Specifically, it converts text-based warning messages into speech using a speech synthesis system and plays them through the terminal's speaker. The input is policy data, and the output is an audio warning.

[0640] Step 7:

[0641] If a transaction is determined to be risky, the terminal automatically blocks the transaction and notifies the user of the result. This reduces the risk of the user engaging in fraudulent transactions. The input is the fraud risk assessment data, and the output is the transaction blocking command.

[0642] 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.

[0643] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0644] 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.

[0645] [Fourth Embodiment]

[0646] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0647] 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.

[0648] 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).

[0649] 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.

[0650] 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.

[0651] 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).

[0652] 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.

[0653] 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.

[0654] 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.

[0655] 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.

[0656] 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.

[0657] 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.

[0658] 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".

[0659] The system of this invention is designed to analyze incoming calls by users in real time and to take appropriate action based on their content. The server, terminal, and user components each perform their respective functions as follows.

[0660] server

[0661] The server plays a central role in receiving and analyzing audio data during calls. Specifically, the server uses a voice analysis engine to perform sentiment analysis, estimating basic emotional states such as reassurance, confusion, and anxiety from the speaker's voice. To detect typical scam phrases, natural language processing techniques are applied, and the call content is analyzed in real time as text data. If fraud is suspected, the server immediately generates a response strategy and creates a dialogue script to psychologically unsettle the scammer.

[0662] terminal

[0663] The device transmits the audio received by the user to the server and streams the data. It then synthesizes the response instructed by the server and transmits it to the scammer as actual audio. The device also notifies the user of the situation in real time and provides safety advice via voice or screen display.

[0664] User

[0665] Users are beneficiaries of this system, which protects them from fraudulent calls. Users understand that their calls are monitored and that advice will be provided as needed, ensuring a safe calling experience. After the call ends, users receive a report generated by the server via their device, allowing them to review call details and a determination based on the suspicion of fraud.

[0666] Specific example

[0667] As a concrete example, consider a scenario where User B receives a message from a scammer. The server detects phrases suggesting potential fraud through voice communication and senses User B's anxiety through sentiment analysis. The server then generates a question-and-answer dialogue designed to expose inconsistencies in the scammer's statements, and the terminal speaks this script. User B receives instructions from the terminal to perform multiple verifications, preventing the scammer from obtaining information, and after the call ends, can review a detailed report to understand the situation. In this way, the system enables immediate prevention of fraud.

[0668] The following describes the processing flow.

[0669] Step 1:

[0670] As soon as the device initiates a call, it streams the audio data to the server in real time.

[0671] Step 2:

[0672] The server processes the received audio data using a speech analysis engine, analyzing the speaker's tone, pitch, speed, etc., to estimate their emotional state.

[0673] Step 3:

[0674] The server uses a natural language processing module to convert audio data into text and detects typical scam phrases and keywords.

[0675] Step 4:

[0676] The server evaluates the sentiment analysis results and phrase detection results to determine whether there is a suspicion of fraud.

[0677] Step 5:

[0678] If fraud is suspected, the server generates a response strategy and creates a dialogue script to psychologically unsettle the fraudster.

[0679] Step 6:

[0680] The terminal receives a dialogue script from the server, synthesizes it into speech, and speaks it to the scammer.

[0681] Step 7:

[0682] The device notifies the user of the call status and provides safety advice via voice or on-screen display.

[0683] Step 8:

[0684] After the call ends, the server generates a report containing the call history and the basis for the fraud determination, and provides it to the user via the terminal.

[0685] (Example 1)

[0686] 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".

[0687] In modern society, telephone fraud is on the rise, and individuals need immediate and effective measures to protect themselves from these scams. However, conventional systems have difficulty detecting fraud during a call and providing appropriate responses in real time. Therefore, there is a need for a system that can detect fraud early during a call and provide users with appropriate responses and assistance information in real time.

[0688] 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.

[0689] In this invention, the server includes means for receiving voice information and analyzing the speaker's emotional state, means for detecting suspicion of fraud based on the analyzed emotional state, and means for generating a response plan and dialogue procedures to psychologically unsettle the fraudster. This makes it possible to detect fraud during a call in real time and to quickly provide the user with appropriate responses and safety information.

