System

Abstract

A system is provided.SOLUTION: The system includes a means for monitoring a communication line to be monitored, a means for converting voice data into text data by using a voice recognition technique, an artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, a means for notifying a person concerned when the possibility of fraud is high, and a means for instructing to temporarily lock a corresponding bank account.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the 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] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] In recent years, the number of fraudulent telephone calls has been increasing, and refund fraud, special fraud, and fictitious billing fraud targeting the elderly and young people have become serious problems. Such fraudulent acts not only cause victims significant economic losses, but also psychological stress. Conventional countermeasures have difficulty detecting and preventing fraudulent telephone calls in real time, and an effective method is needed. Therefore, the objective of this invention is to provide a system that detects fraudulent telephone calls in real time and notifies relevant parties to prevent damage before it occurs. [Means for solving the problem]

[0005] To solve this problem, the present invention provides the following means. First, it provides a means for monitoring the communication lines to be monitored. Then, it adds an artificial intelligence means for converting call voice data into text data using voice recognition technology and analyzing the converted text data. If the analysis determines that there is a high possibility of fraud, it creates a system that includes a means for notifying the relevant parties and a means for issuing an instruction to temporarily lock the relevant bank account. It also adds an interface means for performing identity verification procedures and a means for analyzing abnormal voice patterns in real time, providing more effective fraud prevention measures.

[0006] A "communication line" is a physical or wireless path for transmitting and receiving voice and data.

[0007] "Speech recognition technology" is a technology that allows a computer to understand human speech and convert it into text data.

[0008] "Text data" is data in the form of written information converted from speech using speech recognition technology.

[0009] "Artificial intelligence" is the technology that enables computer systems to mimic human intelligence by learning, reasoning, and self-correcting.

[0010] "Analysis" is the act of analyzing data or events to extract specific patterns or meanings.

[0011] "Temporarily locking a bank account" is a measure to temporarily suspend transactions in a specific account and prevent the transfer of funds.

[0012] "Notification means" refers to the method or device by which the system sends information or warnings to interested parties.

[0013] "Identity verification procedures" are procedures that users go through to prove their identity.

[0014] An "interface" is a method or device by which a user interacts with and operates a system.

[0015] An "anomalous voice pattern" is a voice pattern that differs from the content of a normal call and may indicate fraud or dangerous activity.

[0016] "Real-time" refers to a process or operation that occurs in the present time, without delay. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0018] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0019] First, the terms used in the following description will be explained.

[0020] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0021] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0022] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0023] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0024] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0025] [First embodiment]

[0026] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0027] 1, a 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.

[0028] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0029] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0030] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the 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.

[0031] 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 of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0032] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0033] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 2, in the data processing device 12, a specific process 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" according to the technology of the present 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 process 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.

[0035] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0036] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0037] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0038] The present invention is a system that monitors communication lines to be monitored, detects calls that are likely to be fraudulent, and prevents damage before they occur. This system is mainly composed of a server, a terminal, and a user, and operates as follows.

[0039] Audio data collection and analysis

[0040] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence means to detect abnormal patterns and fraud-related keywords in the text.

[0041] Fraud Assessment and Notification

[0042] After conversion and analysis, the server evaluates the likelihood of fraud. If it determines that there is a high likelihood of fraud, the server notifies the relevant user and relevant parties. The notification is sent via phone, SMS, email, etc., to provide a comprehensive warning.

[0043] Temporary locking of bank accounts

[0044] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from being leaked to the fraudsters.

[0045] Identity verification procedures

[0046] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a verification code entry screen or facial recognition via video call. After successful identity verification, the server unlocks the bank account.

[0047] Specific examples

[0048] Scenario: Senior Citizen Refund Fraud

[0049] 1. Start a call

[0050] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[0051] 2. Collection and analysis of audio data

[0052] The device transmits the audio data of the fraudulent call to the server in real time.

[0053] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[0054] 3. Fraud Assessment and Notification

[0055] The server detects keywords such as "refund," "bank," and "transfer" and determines that there is a high possibility of fraud.

[0056] The server immediately notifies Mr. A and his family of the possible fraud.

[0057] 4. Temporarily locking your bank account

[0058] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[0059] 5. Identity Verification Procedures

[0060] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[0061] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[0062] After verification, the server contacts the bank and unlocks Mr. A's bank account.

[0063] This invention makes it possible to detect fraudulent phone calls to elderly and young people in real time, and to prevent damage by quickly notifying the relevant parties. It also ensures the safety of users' funds by locking their bank accounts and implementing identity verification procedures.

[0064] The processing flow will be explained below.

[0065] Step 1:

[0066] The terminal detects that the user has started a telephone call, and simultaneously starts collecting the call voice data.

[0067] Step 2:

[0068] The terminal transmits the collected call voice data to the server in real time.

[0069] Step 3:

[0070] The server passes the received voice data to a voice recognition engine and converts the voice data into text data.

[0071] Step 4:

[0072] The server passes the converted text data to an AI analysis module to detect abnormal patterns and fraud-related keywords within the text data.

[0073] Step 5:

[0074] The server will raise a fraud warning flag if AI analysis determines that fraud is likely.

[0075] Step 6:

[0076] If a fraud alert flag is raised, the server will immediately notify the user and relevant parties via phone, SMS, email, or other means.

[0077] Step 7:

[0078] The server instructs the user's designated bank to temporarily lock the user's bank account.

[0079] Step 8:

[0080] The user receives the notification and starts the identity verification procedure according to the specified procedure. The terminal provides an interface for identity verification.

[0081] Step 9:

[0082] The device sends the verification code entered by the user and the facial recognition results obtained via video call to the server.

[0083] Step 10:

[0084] The server receives the authentication information and verifies that the user is a legitimate user.

[0085] Step 11:

[0086] If the verification is successful, the server instructs the bank to unlock the user's bank account.

[0087] This series of processes allows for the real-time detection of potentially fraudulent calls and prompt notification and response.

[0088] Example 1

[0089] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0090] In recent years, fraud methods using telecommunications lines have become more sophisticated and diverse. Refund fraud and bank transfer fraud targeting the elderly in particular have been on the rise, causing serious damage. Conventional crime prevention methods have struggled to detect fraudulent activity in real time and prevent damage before it occurs. There is also a lack of methods for quickly and accurately assessing the possibility of fraud and notifying relevant parties. In addition, to prevent fraud once it has occurred, there is a need for a function to quickly lock financial institution accounts and a method for users to easily complete identity verification procedures.

[0091] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0092] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties if a fraudulent call is highly likely, means for issuing instructions to temporarily lock the relevant financial institution's account, means for unlocking the account by providing personal identification information, means for transmitting voice data in real time when the call is made, means for detecting fraud keywords based on the analysis results, and means for transmitting notifications using multiple communication means. This makes it possible to detect potentially fraudulent calls in real time and prevent damage before it occurs. Furthermore, the relevant financial institution's account can be immediately locked and safely unlocked through personal identification procedures.

[0093] A "monitored communication line" is a communication network that is constantly monitored to detect potentially fraudulent calls.

[0094] "Speech recognition technology" is a technology used to convert voice data into text data.

[0095] "Voice data" means voice information, in analog or digital form, captured during a telephone call.

[0096] "Text data" is character information converted from voice data using voice recognition technology.

[0097] "Artificial intelligence means" are machine learning or deep learning algorithms used to analyze text data and detect specific patterns or keywords.

[0098] "Fraudulent keywords" are words or phrases that may be associated with fraudulent activity.

[0099] "Assessing the likelihood of fraud" means determining the risk of fraud based on pattern and keyword analysis within text data.

[0100] "Notifying" means sending warnings and information to users and other interested parties when fraud is likely.

[0101] "Interested Persons" are people or entities that should receive information about fraud risks.

[0102] A "financial institution account," or simply an "account," is an account managed by a financial institution such as a bank that records a user's deposits and funds.

[0103] "Temporarily locking" means suspending activities such as withdrawals from a financial institution account for a certain period of time.

[0104] "Personal identification information" refers to information or data used to prove the identity of a user.

[0105] "Unlocking" means lifting the suspension of an account at a financial institution after identity verification has been completed.

[0106] "Sending in real time when a call is made" means sending voice data to the server immediately at the moment the call is started.

[0107] "Detecting fraud keywords based on the analysis results" means that the artificial intelligence means analyzes the text data and identifies specific words related to fraudulent activities.

[0108] "Sending notifications using multiple communication methods" means sending notifications using different media such as telephone, SMS, and email.

[0109] The present invention relates to a system for monitoring communication lines to detect calls that are likely to be fraudulent and prevent damage before they occur. This system uses a server, a terminal, and a user as its main components.

[0110] The server uses high-performance communication monitoring software (such as WireShark or Snort) to constantly monitor the communication line. When a call is initiated, the device sends the call audio data to the server in real time. The audio data is encoded using VoIP technology (such as the SIP protocol) and sent via a secure communication protocol such as HTTPS.

[0111] The server converts the received voice data into text using the Google Cloud Speech-to-Text API. It then uses speech recognition technology to obtain accurate text data, and analyzes the converted text data using generative AI models such as BERT and GPT-3. The analysis aims to detect abnormal patterns and fraud-related keywords in the text (e.g., "refund," "bank," and "transfer").

[0112] If a fraud risk is determined to be high, the server will notify the relevant parties (e.g., the user and their family) via SMS or phone using the Twilio API. It is also possible to notify them by email using the SendGrid API. This notification allows users to quickly be informed of the risk of fraud.

[0113] After sending the fraud notification, the server uses the API of the partner financial institution (e.g., Plaid API) to issue a command to temporarily lock the user's financial institution account, thereby preventing the user's funds from being stolen due to fraudulent activity.

[0114] When the user receives the fraud notification, they use the provided interface (e.g., a verification code entry screen or video call) to verify their identity. The device provides the interface for identity verification through the API of Jumio or IDNow. The user completes identity verification by entering a verification code or by performing facial recognition via a video call. Once verification is complete, the server again uses the financial institution's API to send instructions to unlock the bank account.

[0115] Specific examples

[0116] Scenario: Senior Citizen Refund Fraud

[0117] 1. Start a call

[0118] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[0119] 2. Collection and analysis of audio data

[0120] The device (smartphone) transmits the audio data of the fraudulent call to the server in real time.

[0121] The server converts the received voice data into text data using Google Cloud Speech-to-Text.

[0122] The converted text data is analyzed using the BERT model to detect keywords such as "refund," "bank," and "transfer."

[0123] 3. Fraud Assessment and Notification

[0124] Based on the keywords detected, the server determines with a high probability that the request is fraudulent.

[0125] The server uses the Twilio API to notify Mr. A and his family via SMS of the possible fraud.

[0126] 4. Temporary locking of financial institution accounts

[0127] The server sends an instruction to the partner bank via API to temporarily lock Mr. A's financial institution account.

[0128] 5. Identity Verification Procedures

[0129] The user, Mr. A, or his / her family member clicks on the link provided in the notification to begin the identity verification process.

[0130] Enter the verification code in the interface provided by your device or use facial recognition via video call.

[0131] After the verification is complete, the server again uses the API to send instructions to unlock the bank account.

[0132] Example prompts for generative AI models

[0133] Using the following system scenario, please explain in natural language how a program to detect potentially fraudulent calls works:

[0134] system:

[0135] The server monitors the communication line and analyzes the voice data.

[0136] The terminal transmits the call voice data to the server.

[0137] The server uses artificial intelligence to analyze the converted text data and assess the likelihood of fraud.

[0138] If fraud is likely, the server will notify the user and lock the financial institution account.

[0139] The user goes through the identity verification procedure and the server unlocks the device.

[0140] Example: Senior Citizen Refund Fraud Scenario

[0141] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0142] Specific explanation of the program based on processing steps

[0143] Step 1:

[0144] Call detection

[0145] The terminal monitors the target communication line and detects when a call has started. As input, it receives a call signal from the communication line and analyzes this signal to detect the call start trigger. The output is event information that a call has started, which starts the subsequent processing.

[0146] Specific operation: When elderly person A answers the phone, the device's monitoring app detects the start of the call and immediately begins collecting voice data.

[0147] Step 2:

[0148] Collection and transmission of voice data

[0149] The device receives call start event information and collects real-time call audio data, which is then encoded using VoIP technology and sent to the server via a secure communication protocol (HTTPS).

[0150] As input, it receives real-time audio data acquired by the terminal, and as output, it transmits encoded audio data to the server.

[0151] Specific operation: While Person A continues the conversation, the device captures audio data and transmits it to the server through a secure channel.

[0152] Step 3:

[0153] Converting audio data to text

[0154] The server receives the voice data sent from the device and converts it into text data using the Google Cloud Speech-to-Text API. The input is the encoded voice data, and the output is the converted text data.

[0155] Specific operation: The server calls the Google Cloud Speech-to-Text API to convert the audio data into text data.

[0156] Step 4:

[0157] Text data analysis

[0158] The server receives text data from the Google Cloud Speech-to-Text API and analyzes it using artificial intelligence models such as BERT and GPT-3. The analysis detects keywords and anomalous patterns that may be indicative of fraud. The output is information about the detected fraud-related keywords and anomalous patterns.

[0159] Specific operation: The server inputs the text data into an artificial intelligence model and detects keywords such as "refund," "bank," and "transfer."

[0160] Step 5:

[0161] Fraud Assessment and Notification

[0162] The server evaluates the likelihood of fraud based on the results of text data analysis. It receives keyword detection results as input, scores the likelihood of fraud, and outputs a notification based on the evaluation results. It notifies users and relevant parties via SMS, phone, or email using the Twilio API, or the SendGrid API.

[0163] Specific operation: If the analysis results indicate a high risk of fraud, the server uses the Twilio API to send an SMS warning to Mr. A and his family.

[0164] Step 6:

[0165] Temporary locking of financial institution accounts

[0166] Based on the fraud risk assessment results, the server uses the API (such as Plaid API) of the partner financial institution to send an instruction to temporarily lock the user's financial institution account. The input receives the fraud risk assessment result, and the output is the completion of the lock instruction.

[0167] Specific operation: If it is determined that there is a high possibility of fraud, the server will generate an API request to immediately temporarily lock Mr. A's financial institution account.

[0168] Step 7:

[0169] Identity verification procedures

[0170] The user follows the notification to complete the identity verification procedure. Using the provided interface (verification code entry screen or video call), the user sends the verification information from the device to the server. The server receives the verification code entered by the user and facial recognition data as input, and the output is the completion of identity verification.

[0171] Specific actions: Person A or his / her family follows the instructions in the notification, enters the verification code on the device's browser, and sends it to the server.

[0172] Step 8:

[0173] Unlock

[0174] After the identity verification is successful, the server again uses the financial institution's API to send an unlock instruction. The input is the successful identity verification information, and the output is the unlock of the financial institution account.

[0175] Specific operation: Once identity verification is complete, the server contacts the financial institution via API and unlocks Mr. A's account.

[0176] This process flow allows for real-time detection of potentially fraudulent calls, preventing damage before it occurs, and providing rapid response to protect users' financial assets.

[0177] (Application example 1)

[0178] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0179] In modern society, voice-based fraud is on the rise, and fraud victimization of elderly people is becoming particularly serious. There is also a need for a method to monitor fraudulent call content in real time and respond quickly, but current technology makes this difficult to achieve. With current systems, it takes time for users receiving fraudulent calls to realize they are being scammed, making it difficult to prevent fraud before it occurs. To solve this problem, a system is needed that can assess the possibility of fraud in real time and respond quickly.

[0180] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0181] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties when there is a high possibility of fraud, means for issuing an instruction to temporarily lock the relevant bank account, means for notifying the user of a fraud warning in real time based on the results of the voice recognition and analysis, communication means for monitoring voice call data that is likely to be fraudulent and sending the notification, means for evaluating abnormal voice patterns using a generative AI model, and means for generating prompt sentences to be input to the generative AI model. This makes it possible to monitor the communication lines to be monitored in real time, quickly evaluate and respond to the possibility of fraud, and immediately notify the user and relevant parties.