[0690] "Voice information" refers to the waveform data of sounds contained in phone calls or voice input, and by using this, it becomes possible to analyze the voice spoken by the speaker.

[0691] "Emotional state" refers to the speaker's internal psychological state as analyzed from auditory information, and is categorized into states such as reassurance, confusion, and anxiety.

[0692] "Suspicion of fraud" refers to a situation that may indicate fraudulent activity, as judged from the content of the call and the speaker's emotional state.

[0693] A "response plan" refers to a series of responses and actions taken to prevent fraud when a suspected fraud is detected.

[0694] A "dialogue procedure" refers to a consistent flow of conversation that consists of questions and answers and enticing questions designed to psychologically manipulate the con artist.

[0695] "Speech synthesis" refers to a technology that converts text-based dialogue procedures into speech, creating digital audio data that can be spoken through a speaker.

[0696] "Natural language processing" refers to artificial intelligence technology that enables computers to understand, analyze, and generate natural language used by humans.

[0697] A "report" refers to a document provided to the user that contains a detailed history of the call and the grounds for determining that it was a scam.

[0698] The system of this invention is configured to analyze calls received by users, detect potential fraud in real time, and ensure appropriate responses. It is implemented using three main components: a server, a terminal, and the user.

[0699] Server Embodiment

[0700] The server plays a central role in receiving and analyzing audio information during a call. Equipped with an advanced speech analysis engine (e.g., speech recognition technology), the server estimates the speaker's emotional state from the audio. This information is further converted into text data by a natural language processing engine (e.g., natural language processing technology) and used to detect typical patterns and phrases of fraud. If fraud is suspected, a generative AI model is used to generate dialogue sequences designed to psychologically manipulate the fraudster.

[0701] Terminal embodiment

[0702] The device is responsible for transmitting the audio data of calls received by the user to the server. It streams the audio in real time, enabling secure and rapid communication. Based on the response plan sent from the server, it also uses speech synthesis technology to create voice commands for the conversation and transmit them to the fraudster. The device also notifies the user of the situation and provides safety advice via voice or on-screen display.

[0703] User Embodiment

[0704] Users are subject to the protection of this system and should be provided with a safe calling experience. They understand that calls are monitored and appropriate advice will be provided when necessary. Furthermore, after the call ends, they can receive a report sent from the server via their device, allowing them to review call details and fraud detection results.

[0705] Specific example

[0706] For example, consider a scenario where user B receives a suspicious phone call. The server detects user B's anxiety through voice analysis and identifies characteristic phrases of fraud from the call content. Next, a generative AI model is used to generate a dialogue procedure incorporating effective questions for the scammer, which the terminal then synthesizes and speaks aloud. This prevents malicious information acquisition for user B, and after the call, user B can review the report to ensure their safety.

[0707] An example of a prompt for a generative AI model might be: "Analyze the call received by the user in real time and detect potential fraud. If fraud is suspected, create a dialogue script that suggests specific countermeasures."

[0708] This allows the system to effectively protect users from fraud and provide a safe and secure calling environment.

[0709] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0710] Step 1:

[0711] The device captures the audio of calls received by the user and streams that audio information to the server in real time. The input is the audio data of the call, and the output is the live streaming data to the server. Efficient audio compression technology is used in this process to minimize latency.

[0712] Step 2:

[0713] The server inputs the received audio information into a speech analysis engine to estimate the speaker's emotional state. The input is the audio information, and the output is the analysis result of the speaker's emotions. A specific algorithm is used to analyze the tone and pitch changes of the voice and identify emotions such as reassurance or anxiety.

[0714] Step 3:

[0715] The server inputs audio information into a natural language processing engine, converts it into text data, and detects typical phrases and patterns of fraud. The input consists of audio information and sentiment analysis results, while the output is text data and analysis results indicating suspected fraud. This process analyzes the content extracted from the audio through a language model.