[0182] 1. "Monitored communication lines" refer to lines that carry communications such as telephone and internet that are subject to monitoring.

[0183] 2. "Monitoring means" means devices or programs that constantly monitor communication lines and collect data.

[0184] 3. "Speech recognition technology" refers to technology that converts voice data into text data.

[0185] 4. "Voice data" refers to sound information collected through telephone conversations, recordings, etc.

[0186] 5. "Text data" means data that has been converted from audio data into text information.

[0187] 6. "Artificial intelligence means" means systems or software that use machine learning or natural language processing to perform specific tasks.

[0188] 7. "Assessing the likelihood of fraud" means analyzing voice and text data to determine whether there is a possibility of fraud.

[0189] 8. "Means of notification" means the means (e.g., telephone, SMS, email, etc.) used to send warnings or information to interested parties.

[0190] 9. "Means for issuing instructions to temporarily lock a bank account" means a means for sending an instruction to an affiliated financial institution to temporarily suspend the use of a user's account.

[0191] 10. "Means for providing real-time fraud alerts to users based on speech recognition and analysis results" means means for instantly sending alerts to users based on analysis results from speech recognition technology and artificial intelligence means.

[0192] 11. "Communication means for monitoring voice call data that may be fraudulent and sending notifications" refers to communication devices or programs that constantly monitor call data that may be fraudulent and notify the results.

[0193] 12. "Means for assessing anomalous speech patterns using a generative AI model" means means for identifying and assessing anomalous speech patterns using a generative AI model.

[0194] 13. "Means for generating prompt sentences to be input to a generative AI model" means means for creating initial input sentences to provide appropriate analytical instructions to a generative AI model.

[0195] The system for implementing this invention mainly comprises a server, a terminal, and a user. The server constantly monitors the communication line to be monitored and receives call voice data transmitted from the terminal in real time. The server converts the received voice data into text data using speech recognition technology. The converted text data is then analyzed using artificial intelligence to detect abnormal patterns and fraud-related keywords in the text.

[0196] Audio data collection and analysis

[0197] The server constantly monitors the target communication line. The device sends the call audio data to the server in real time when the call starts. The server converts this audio data into text data using speech recognition technology (e.g., Google Speech Recognition API). This text data is then analyzed using artificial intelligence means (e.g., machine learning algorithms).

[0198] Fraud Assessment and Notification

[0199] Once the conversion and analysis are complete, the server evaluates the likelihood of fraud. If it determines that there is a high likelihood of fraud, the server notifies the user and relevant parties. Notifications are sent with high speed via phone, SMS, email, etc. The server also issues instructions to temporarily lock the relevant bank account.

[0200] Temporary locking of bank accounts

[0201] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from being leaked to the fraudsters.

[0202] Identity verification procedures

[0203] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a verification code entry screen or facial recognition via video call. After successful identity verification, the server unlocks the bank account.

[0204] Evaluation by generative AI model

[0205] Based on the results of speech recognition and analysis, the server evaluates abnormal speech patterns using a generative AI model, to which an appropriate prompt is input. For example, the following prompt might be used:

[0206] Prompt text: Convert the voice data into text and detect the keywords "refund," "bank," or "transfer" to determine the possibility of fraud. If there are signs of fraud, send a warning to the user via email.

[0207] Specific examples

[0208] Below are some specific scenarios where this invention can be used:

[0209] Preventing refund fraud among the elderly

[0210] 1. Start a call:

[0211] An elderly person, Mr. A, receives a fraudulent phone call and is asked for "information regarding receiving a refund."

[0212] 2. Audio data collection and analysis:

[0213] The device transmits the voice data of the fraudulent call in real time to a server, which uses voice recognition technology to convert the voice data into text data, which is then analyzed using artificial intelligence means.

[0214] 3. Potential Fraud Assessment and Notification:

[0215] The server detects keywords such as "refund," "bank," and "transfer," and determines that there is a high possibility of fraud. The server immediately notifies Mr. A and his family of the possibility of fraud.

[0216] 4. Temporarily locking your bank account:

[0217] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[0218] 5. Identity Verification Procedure:

[0219] The user, Mr. A, or his / her family member, confirms the notification and begins the identity verification process. They enter a verification code using the interface provided by the device, and the server receives the information. After verification, the server contacts the bank and unlocks Mr. A's bank account.

[0220] As described above, this invention allows for real-time monitoring of potentially fraudulent calls and prompt response. This makes it possible to prevent fraudulent calls from affecting both the elderly and young. Furthermore, the safety of users' funds can be ensured by locking their bank accounts and implementing identity verification procedures.

[0221] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0222] Step 1:

[0223] Initiating a call and collecting audio data

[0224] When a terminal initiates a user's call, voice data is collected in real time. This voice data is sent by the terminal to a server. The input is the user's call voice, and the output is the voice data sent to the server. The specific operations include the terminal capturing the voice of the call and sending the data to the server.

[0225] Step 2:

[0226] Converting audio data to text

[0227] The server converts the received voice data into text data using voice recognition technology (e.g., Google Speech Recognition API). The input is the voice data sent from the device, and the output is text data. Specific operations include calling the voice recognition API and converting the voice data into text data.

[0228] Step 3:

[0229] Text data analysis

[0230] The server analyzes the text data using artificial intelligence means (e.g., machine learning algorithms). The input is the text data and the output is the analysis results. Specific operations include using machine learning models to detect anomalous patterns and fraud-related keywords in the text data.

[0231] Step 4:

[0232] Fraud likelihood assessment

[0233] The server evaluates the possibility of fraud based on the analysis results. The input is the analysis results, and the output is the fraud possibility evaluation result. Specific operations include scoring the possibility of fraud based on the analysis results and determining whether the score exceeds a threshold.

[0234] Step 5:

[0235] Sending notifications

[0236] If it determines that there is a high possibility of fraud, the server notifies the user and relevant parties. The input is the fraud possibility assessment result, and the output is a notification message. Specific operations include sending warnings to the user and relevant parties via email, SMS, phone, etc.

[0237] Step 6:

[0238] Temporary locking of bank accounts

[0239] The server issues an instruction to temporarily lock the relevant bank account. The input is the fraud probability assessment result and user information, and the output is a lock instruction to be sent to the bank. Specific operations include sending a request to temporarily lock the account to the partner bank's API.

[0240] Step 7:

[0241] Start the identity verification process

[0242] Upon receiving the notification, the user begins the identity verification process. The device then provides a screen for entering a verification code or facial recognition via video call. The input is the notification message and user information, and the output is progress information on identity verification. Specific actions include the user entering a verification code or performing facial recognition via video call.

[0243] Step 8:

[0244] Unlocking a bank account

[0245] After successful identity verification, the server issues a command to unlock the bank account. The input is the successful identity verification information, and the output is an unlock command sent to the bank. Specific operations include sending an unlock request to the partner bank's API.

[0246] Step 9:

[0247] Analyzing abnormal voice patterns using AI models

[0248] The server analyzes abnormal voice patterns using a generative AI model. The input is the voice recognition and analysis results, and the output is the evaluation result of the abnormal voice pattern. Specific operations include inputting appropriate prompt sentences to the generative AI model and evaluating the abnormal voice pattern.

[0249] Step 10:

[0250] Generate prompt statement

[0251] The server generates prompt sentences to be input to the generative AI model. The inputs are the AI ​​model specifications and the information to be analyzed, and the output is the prompt sentence. Specific operations include creating prompt sentences appropriate to the situation to be analyzed (e.g., "Convert the voice data into text and detect any of the keywords 'refund,' 'bank,' or 'transfer' to determine the possibility of fraud. If there are signs of fraud, send a warning to the user via email.").

[0252] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0253] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing the user's emotions. This system mainly consists of a server, a terminal, and a user, and operates as follows.