[0716] Step 4:

[0717] If fraud is suspected, the server inputs prompt text into a generating AI model, which then generates a dialogue procedure for the fraudster. The input consists of sentiment analysis results and text analysis results, and the output is the dialogue procedure. Here, the AI ​​model designs a conversation flow that includes questions designed to exploit the fraudster's psychology.

[0718] Step 5:

[0719] The terminal receives dialogue instructions from the server as input and converts them into speech using a speech synthesis engine. The input is the dialogue instructions, and the output is the synthesized speech. In this step, the generated script is output in a voice that resembles the user's voice to create a natural response.

[0720] Step 6:

[0721] The terminal transmits synthesized voice to the caller and notifies the user of the situation. The input is synthesized voice, and the output is actual voice communication and screen display. The terminal also provides warnings and additional instructions to the user.

[0722] Step 7:

[0723] After the call ends, the server generates a report based on the call records and provides it to the user via the terminal. The input is the call data and analysis results, and the output is a detailed report. This report includes the grounds for suspected fraud and a summary of the call.

[0724] This process allows the system to detect suspected fraud in real time and provide advanced security support to users.

[0725] (Application Example 1)

[0726] 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".

[0727] Traditional communication methods make it difficult to detect phone scams and fraudulent activities, potentially threatening user safety and privacy. Furthermore, users lack effective means to immediately defend themselves against potential fraud. Therefore, there is a need to provide safe and secure communication by analyzing call content in real time and notifying users of potential fraud.

[0728] 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.

[0729] In this invention, the server includes means for acquiring voice information and analyzing the speaker's emotions in real time; means for identifying the possibility of fraudulent activity when the analyzed emotions indicate confusion or anxiety; and means for analyzing the content of the call and detecting phrases indicating fraud or changes in the user's emotions, thereby presenting a safety alert to the user in real time during the call. This enables immediate detection of fraudulent activity during a call and notification to the user.

[0730] "Voice information" refers to voice data obtained from phone calls, recordings, and other sources.

[0731] "Speaker" refers to the person providing the audio information.

[0732] "Emotional analysis" is a technology that analyzes audio information to evaluate the speaker's psychological state.

[0733] "Fraudulent activity" refers to communication conducted for fraudulent or illicit purposes.

[0734] "Real-time" refers to processing and analyzing events at the very moment they occur.

[0735] "Identification" is the act of detecting, classifying, or recognizing specific features or patterns.

[0736] "Analysis" is the act of thoroughly examining information and extracting meaning and patterns.

[0737] A "phrase" refers to a combination of words that have a specific meaning.

[0738] "User" refers to an individual or group that utilizes the system.

[0739] An "alert" refers to a notification that draws attention or warning.

[0740] The server acquires audio information in real time from smartphones and other digital devices. The Google Speech-to-Text API is used as the speech analysis engine to convert the audio information into text. Next, natural language processing techniques are used with this text data to identify phrases that indicate the speaker's emotions and specific fraudulent activities. For example, libraries such as spaCy and NLTK can be used to identify potentially fraudulent phrases.

[0741] The device uses a speech synthesis engine (e.g., Amazon Polly) to convert information obtained based on sentiment analysis into speech and provides real-time alerts to the user. The device displays the speech data and analysis results on the screen to help the user understand the current situation. This notification function allows the user to receive immediate warnings of fraudulent activity and take appropriate action.

[0742] After the call ends, users can refer to a detailed report provided by the server to verify the basis for any alleged misconduct during the call. This report can serve as a reference for users to avoid similar risks in the future.

[0743] As a concrete example, consider a scenario where a user is at home when a fraudulent person posing as a delivery service contacts them about a lost package. The server immediately detects phrases indicating fraudulent activity, such as "You need to verify the delivery company," and uses sentiment analysis to identify the user's confusion. In response, the device issues a warning to the user, prompting them to take precautions. This allows the user to understand the risk of fraud and protect their personal information.

[0744] An example of a prompt message might be, "What is the best course of action if I receive a fraudulent phone call?" Based on this message, the generative AI model generates appropriate countermeasures.

[0745] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0746] Step 1:

[0747] The server acquires audio information in real time from smartphones and digital devices. It sends the audio data as input to the Google Speech-to-Text API, which generates text data as output. This text data forms the basis for subsequent analysis processes.

[0748] Step 2:

[0749] The server analyzes the acquired text data using natural language processing libraries (e.g., spaCy or NLTK). It performs specific pattern recognition and sentiment evaluation on the input text data, and identifies relevant fraudulent phrases and emotional states as output. Specifically, it classifies phrases that suggest fraud and the emotions of the user, and evaluates the likelihood of fraud.

[0750] Step 3:

[0751] The server assesses the likelihood of fraud based on sentiment analysis and detection of fraudulent phrases, and generates appropriate dialogue instructions. Using the analysis results as input, it converts the dialogue script for a speech synthesis engine (e.g., Amazon Polly) and prepares an alert to send to the user as output.

[0752] Step 4:

[0753] The terminal converts the dialogue script received from the server into speech and issues warnings to the user in real time. It takes the dialogue script as input and converts it into speech, while outputting notifications and warnings to the user. Specifically, it emits warnings from the terminal's speaker.

[0754] Step 5:

[0755] After the call ends, the server generates a detailed report based on the call history and analysis results, and provides it to the user. The report is created by combining the call content (input) and analysis results (output). This allows the user to gain a deeper understanding of the call details and the possibility of fraud.

[0756] 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.

[0757] This invention is a fraud prevention system that combines an emotion engine, enabling more effective fraud response by analyzing user emotions in real time. Specifically, it provides a mechanism in which a server and a terminal work together to detect suspected fraud early and take appropriate countermeasures.

[0758] server

[0759] The server is equipped with a speech analysis engine and an emotion engine for analyzing audio data received from terminals. The speech analysis engine is responsible for the basic processing of estimating the speaker's emotional state and determining the likelihood of fraud based on this. The emotion engine further recognizes emotions from the user's voice during a call, and if the user is feeling confused or anxious, it immediately assesses the risk and dynamically adjusts the priority of fraud prevention measures. This emotion data is also accumulated and used for long-term pattern analysis.

[0760] terminal

[0761] The device streams voice data to a server, and the server generates a response which is then synthesized and spoken aloud. Based on notifications from the emotion engine, safety advice is provided to the user via on-screen or audio notifications. This helps users anticipate fraud risks and respond appropriately.

[0762] User

[0763] Users are protected from fraud through this system. During a call, the system performs background analysis and takes action as needed. After the call ends, users receive a report generated by the server, allowing them to review details including their emotional state and the basis for the fraud assessment. This provides users with feedback that can help them take future preventative measures.

[0764] Specific example

[0765] As a concrete example, when user C receives an unexpected sales call, the emotion engine detects subtle confusion and anxiety in the voice. Based on this, the server determines the call may be fraudulent and immediately generates an advanced dialogue strategy to unsettle the scammer. The terminal then uses this script to communicate with the scammer. User C is encouraged to end the call before their anxiety escalates, and simultaneously gains reassurance through a post-call report. In this way, the system achieves flexible fraud response that reflects the user's emotions.

[0766] The following describes the processing flow.

[0767] Step 1:

[0768] The device initiates a call and streams audio data to the server in real time. The emotion engine also starts simultaneously and begins collecting audio data.

[0769] Step 2:

[0770] The server uses a speech analysis engine to analyze the tone and speed of the speaker's voice from the audio data and estimate the speaker's emotional state. The user's voice is specifically analyzed by the emotion engine, and its emotional state is evaluated in real time.

[0771] Step 3:

[0772] The server applies natural language processing technology to convert the audio data into text and detects typical phrases and keywords used in scams.

[0773] Step 4:

[0774] The server integrates sentiment analysis and phrase detection results to determine if there is a suspicion of fraud. If the user shows signs of confusion or anxiety, fraud prevention measures are prioritized.

[0775] Step 5:

[0776] If fraud is suspected, the server generates a dialogue script designed to psychologically unsettle the fraudster. If necessary, the script's content is adjusted based on the user's emotions.

[0777] Step 6:

[0778] The terminal receives a dialogue script from the server, converts it using speech synthesis technology, and speaks it to the scammer.