[0254] Audio data collection and analysis

[0255] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence tools to detect abnormal patterns and fraud-related keywords.

[0256] Emotion recognition by emotion engine

[0257] The server then passes the call audio data to an emotion engine to recognize the user's emotions. This emotion engine assesses whether the user is experiencing stress or anxiety. The user's emotion data is then used to augment the assessment of the likelihood of fraud.

[0258] Fraud Assessment and Notification

[0259] If the analysis determines that there is a high possibility of fraud, the server will raise a fraud warning flag and immediately notify the user and relevant parties via phone, SMS, email, etc. to quickly warn them.

[0260] Temporary locking of bank accounts

[0261] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from reaching the fraudsters.

[0262] Identity verification procedures

[0263] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a screen for entering a verification code or facial recognition via video call. After successfully verifying the user's identity, the server unlocks the bank account.

[0264] Specific examples

[0265] Scenario: Senior Citizen Refund Fraud

[0266] 1. Start a call

[0267] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[0268] 2. Collection and analysis of audio data

[0269] The device transmits the audio data of the fraudulent call to the server in real time.

[0270] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[0271] At the same time, the server uses an emotion engine to analyze Mr. A's emotional state and assess his level of stress and anxiety.

[0272] 3. Fraud Assessment and Notification

[0273] The server detects keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on the results of AI analysis and emotional data.

[0274] The server immediately notifies Mr. A and his family of the possible fraud.

[0275] 4. Temporarily locking your bank account

[0276] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[0277] 5. Identity Verification Procedures

[0278] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[0279] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[0280] After verification, the server contacts the bank and unlocks Mr. A's bank account.

[0281] This invention makes it possible to detect fraudulent phone calls suffered by the elderly and young in real time, and by utilizing user emotion data, it is possible to improve the accuracy of fraud detection and prevent damage by quickly notifying relevant parties.In addition, the safety of users' funds can be ensured by locking bank accounts and implementing identity verification procedures.

[0282] The processing flow will be explained below.

[0283] Step 1:

[0284] The terminal detects that the user has started a telephone call, and simultaneously starts collecting the call voice data.

[0285] Step 2:

[0286] The terminal transmits the collected call voice data to the server in real time.

[0287] Step 3:

[0288] The server passes the received voice data to a voice recognition engine and converts the voice data into text data.

[0289] Step 4:

[0290] The server passes the converted text data to an AI analysis module to detect abnormal patterns and fraud-related keywords within the text data.

[0291] Step 5:

[0292] At the same time, the server passes the voice data to the emotion engine, which analyzes the user's emotions by analyzing the tone, tempo, and speaking style of the voice to assess whether the user is feeling stressed or anxious.

[0293] Step 6:

[0294] The server makes a comprehensive judgment based on the results of the AI ​​analysis and the evaluation results of the emotion engine, and if there is a high possibility of fraud, it raises a fraud warning flag.

[0295] Step 7:

[0296] If a fraud alert flag is raised, the server will immediately notify the user and relevant parties via phone, SMS, email, or other means.

[0297] Step 8:

[0298] The server instructs the user's designated bank to temporarily lock the user's bank account.

[0299] Step 9:

[0300] The user receives the notification and starts the identity verification procedure according to the specified procedure. The terminal provides an interface for identity verification.

[0301] Step 10:

[0302] The device sends the verification code entered by the user and the facial recognition results obtained via video call to the server.

[0303] Step 11:

[0304] The server receives the authentication information and verifies that the user is a legitimate user.

[0305] Step 12:

[0306] If the verification is successful, the server instructs the bank to unlock the user's bank account.

[0307] This process allows for real-time detection of potentially fraudulent calls, and prompt notification and response. The emotion engine also enables advanced fraud detection that takes into account the user's emotional state.

[0308] Example 2

[0309] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0310] Conventional communication monitoring systems simply convert voice data into text data and detect fraud through simple keyword matching, but this method is insufficient to counter increasingly sophisticated fraud techniques. Furthermore, because the system ignores the user's emotions and psychological state, the accuracy of fraud detection can be reduced. Furthermore, it is difficult to respond quickly after fraud is detected, making it difficult to effectively prevent users from losing funds.

[0311] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0312] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, emotion recognition means for analyzing the call voice data and recognizing the user's emotion, means for notifying relevant parties when there is a high possibility of fraud, means for issuing instructions to temporarily lock the relevant bank account, and interface means for carrying out identity verification procedures. This enables highly sophisticated fraud to be detected with high accuracy, and by utilizing user emotion data, the accuracy of fraud detection is improved, enabling quick and effective response.

[0313] A "monitored communication line" is a communication infrastructure that is monitored and through which information is sent and received.

[0314] "Speech recognition technology" is a technology for converting voice data into text data, and typically uses an algorithm to extract linguistic information from voice.

[0315] "Voice data" refers to data that is a digital recording of the voice generated during a call.

[0316] "Text data" refers to the format of character information obtained after voice data is converted using voice recognition technology.

[0317] "Artificial intelligence means" means artificial intelligence-based techniques for analyzing the converted text data and assessing any abnormal patterns or potential fraud.

[0318] The "emotion recognition means" is a system that analyzes call voice data and evaluates the user's emotions and psychological state.

[0319] "Notification means" refers to methods or tools used to warn or inform relevant parties when a high possibility of fraud is determined.

[0320] A "bank account locking mechanism" is a system that executes instructions to temporarily freeze a user's bank account if fraud is suspected.

[0321] "Interface means" refers to input and output means for users and related parties to interact with the system, and includes, for example, a confirmation code input screen and a function for video calling.

[0322] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing user emotions. This system mainly consists of three components: a server, a terminal, and a user. Specific embodiments of the system are described below.

[0323] Audio data collection and analysis

[0324] The server constantly monitors the target communication line. Network monitoring software (e.g., Wireshark or PRTG) is used for monitoring. When a call is initiated, the device sends the call audio data to the server in real time. A streaming API (e.g., WebRTC) is used to collect the audio data, minimizing the delay in the audio data.

[0325] The server converts the received voice data into text data using the Google Cloud Speech-to-Text API. This conversion process is performed in parallel, and text is generated in real time.

[0326] The server then analyzes the converted text data using natural language processing (NLP) techniques such as GPT-3, which can identify unusual patterns and fraud-related keywords (e.g., "refund," "bank," "transfer," etc.).

[0327] Emotion recognition by emotion engine

[0328] The server then passes the call audio data to an emotion recognition engine (Emotion AI) to recognize the user's emotions. The emotion engine analyzes the tone and pace of the voice to assess whether the user is feeling stressed or anxious. For example, a trembling voice or a high-pitched tone can be detected as a sign of stress.

[0329] Fraud Assessment and Notification

[0330] The server combines the analysis results with sentiment data to assess the likelihood of fraud, and if it determines that fraud is likely, it raises a fraud warning flag.

[0331] If a potential fraud is detected, the server immediately notifies the user and relevant parties. The notification is sent in real time via SMS or email, allowing the user and relevant parties to take immediate action. For example, the notification may say, "A potential fraudulent call has been detected. Please take action immediately."

[0332] Temporary locking of bank accounts

[0333] After the fraud notification is complete, the server sends an instruction to the partner financial institution via API to temporarily lock the user's bank account, thereby preventing the outflow of funds.

[0334] Identity verification procedures

[0335] Upon receiving the notification, the user begins the identity verification process. The device provides facial recognition via a verification code input screen or video call. For example, the user receives a verification code via SMS and enters it on the device's input screen. Alternatively, facial recognition is performed via a video call to verify the user's identity. The server then contacts the financial institution and unlocks the bank account.

[0336] Specific examples

[0337] For example, consider the case of an elderly person named Mr. A who receives a call about a tax refund scam. The moment the call starts, the device sends a call start message to the server via the API, and captures and transmits voice data in real time.

[0338] The server converts the voice data into text using the Google Cloud Speech-to-Text API and analyzes the text using GPT-3. At the same time, it uses an emotion engine to analyze Mr. A's emotional state, detecting that he is experiencing increased stress and anxiety.