[0779] Step 7:

[0780] The device notifies the user of the call status via voice or on-screen display, along with safety advice.

[0781] Step 8:

[0782] After the call ends, the server generates a report containing the call history and the basis for the fraud detection, and provides it to the user via their device. This allows the user to understand the overall picture of the call, including their emotional state.

[0783] (Example 2)

[0784] 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".

[0785] In recent years, fraudulent activities have diversified, with telephone scams being particularly on the rise. Traditional fraud prevention methods are often limited to detecting simple phrases and issuing delayed warnings, and real-time sentiment analysis and dynamic responses are frequently difficult to implement. As a result, the risk of users becoming victims of fraud is increasing. There is a need for a system that can prevent this and protect users more safely.

[0786] 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.

[0787] In this invention, the server includes means for acquiring voice data and transmitting it to a data processing device via a communication device; means for analyzing the received voice data in detail and estimating the speaker's emotions; and means for evaluating the likelihood of a high level of fraud if the estimated emotions indicate confusion or anxiety. This enables users to reduce the risk of fraud in real time and take quick and appropriate action.

[0788] "Voice data" refers to digital information containing the user's spoken content, which is processed via a communication device.

[0789] "Communication equipment" is a general term for hardware and software used to acquire voice data and transfer it to another device or server.

[0790] A "data processing device" is a computer system that analyzes received audio data and processes the information.

[0791] "Means for estimating emotions" refers to algorithms and analysis engines for identifying a speaker's emotional state from audio data.

[0792] "Means of assessing suspicion of fraud" refers to a process of determining the likelihood of fraud based on sentiment estimation results, in light of numerical values ​​and evaluation criteria.

[0793] A "generative AI model" is an artificial intelligence technology that generates new outputs using past data and learning algorithms.

[0794] A "dialogue script" is a script that outlines the expected responses to a fraudster, and is played back using speech synthesis.

[0795] "Speech synthesis" is a technology that converts text information into speech information and is used to produce virtual voice responses.

[0796] This fraud prevention system consists of three components: a server, a terminal, and a user. Specific embodiments of each component are shown below.

[0797] The server features a voice analysis engine and an emotion engine, which analyze voice data transmitted from the terminal in real time. The voice analysis engine uses a high-performance processor and database server to analyze the speaker's voice characteristics and estimate their emotional state. This data is further analyzed by the emotion engine, and if negative emotions such as confusion or anxiety are detected, the risk of fraud is immediately assessed. The emotion data is accumulated and used for long-term trend analysis.

[0798] The device is responsible for acquiring the user's voice data and streaming it to the server in real time. The device is equipped with a microphone and speaker and runs an audio streaming application. Upon receiving a response strategy generated by the server, it uses speech synthesis technology to generate and play a response to the scammer. The device also informs the user of the risk of fraud and provides safety advice through screen displays and audio notifications.

[0799] This system protects users from the risk of fraud. The system performs sentiment analysis in the background and takes prompt action if fraud is suspected. After the call ends, users can review the report provided by the server and understand the detected sentiment and the reasons for the fraud assessment, which can help them take future preventative measures.

[0800] As a concrete example, consider a scenario where a user receives an unexpected sales call. In this case, if the emotion engine detects slight confusion and anxiety, the server determines that the sales call may be a scam. Based on this, the server generates sophisticated dialogue strategies to psychologically unsettle the scammer. The terminal synthesizes these strategies into speech and plays them back to the scammer, helping the user end the call before feeling anxious. In this example, the prompt message "Evaluate the likelihood of fraud based on the anxiety level detected from the user's voice and generate sophisticated dialogue strategies" is input into the generating AI model, providing the optimal solution.

[0801] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0802] Step 1:

[0803] The device acquires the user's voice data. A microphone is used to capture the user's voice during a call in digital format. The input is the user's raw voice, and this data is temporarily stored within the device. The output is digitized voice data.

[0804] Step 2:

[0805] The terminal streams the acquired audio data to the server. Wi-Fi or mobile data communication is used for this purpose. The input is digital audio data stored within the terminal, and the output is the process of that data being wirelessly transferred to the server. An efficient data transfer protocol is applied to ensure real-time performance during this process.