[0339] The server detects fraud-related keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on AI analysis and emotional data. It immediately notifies Mr. A and his family of the possibility of fraud and instructs the financial institution to temporarily lock Mr. A's bank account.

[0340] Upon receiving the notification, Mr. A or his family members can use the device interface to complete identity verification procedures, and once verification is complete, the bank account will be unlocked.

[0341] Prompt Sentence Examples

[0342] "They mentioned a refund, is this a scam?"

[0343] "How can I identify a fraudulent phone call?"

[0344] "Please let me know the analysis results of calls that are likely to be fraudulent."

[0345] The present invention makes it possible to detect fraudulent phone call damage in real time, improve the accuracy of fraud detection by utilizing user emotion data, and respond quickly.

[0346] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0347] Step 1:

[0348] The server constantly monitors the target communication line. The input is the connection information of the communication line, and the output is the call start trigger. Specifically, it uses network monitoring software (e.g., Wireshark or PRTG) to detect the start of a call in real time.

[0349] Step 2:

[0350] When a call is started, the device notifies the server of the call start information. The input is a call start event, and the output is a notification message to the server. Specifically, the device sends a call start message to the server through the API. This message contains the call identification information.

[0351] Step 3:

[0352] The device captures the call audio data in real time and sends it to the server. The input is the call audio data, and the output is a stream of audio data to the server. Specifically, it uses a streaming API (e.g., WebRTC) to capture the audio data and sends it to the server in an encrypted format.

[0353] Step 4:

[0354] The server converts the received voice data into text data using the Google Cloud Speech-to-Text API. The input is voice data and the output is text data. Specifically, the server sends voice data to the API and generates text data in real time. Parallel processing is used to improve conversion efficiency.

[0355] Step 5:

[0356] The server analyzes the converted text data using natural language processing technology such as GPT-3. The input is text data, and the output is the analysis results. Specifically, the text data is input into a generative AI model to detect fraud-related keywords such as "refund," "bank," and "transfer."

[0357] Step 6:

[0358] The server passes the call voice data to the Emotion AI engine, which analyzes the user's emotions. The input is voice data, and the output is emotional evaluation data. Specifically, it analyzes the voice tone and pace to evaluate whether the user is feeling stressed or anxious.

[0359] Step 7:

[0360] The server integrates the analysis results and emotion data to assess the likelihood of fraud. The input is the text analysis results and emotion evaluation data, and the output is a fraud warning flag. Specifically, it determines the likelihood of fraud based on abnormal patterns, fraud-related keywords, and user emotion data.

[0361] Step 8:

[0362] If it is determined that there is a high possibility of fraud, the server will notify the relevant parties. The input is a fraud warning flag and the output is a notification message. Specifically, a notification is sent to the relevant parties via SMS or email. The message sent will say something like, "A call with a high possibility of fraud has been detected. Please take immediate action."

[0363] Step 9:

[0364] After the fraud notification is complete, the server issues an instruction to the partner financial institution to temporarily lock the user's bank account. The input is the fraud warning flag and the user's bank information, and the output is a lock instruction to the bank. Specifically, the server sends the lock instruction to the bank via API to ensure the safety of the user's funds.

[0365] Step 10:

[0366] Upon receiving the notification, the user begins the identity verification procedure. The input is the notification message, and the output is the identity verification result. Specifically, the procedure is carried out using the verification code input screen or video call function provided by the device. After verification is complete, the server contacts the bank and unlocks the bank account.

[0367] (Application example 2)

[0368] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0369] In recent years, the number of victims of fraudulent phone calls targeting the elderly and young people has been increasing. These fraudulent acts have become more diverse, making it difficult to deal with them using conventional fraud detection methods alone. Furthermore, since the damage often occurs before the user even realizes it, new technologies are needed to respond to fraud early.

[0370] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties when there is a high possibility of fraud, means for issuing an instruction to temporarily lock the relevant trading account, and emotion recognition means for evaluating the user's emotional state. This enables highly accurate fraud detection and rapid response that takes the user's emotional state into account.

[0371] A "monitored communication line" is a line on a communication network that is continuously monitored to monitor activity on specific calls or data traffic.

[0372] "Voice recognition technology" is a technology that captures voice as a digital signal and converts it into text data or commands.

[0373] "Text data" is character information converted using speech recognition technology, and is the data that is the subject of analysis.

[0374] "Artificial intelligence means" are means that use technologies such as machine learning and deep learning to analyze data and detect patterns and anomalies.

[0375] "Means for notifying relevant parties" refers to means for conveying warnings and information to users and relevant third parties when possible fraud is detected.

[0376] "Means of issuing instructions to temporarily lock a trading account" refers to a means of instructing a bank or financial institution to freeze a trading account in order to prevent further damage when fraud is suspected.

[0377] "Emotion recognition means" refers to a technical means for assessing a user's emotional state from voice and other data, and determining the level of stress or anxiety.

[0378] The "interface means for carrying out identity verification procedures" refers to means including input devices and software for the user to verify their identity.

[0379] An "anomalous voice pattern" refers to a unique voice characteristic or word usage that differs from normal speech, and is a pattern that is an indicator of possible fraud.

[0380] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing the user's emotions. This system mainly consists of a server, a terminal, and a user, and operates as follows.

[0381] Audio data collection and analysis

[0382] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence means to detect abnormal patterns and fraud-related keywords. This identification is performed using a generative AI model.

[0383] Emotion recognition by emotion engine

[0384] The server then passes the call audio data to an emotion recognition engine to recognize the user's emotions. This emotion engine assesses whether the user is experiencing stress or anxiety. The user's emotion data is then used to augment the assessment of the likelihood of fraud.

[0385] Fraud Assessment and Notification

[0386] If the analysis determines that there is a high possibility of fraud, the server will raise a fraud warning flag and immediately notify the user and relevant parties via phone, SMS, email, etc. to quickly warn them.

[0387] Temporary locking of trading accounts

[0388] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's trading account, preventing the user's funds from being leaked to the fraudsters.

[0389] Identity verification procedures

[0390] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device will provide a verification code entry screen or face recognition via video call. After successful identity verification, the server will unlock the trading account.

[0391] Specific examples

[0392] Scenario: Senior Citizen Refund Fraud

[0393] 1. Start a call

[0394] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[0395] 2. Collection and analysis of audio data

[0396] The device transmits the audio data of the fraudulent call to the server in real time.

[0397] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[0398] At the same time, the server uses an emotion engine to analyze Mr. A's emotional state and assess his level of stress and anxiety.

[0399] 3. Fraud Assessment and Notification

[0400] The server detects keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on the results of AI analysis and emotional data.

[0401] The server immediately notifies Mr. A and his family of the possible fraud.

[0402] 4. Temporary Locking of Trading Account

[0403] The server instructs the partner bank to temporarily lock Mr. A's trading account.

[0404] 5. Identity Verification Procedures

[0405] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[0406] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[0407] After confirmation, the server contacts the bank and unlocks Mr. A's trading account.

[0408] This technology can detect fraudulent phone calls in real time, and by utilizing user emotional data, it is possible to improve the accuracy of fraud detection and prevent damage by quickly notifying relevant parties.In addition, the safety of users' funds can be ensured by locking their trading accounts and implementing identity verification procedures.

[0409] An example of a prompt to be input to the generative AI model is as follows:

[0410] "Develop a system that analyzes human emotions and detects fraudulent phone calls. Design an application that detects calls containing keywords such as 'refund,' 'bank,' and 'transfer,' as well as the user's feelings of stress and anxiety, and displays a warning if there is a high possibility of fraud."

[0411] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0412] Step 1:

[0413] The terminal starts collecting the call voice data when the call starts. The terminal transmits the collected voice data to the server in real time. The input is the call voice data, and the output is the transmitted voice data.

[0414] Step 2:

[0415] The server converts the received voice data into text data using voice recognition technology. Specifically, the server converts the voice data into text data using a voice recognition engine (e.g., Google Speech-to-Text API). The input is voice data, and the output is text data.