[0806] Step 3:

[0807] The server analyzes the received audio data using a speech analysis engine. The analysis extracts characteristics such as intonation, speed, and volume. The input is the digital audio data sent to the server, and the output is a dataset of these characteristics quantified. This dataset is then used for further emotion estimation.

[0808] Step 4:

[0809] The server uses an emotion engine to estimate emotions based on the results of speech analysis. Specifically, it estimates the speaker's emotional state (e.g., confusion, anxiety) using the emotion engine's model based on extracted features. The input is feature data generated by the speech analysis engine, and the output is the estimated emotion label.

[0810] Step 5:

[0811] The server assesses the likelihood of fraud based on estimated sentiment. Here, a generative AI model is used to compare sentiment data with past fraud patterns. The input is the sentiment label estimated by the sentiment engine, and the output is a fraud risk assessment index. If the assessment index is high, the process proceeds to the next step.

[0812] Step 6:

[0813] The server generates a response strategy for the fraudster when it assesses the likelihood of fraud. Specifically, it uses a generative AI model to generate a dialogue script from the prompt text to psychologically influence the fraudster. The input is an assessment of fraud risk, and the output is the generated dialogue script.

[0814] Step 7:

[0815] The terminal plays back the dialogue script received from the server using speech synthesis technology. The audio is played through the speaker. The input is the voice script sent from the server, and the output is a voice message directed at the fraudster. This process employs a psychological approach to the fraudster.

[0816] Step 8:

[0817] The device notifies the user of potential fraud and suggests safe countermeasures. Notifications are delivered via screen displays and additional audio messages. Input is the server's evaluation result, and output is a warning and suggested action to the user. This allows the user to make informed decisions and mitigate the risk of fraud.

[0818] (Application Example 2)

[0819] 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".

[0820] With the widespread adoption of electronic payments, the risk of users becoming victims of fraud is increasing. Therefore, there is a need for a system that analyzes user emotions in real time, quickly detects potential fraud, and provides appropriate responses. Furthermore, a mechanism to verify the security of transactions and immediately terminate risky transactions is also necessary.

[0821] 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.

[0822] In this invention, the server includes means for receiving voice information and immediately analyzing the speaker's emotions, means for detecting the possibility of fraud if the analyzed emotions express confusion or anxiety, and means for creating countermeasures and providing appropriate responses to the responder when the possibility of fraud is detected. This makes it possible for users to detect the risk of fraud more quickly and for a safer trading environment to be provided.

[0823] "Voice information" refers to digital data used to collect and analyze the content of users' phone calls in real time.

[0824] A "device that instantly analyzes the speaker's emotions" is a technology that quickly estimates the user's emotional state from received audio information.

[0825] A "device for detecting potential fraud" is a function that identifies potential fraud when suspicious signs are found in the analyzed emotions.

[0826] A "device for generating response measures" is a technology for generating specific actions and intervention methods to be taken when fraud is suspected.

[0827] A "device that provides appropriate responses to those involved" is a function that provides appropriate support and guidance to users and other stakeholders based on the policies that have been created.

[0828] "Payment communication" refers to the exchange of data that occurs during the process of conducting financial transactions and payments using electronic means.

[0829] A "device for verifying the security of transactions" is a technology that detects signs of fraudulent activity during settlement communications and performs a risk assessment in response.

[0830] A "mechanism to immediately terminate risky trades" is a process that automatically stops a trade when it is deemed to be high-risk, thereby preventing potential losses.

[0831] This invention relates to a system that uses voice analysis and sentiment analysis to reduce the risk of fraud during electronic payments. This system consists of a smartphone terminal and server software running in the background.

[0832] The server receives audio data and analyzes the audio information in real time. A speech analysis engine, such as Google Cloud Speech-to-Text, is used for the audio analysis. From the analyzed audio information, an emotion analysis engine instantly estimates the speaker's emotions and detects potential fraud based on the results. If fraud is detected, the server creates countermeasures and generates an appropriate response based on those measures.