[0416] Step 3:

[0417] The server analyzes the converted text data using artificial intelligence means to detect fraud-related keywords and abnormal patterns. Specifically, the server uses natural language processing (NLP) techniques and deep learning models to extract fraud-related keywords (e.g., "refund," "bank," "transfer," etc.) from the text data. The input is the text data, and the output is the analysis results.

[0418] Step 4:

[0419] At the same time, the server passes the call voice data to an emotion recognition engine to evaluate the user's emotional state. Specifically, the server uses a voice emotion recognition model (e.g., IBM Watson Tone Analyzer) to evaluate emotional states such as stress and anxiety from the voice data. The input is voice data, and the output is emotional state data.

[0420] Step 5:

[0421] The server comprehensively evaluates the text data analysis results and the emotional state data to determine the likelihood of fraud. If the likelihood of fraud is high, the server raises a fraud warning flag. In this step, the text analysis results and the emotional state data are used as inputs, and a fraud warning flag is generated as output.

[0422] Step 6:

[0423] The server notifies users and other interested parties when a fraud alert flag is raised. Specifically, the server uses a notification method (e.g., phone call, SMS, email) to send a message informing them of a possible fraud. The input is the fraud alert flag, and the output is the notification message sent.

[0424] Step 7:

[0425] After the server is notified of the fraud, it instructs the affiliated financial institution to temporarily lock the user's trading account. The input is a fraud warning flag, and the output is an instruction to lock the trading account.

[0426] Step 8:

[0427] The user receives the fraud notification and starts the identity verification procedure using the device. Specifically, the device provides a verification code input screen and facial recognition functionality via video call. The input is the verification code and facial recognition data, and the output is the identity verification result.

[0428] Step 9:

[0429] After the identity verification procedure is successful, the server contacts the financial institution and unlocks the trading account. The input is the identity verification result, and the output is the instruction to unlock the trading account.

[0430] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the 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.

[0431] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0432] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0433] [Second embodiment]

[0434] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0435] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0436] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0437] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0438] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0439] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0440] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0441] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0442] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[0443] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0444] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0445] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[0446] The present invention is a system that monitors communication lines to be monitored, detects calls that are likely to be fraudulent, and prevents damage before they occur. This system is mainly composed of a server, a terminal, and a user, and operates as follows.

[0447] Audio data collection and analysis

[0448] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence means to detect abnormal patterns and fraud-related keywords in the text.

[0449] Fraud Assessment and Notification

[0450] After conversion and analysis, the server evaluates the likelihood of fraud. If it determines that there is a high likelihood of fraud, the server notifies the relevant user and relevant parties. The notification is sent via phone, SMS, email, etc., to provide a comprehensive warning.

[0451] Temporary locking of bank accounts

[0452] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from being leaked to the fraudsters.

[0453] Identity verification procedures

[0454] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a verification code entry screen or facial recognition via video call. After successful identity verification, the server unlocks the bank account.

[0455] Specific examples

[0456] Scenario: Senior Citizen Refund Fraud

[0457] 1. Start a call

[0458] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[0459] 2. Collection and analysis of audio data

[0460] The device transmits the audio data of the fraudulent call to the server in real time.

[0461] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[0462] 3. Fraud Assessment and Notification

[0463] The server detects keywords such as "refund," "bank," and "transfer" and determines that there is a high possibility of fraud.

[0464] The server immediately notifies Mr. A and his family of the possible fraud.

[0465] 4. Temporarily locking your bank account

[0466] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[0467] 5. Identity Verification Procedures

[0468] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[0469] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[0470] After verification, the server contacts the bank and unlocks Mr. A's bank account.

[0471] This invention makes it possible to detect fraudulent phone calls to elderly and young people in real time, and to prevent damage by quickly notifying the relevant parties. It also ensures the safety of users' funds by locking their bank accounts and implementing identity verification procedures.

[0472] The processing flow will be explained below.

[0473] Step 1:

[0474] The terminal detects that the user has started a telephone call, and simultaneously starts collecting the call voice data.

[0475] Step 2:

[0476] The terminal transmits the collected call voice data to the server in real time.

[0477] Step 3:

[0478] The server passes the received voice data to a voice recognition engine and converts the voice data into text data.

[0479] Step 4:

[0480] The server passes the converted text data to an AI analysis module to detect abnormal patterns and fraud-related keywords within the text data.

[0481] Step 5:

[0482] The server will raise a fraud warning flag if AI analysis determines that fraud is likely.

[0483] Step 6:

[0484] If a fraud alert flag is raised, the server will immediately notify the user and relevant parties via phone, SMS, email, or other means.

[0485] Step 7:

[0486] The server instructs the user's designated bank to temporarily lock the user's bank account.

[0487] Step 8:

[0488] The user receives the notification and starts the identity verification procedure according to the specified procedure. The terminal provides an interface for identity verification.

[0489] Step 9:

[0490] The device sends the verification code entered by the user and the facial recognition results obtained via video call to the server.

[0491] Step 10:

[0492] The server receives the authentication information and verifies that the user is a legitimate user.

[0493] Step 11:

[0494] If the verification is successful, the server instructs the bank to unlock the user's bank account.

[0495] This series of processes allows for the real-time detection of potentially fraudulent calls and prompt notification and response.

[0496] Example 1

[0497] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0498] In recent years, fraud methods using telecommunications lines have become more sophisticated and diverse. Refund fraud and bank transfer fraud targeting the elderly in particular have been on the rise, causing serious damage. Conventional crime prevention methods have struggled to detect fraudulent activity in real time and prevent damage before it occurs. There is also a lack of methods for quickly and accurately assessing the possibility of fraud and notifying relevant parties. In addition, to prevent fraud once it has occurred, there is a need for a function to quickly lock financial institution accounts and a method for users to easily complete identity verification procedures.

[0499] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0500] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties if a fraudulent call is highly likely, means for issuing instructions to temporarily lock the relevant financial institution's account, means for unlocking the account by providing personal identification information, means for transmitting voice data in real time when the call is made, means for detecting fraud keywords based on the analysis results, and means for transmitting notifications using multiple communication means. This makes it possible to detect potentially fraudulent calls in real time and prevent damage before it occurs. Furthermore, the relevant financial institution's account can be immediately locked and safely unlocked through personal identification procedures.

[0501] A "monitored communication line" is a communication network that is constantly monitored to detect potentially fraudulent calls.

[0502] "Speech recognition technology" is a technology used to convert voice data into text data.

[0503] "Voice data" means voice information, in analog or digital form, captured during a telephone call.

[0504] "Text data" is character information converted from voice data using voice recognition technology.

[0505] "Artificial intelligence means" are machine learning or deep learning algorithms used to analyze text data and detect specific patterns or keywords.

[0506] "Fraudulent keywords" are words or phrases that may be associated with fraudulent activity.

[0507] "Assessing the likelihood of fraud" means determining the risk of fraud based on pattern and keyword analysis within text data.

[0508] "Notifying" means sending warnings and information to users and other interested parties when fraud is likely.

[0509] "Interested Persons" are people or entities that should receive information about fraud risks.

[0510] A "financial institution account," or simply an "account," is an account managed by a financial institution such as a bank that records a user's deposits and funds.

[0511] "Temporarily locking" means suspending activities such as withdrawals from a financial institution account for a certain period of time.

[0512] "Personal identification information" refers to information or data used to prove the identity of a user.

[0513] "Unlocking" means lifting the suspension of an account at a financial institution after identity verification has been completed.

[0514] "Sending in real time when a call is made" means sending voice data to the server immediately at the moment the call is started.

[0515] "Detecting fraud keywords based on the analysis results" means that the artificial intelligence means analyzes the text data and identifies specific words related to fraudulent activities.

[0516] "Sending notifications using multiple communication methods" means sending notifications using different media such as telephone, SMS, and email.

[0517] The present invention relates to a system for monitoring communication lines to detect calls that are likely to be fraudulent and prevent damage before they occur. This system uses a server, a terminal, and a user as its main components.