[0833] The terminal plays this response as an audio message to notify the user. Furthermore, if a risky transaction occurs, the terminal has a function to automatically block the transaction and notify the user immediately to terminate it.

[0834] For example, if anxiety or confusion is detected from voice information when a user attempts to make an electronic payment, the server will immediately issue a warning, and the terminal will provide advice such as, "This transaction may be risky." Another example of a prompt message is, "I feel uneasy about this new online payment method, and I would like to confirm if this transaction is safe. Please let me know if the system has detected anything unusual." This prompt message forms the basis for the generative AI model to provide appropriate support.

[0835] In this way, the server and terminal work together to provide users with consistent support for fraud prevention.

[0836] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0837] Step 1:

[0838] The server receives audio data from the terminal. This audio data is a recording of the user's phone call and is transmitted in real time. Audio information is provided as input, and basic data for audio analysis is prepared as output.

[0839] Step 2:

[0840] The server converts the received audio data into text using Google Cloud Speech-to-Text. This process transforms the audio information into text data. The input is audio data, and the output is text data.

[0841] Step 3:

[0842] The server passes text data to an emotion analysis engine to estimate the user's emotions. The emotion analysis engine analyzes the text data to identify the user's emotional state (e.g., confusion, anxiety). The input is text data, and the output is emotional state data.

[0843] Step 4:

[0844] The server assesses the likelihood of fraud based on emotional state data. This assessment is performed by an algorithm that generates a fraud warning if confusion or anxiety exceeds a certain threshold. The input is emotional state data, and the output is fraud risk assessment data.

[0845] Step 5:

[0846] If fraud is detected, the server generates countermeasures. These measures include warning messages to the user and orders to suspend transactions. The input is data indicating fraud risk, and the output is data indicating the countermeasures.

[0847] Step 6:

[0848] The terminal receives policy data provided by the server and issues an audio warning to the user. Specifically, it converts text-based warning messages into speech using a speech synthesis system and plays them through the terminal's speaker. The input is policy data, and the output is an audio warning.

[0849] Step 7:

[0850] If a transaction is determined to be risky, the terminal automatically blocks the transaction and notifies the user of the result. This reduces the risk of the user engaging in fraudulent transactions. The input is the fraud risk assessment data, and the output is the transaction blocking command.

[0851] 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.

[0852] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0853] 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.

[0854] 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.

[0855] 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.

[0856] 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.

[0857] 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.

[0858] 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.

[0859] 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."

[0860] 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.

[0861] 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.

[0862] 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.

[0863] 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.

[0864] 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.

[0865] 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.

[0866] 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.

[0867] 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.

[0868] 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.

[0869] 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.

[0870] 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.

[0871] 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.

[0872] The following is further disclosed regarding the embodiments described above.

[0873] (Claim 1)

[0874] A means of receiving audio data and analyzing the speaker's emotions in real time,

[0875] A method for detecting suspected fraud when analyzed emotions indicate confusion or anxiety,

[0876] A means to generate a response strategy when fraud is suspected and to respond appropriately to the other party,

[0877] A method of generating and speaking dialogue scripts that are used to psychologically unsettle con artists,

[0878] A means of notifying users of the situation and providing safety advice,

[0879] A system that includes this.

[0880] (Claim 2)

[0881] The system according to claim 1, further comprising means for detecting typical phrases of fraud and assessing the risk level of fraud.

[0882] (Claim 3)

[0883] The system according to claim 1, further comprising means for generating and providing to the user a report after the end of a call, which includes the call history and the grounds for determining that the call was fraudulent.

[0884] "Example 1"

[0885] (Claim 1)

[0886] A means of receiving audio information and analyzing the speaker's emotional state in real time,

[0887] A means of detecting suspected fraud when the analyzed emotional state indicates confusion or anxiety,

[0888] A means to generate a response plan and take an appropriate response to the other party when fraud is suspected,

[0889] A method for generating and speaking dialogue procedures that are designed to psychologically unsettle con artists,

[0890] A means of notifying users of the situation and providing safety advice,

[0891] A natural language processing method that analyzes call information and assesses the risk of fraud,

[0892] A means of generating and providing to the user a report containing call records and evidence of fraud after the call has ended,

[0893] A system that includes this.

[0894] (Claim 2)

[0895] The system according to claim 1, further comprising means for detecting typical language patterns of fraud and assessing their risk.

[0896] (Claim 3)

[0897] The system according to claim 1, further comprising means for using an artificial intelligence model to carry out the generated dialogue procedure.

[0898] "Application Example 1"

[0899] (Claim 1)

[0900] A means of acquiring audio information and analyzing the speaker's emotions in real time,

[0901] A means of identifying the possibility of misconduct when analyzed emotions indicate confusion or anxiety,

[0902] A means of generating response methods and taking appropriate action against the other party when potential fraudulent activity is identified,

[0903] A means of generating and speaking dialogue instructions that are generated to psychologically unsettle the cheater,

[0904] A means of notifying users of the situation and providing safety advice,

[0905] By analyzing the content of calls and detecting phrases that indicate fraud or changes in the user's emotions, a means of presenting users with real-time safety alerts during calls is being developed.

[0906] A system that includes this.

[0907] (Claim 2)

[0908] The system according to claim 1, further comprising means for identifying phrases indicating fraudulent activity and for assessing the risk of fraudulent activity.

[0909] (Claim 3)

[0910] The system according to claim 1, further comprising means for generating a report after the end of a call, including the call history and the grounds for determining fraudulent activity, and presenting it to the user.

[0911] "Example 2 of combining an emotion engine"

[0912] (Claim 1)

[0913] A means for acquiring voice data and transmitting it to a data processing device via a communication device,

[0914] A means of analyzing received audio data in detail and estimating the speaker's emotions,

[0915] A means of assessing a high degree of suspicion of fraud when the estimated emotions indicate confusion or anxiety,

[0916] A means of dynamically generating response strategies based on the degree of suspicion of fraud and responding to fraudsters with a psychological effect,

[0917] A means of creating an optimal dialogue script using a generative AI model and playing it back using speech synthesis,

[0918] A means of notifying users of changes in the situation in real time and providing safety instructions,

[0919] A system that includes this.

[0920] (Claim 2)

[0921] The system according to claim 1, characterized by having means for automatically detecting typical phrases related to fraud and performing a fraud risk assessment.

[0922] (Claim 3)

[0923] The system according to claim 1, characterized in that it includes means for generating a report after the end of a call, including details of the conversation and the basis for the fraud assessment, and supplying this report to the user.

[0924] "Application example 2 when combining with an emotional engine"

[0925] (Claim 1)

[0926] A device that receives audio information and instantly analyzes the speaker's emotions,

[0927] A device that detects the possibility of fraud when the analyzed emotions express confusion or anxiety,

[0928] A device that creates countermeasures when potential fraud is detected and provides appropriate responses to those responding,

[0929] A device that synthesizes and speaks dialogue scenarios generated to psychologically unsettle cheaters,

[0930] A device that notifies users of the situation and provides security advice,

[0931] A device that evaluates voice data during payment communication to confirm the security of the transaction,

[0932] A device that terminates transactions when the risk is high and blocks malicious transactions,

[0933] A system that includes this.

[0934] (Claim 2)

[0935] The system according to claim 1, further comprising a device for detecting common fraudulent phrases and assessing the risk of fraud.

[0936] (Claim 3)

[0937] The system according to claim 1, further comprising a device that, after the termination of communication, generates and provides to the user a report containing the communication history and the grounds for determining fraud. [Explanation of Symbols]

[0938] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of receiving audio data and analyzing the speaker's emotions in real time, A method for detecting suspected fraud when analyzed emotions indicate confusion or anxiety, A means to generate a response strategy when fraud is suspected and to respond appropriately to the other party, A method of generating and speaking dialogue scripts that are used to psychologically unsettle con artists, A means of notifying users of the situation and providing safety advice, A system that includes this.

2. The system according to claim 1, further comprising means for detecting typical phrases of fraud and assessing the risk level of fraud.

3. The system according to claim 1, further comprising means for generating and providing to the user a report after the end of a call, which includes the call history and the grounds for determining that the call was fraudulent.

Images