[0518] The server uses high-performance communication monitoring software (such as WireShark or Snort) to constantly monitor the communication line. When a call is initiated, the device sends the call audio data to the server in real time. The audio data is encoded using VoIP technology (such as the SIP protocol) and sent via a secure communication protocol such as HTTPS.

[0519] The server converts the received voice data into text using the Google Cloud Speech-to-Text API. It then uses speech recognition technology to obtain accurate text data, and analyzes the converted text data using generative AI models such as BERT and GPT-3. The analysis aims to detect abnormal patterns and fraud-related keywords in the text (e.g., "refund," "bank," and "transfer").

[0520] If a fraud risk is determined to be high, the server will notify the relevant parties (e.g., the user and their family) via SMS or phone using the Twilio API. It is also possible to notify them by email using the SendGrid API. This notification allows users to quickly be informed of the risk of fraud.

[0521] After sending the fraud notification, the server uses the API of the partner financial institution (e.g., Plaid API) to issue a command to temporarily lock the user's financial institution account, thereby preventing the user's funds from being stolen due to fraudulent activity.

[0522] When the user receives the fraud notification, they use the provided interface (e.g., a verification code entry screen or video call) to verify their identity. The device provides the interface for identity verification through the API of Jumio or IDNow. The user completes identity verification by entering a verification code or by performing facial recognition via a video call. Once verification is complete, the server again uses the financial institution's API to send instructions to unlock the bank account.

[0523] Specific examples

[0524] Scenario: Senior Citizen Refund Fraud

[0525] 1. Start a call

[0526] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[0527] 2. Collection and analysis of audio data

[0528] The device (smartphone) transmits the audio data of the fraudulent call to the server in real time.

[0529] The server converts the received voice data into text data using Google Cloud Speech-to-Text.

[0530] The converted text data is analyzed using the BERT model to detect keywords such as "refund," "bank," and "transfer."

[0531] 3. Fraud Assessment and Notification

[0532] Based on the keywords detected, the server determines with a high probability that the request is fraudulent.

[0533] The server uses the Twilio API to notify Mr. A and his family via SMS of the possible fraud.

[0534] 4. Temporary locking of financial institution accounts

[0535] The server sends an instruction to the partner bank via API to temporarily lock Mr. A's financial institution account.

[0536] 5. Identity Verification Procedures

[0537] The user, Mr. A, or his / her family member clicks on the link provided in the notification to begin the identity verification process.

[0538] Enter the verification code in the interface provided by your device or use facial recognition via video call.

[0539] After the verification is complete, the server again uses the API to send instructions to unlock the bank account.

[0540] Example prompts for generative AI models

[0541] Using the following system scenario, please explain in natural language how a program to detect potentially fraudulent calls works:

[0542] system:

[0543] The server monitors the communication line and analyzes the voice data.

[0544] The terminal transmits the call voice data to the server.

[0545] The server uses artificial intelligence to analyze the converted text data and assess the likelihood of fraud.

[0546] If fraud is likely, the server will notify the user and lock the financial institution account.

[0547] The user goes through the identity verification procedure and the server unlocks the device.

[0548] Example: Senior Citizen Refund Fraud Scenario

[0549] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0550] Specific explanation of the program based on processing steps

[0551] Step 1:

[0552] Call detection

[0553] The terminal monitors the target communication line and detects when a call has started. As input, it receives a call signal from the communication line and analyzes this signal to detect the call start trigger. The output is event information that a call has started, which starts the subsequent processing.

[0554] Specific operation: When elderly person A answers the phone, the device's monitoring app detects the start of the call and immediately begins collecting voice data.

[0555] Step 2:

[0556] Collection and transmission of voice data

[0557] The device receives call start event information and collects real-time call audio data, which is then encoded using VoIP technology and sent to the server via a secure communication protocol (HTTPS).

[0558] As input, it receives real-time audio data acquired by the terminal, and as output, it transmits encoded audio data to the server.

[0559] Specific operation: While Person A continues the conversation, the device captures audio data and transmits it to the server through a secure channel.

[0560] Step 3:

[0561] Converting audio data to text

[0562] The server receives the voice data sent from the device and converts it into text data using the Google Cloud Speech-to-Text API. The input is the encoded voice data, and the output is the converted text data.

[0563] Specific operation: The server calls the Google Cloud Speech-to-Text API to convert the audio data into text data.

[0564] Step 4:

[0565] Text data analysis

[0566] The server receives text data from the Google Cloud Speech-to-Text API and analyzes it using artificial intelligence models such as BERT and GPT-3. The analysis detects keywords and anomalous patterns that may be indicative of fraud. The output is information about the detected fraud-related keywords and anomalous patterns.

[0567] Specific operation: The server inputs the text data into an artificial intelligence model and detects keywords such as "refund," "bank," and "transfer."

[0568] Step 5:

[0569] Fraud Assessment and Notification

[0570] The server evaluates the likelihood of fraud based on the results of text data analysis. It receives keyword detection results as input, scores the likelihood of fraud, and outputs a notification based on the evaluation results. It notifies users and relevant parties via SMS, phone, or email using the Twilio API, or the SendGrid API.

[0571] Specific operation: If the analysis results indicate a high risk of fraud, the server uses the Twilio API to send an SMS warning to Mr. A and his family.

[0572] Step 6:

[0573] Temporary locking of financial institution accounts

[0574] Based on the fraud risk assessment results, the server uses the API (such as Plaid API) of the partner financial institution to send an instruction to temporarily lock the user's financial institution account. The input receives the fraud risk assessment result, and the output is the completion of the lock instruction.

[0575] Specific operation: If it is determined that there is a high possibility of fraud, the server will generate an API request to immediately temporarily lock Mr. A's financial institution account.

[0576] Step 7:

[0577] Identity verification procedures

[0578] The user follows the notification to complete the identity verification procedure. Using the provided interface (verification code entry screen or video call), the user sends the verification information from the device to the server. The server receives the verification code entered by the user and facial recognition data as input, and the output is the completion of identity verification.

[0579] Specific actions: Person A or his / her family follows the instructions in the notification, enters the verification code on the device's browser, and sends it to the server.

[0580] Step 8:

[0581] Unlock

[0582] After the identity verification is successful, the server again uses the financial institution's API to send an unlock instruction. The input is the successful identity verification information, and the output is the unlock of the financial institution account.

[0583] Specific operation: Once identity verification is complete, the server contacts the financial institution via API and unlocks Mr. A's account.

[0584] This process flow allows for real-time detection of potentially fraudulent calls, preventing damage before it occurs, and providing rapid response to protect users' financial assets.

[0585] (Application example 1)

[0586] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0587] In modern society, voice-based fraud is on the rise, and fraud victimization of elderly people is becoming particularly serious. There is also a need for a method to monitor fraudulent call content in real time and respond quickly, but current technology makes this difficult to achieve. With current systems, it takes time for users receiving fraudulent calls to realize they are being scammed, making it difficult to prevent fraud before it occurs. To solve this problem, a system is needed that can assess the possibility of fraud in real time and respond quickly.

[0588] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0589] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties when there is a high possibility of fraud, means for issuing an instruction to temporarily lock the relevant bank account, means for notifying the user of a fraud warning in real time based on the results of the voice recognition and analysis, communication means for monitoring voice call data that is likely to be fraudulent and sending the notification, means for evaluating abnormal voice patterns using a generative AI model, and means for generating prompt sentences to be input to the generative AI model. This makes it possible to monitor the communication lines to be monitored in real time, quickly evaluate and respond to the possibility of fraud, and immediately notify the user and relevant parties.

[0590] 1. "Monitored communication lines" refer to lines that carry communications such as telephone and internet that are subject to monitoring.

[0591] 2. "Monitoring means" means devices or programs that constantly monitor communication lines and collect data.

[0592] 3. "Speech recognition technology" refers to technology that converts voice data into text data.

[0593] 4. "Voice data" refers to sound information collected through telephone conversations, recordings, etc.

[0594] 5. "Text data" means data that has been converted from audio data into text information.

[0595] 6. "Artificial intelligence means" means systems or software that use machine learning or natural language processing to perform specific tasks.

[0596] 7. "Assessing the likelihood of fraud" means analyzing voice and text data to determine whether there is a possibility of fraud.

[0597] 8. "Means of notification" means the means (e.g., telephone, SMS, email, etc.) used to send warnings or information to interested parties.

[0598] 9. "Means for issuing instructions to temporarily lock a bank account" means a means for sending an instruction to an affiliated financial institution to temporarily suspend the use of a user's account.

[0599] 10. "Means for providing real-time fraud alerts to users based on speech recognition and analysis results" means means for instantly sending alerts to users based on analysis results from speech recognition technology and artificial intelligence means.

[0600] 11. "Communication means for monitoring voice call data that may be fraudulent and sending notifications" refers to communication devices or programs that constantly monitor call data that may be fraudulent and notify the results.

[0601] 12. "Means for assessing anomalous speech patterns using a generative AI model" means means for identifying and assessing anomalous speech patterns using a generative AI model.

[0602] 13. "Means for generating prompt sentences to be input to a generative AI model" means means for creating initial input sentences to provide appropriate analytical instructions to a generative AI model.

[0603] The system for implementing this invention mainly comprises a server, a terminal, and a user. The server constantly monitors the communication line to be monitored and receives call voice data transmitted from the terminal in real time. The server converts the received voice data into text data using speech recognition technology. The converted text data is then analyzed using artificial intelligence to detect abnormal patterns and fraud-related keywords in the text.

[0604] Audio data collection and analysis

[0605] The server constantly monitors the target communication line. The device sends the call audio data to the server in real time when the call starts. The server converts this audio data into text data using speech recognition technology (e.g., Google Speech Recognition API). This text data is then analyzed using artificial intelligence means (e.g., machine learning algorithms).

[0606] Fraud Assessment and Notification

[0607] Once the conversion and analysis are complete, the server evaluates the likelihood of fraud. If it determines that there is a high likelihood of fraud, the server notifies the user and relevant parties. Notifications are sent with high speed via phone, SMS, email, etc. The server also issues instructions to temporarily lock the relevant bank account.

[0608] Temporary locking of bank accounts

[0609] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from being leaked to the fraudsters.

[0610] Identity verification procedures

[0611] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a verification code entry screen or facial recognition via video call. After successful identity verification, the server unlocks the bank account.

[0612] Evaluation by generative AI model

[0613] Based on the results of speech recognition and analysis, the server evaluates abnormal speech patterns using a generative AI model, to which an appropriate prompt is input. For example, the following prompt might be used:

[0614] Prompt text: Convert the voice data into text and detect the keywords "refund," "bank," or "transfer" to determine the possibility of fraud. If there are signs of fraud, send a warning to the user via email.

[0615] Specific examples

[0616] Below are some specific scenarios where this invention can be used:

[0617] Preventing refund fraud among the elderly

[0618] 1. Start a call:

[0619] An elderly person, Mr. A, receives a fraudulent phone call and is asked for "information regarding receiving a refund."

[0620] 2. Audio data collection and analysis:

[0621] The device transmits the voice data of the fraudulent call in real time to a server, which uses voice recognition technology to convert the voice data into text data, which is then analyzed using artificial intelligence means.

[0622] 3. Potential Fraud Assessment and Notification:

[0623] The server detects keywords such as "refund," "bank," and "transfer," and determines that there is a high possibility of fraud. The server immediately notifies Mr. A and his family of the possibility of fraud.

[0624] 4. Temporarily locking your bank account:

[0625] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[0626] 5. Identity Verification Procedure:

[0627] The user, Mr. A, or his / her family member, confirms the notification and begins the identity verification process. They enter a verification code using the interface provided by the device, and the server receives the information. After verification, the server contacts the bank and unlocks Mr. A's bank account.

[0628] As described above, this invention allows for real-time monitoring of potentially fraudulent calls and prompt response. This makes it possible to prevent fraudulent calls from affecting both the elderly and young. Furthermore, the safety of users' funds can be ensured by locking their bank accounts and implementing identity verification procedures.

[0629] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0630] Step 1:

[0631] Initiating a call and collecting audio data

[0632] When a terminal initiates a user's call, voice data is collected in real time. This voice data is sent by the terminal to a server. The input is the user's call voice, and the output is the voice data sent to the server. The specific operations include the terminal capturing the voice of the call and sending the data to the server.

[0633] Step 2:

[0634] Converting audio data to text

[0635] The server converts the received voice data into text data using voice recognition technology (e.g., Google Speech Recognition API). The input is the voice data sent from the device, and the output is text data. Specific operations include calling the voice recognition API and converting the voice data into text data.

[0636] Step 3:

[0637] Text data analysis

[0638] The server analyzes the text data using artificial intelligence means (e.g., machine learning algorithms). The input is the text data and the output is the analysis results. Specific operations include using machine learning models to detect anomalous patterns and fraud-related keywords in the text data.

[0639] Step 4:

[0640] Fraud likelihood assessment

[0641] The server evaluates the possibility of fraud based on the analysis results. The input is the analysis results, and the output is the fraud possibility evaluation result. Specific operations include scoring the possibility of fraud based on the analysis results and determining whether the score exceeds a threshold.

[0642] Step 5:

[0643] Sending notifications

[0644] If it determines that there is a high possibility of fraud, the server notifies the user and relevant parties. The input is the fraud possibility assessment result, and the output is a notification message. Specific operations include sending warnings to the user and relevant parties via email, SMS, phone, etc.

[0645] Step 6:

[0646] Temporary locking of bank accounts

[0647] The server issues an instruction to temporarily lock the relevant bank account. The input is the fraud probability assessment result and user information, and the output is a lock instruction to be sent to the bank. Specific operations include sending a request to temporarily lock the account to the partner bank's API.

[0648] Step 7:

[0649] Start the identity verification process

[0650] Upon receiving the notification, the user begins the identity verification process. The device then provides a screen for entering a verification code or facial recognition via video call. The input is the notification message and user information, and the output is progress information on identity verification. Specific actions include the user entering a verification code or performing facial recognition via video call.

[0651] Step 8:

[0652] Unlocking a bank account

[0653] After successful identity verification, the server issues a command to unlock the bank account. The input is the successful identity verification information, and the output is an unlock command sent to the bank. Specific operations include sending an unlock request to the partner bank's API.

[0654] Step 9:

[0655] Analyzing abnormal voice patterns using AI models

[0656] The server analyzes abnormal voice patterns using a generative AI model. The input is the voice recognition and analysis results, and the output is the evaluation result of the abnormal voice pattern. Specific operations include inputting appropriate prompt sentences to the generative AI model and evaluating the abnormal voice pattern.

[0657] Step 10:

[0658] Generate prompt statement

[0659] The server generates prompt sentences to be input to the generative AI model. The inputs are the AI ​​model specifications and the information to be analyzed, and the output is the prompt sentence. Specific operations include creating prompt sentences appropriate to the situation to be analyzed (e.g., "Convert the voice data into text and detect any of the keywords 'refund,' 'bank,' or 'transfer' to determine the possibility of fraud. If there are signs of fraud, send a warning to the user via email.").

[0660] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0661] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing the user's emotions. This system mainly consists of a server, a terminal, and a user, and operates as follows.

[0662] Audio data collection and analysis

[0663] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence tools to detect abnormal patterns and fraud-related keywords.

[0664] Emotion recognition by emotion engine

[0665] The server then passes the call audio data to an emotion engine to recognize the user's emotions. This emotion engine assesses whether the user is experiencing stress or anxiety. The user's emotion data is then used to augment the assessment of the likelihood of fraud.

[0666] Fraud Assessment and Notification

[0667] If the analysis determines that there is a high possibility of fraud, the server will raise a fraud warning flag and immediately notify the user and relevant parties via phone, SMS, email, etc. to quickly warn them.

[0668] Temporary locking of bank accounts

[0669] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from reaching the fraudsters.

[0670] Identity verification procedures

[0671] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a screen for entering a verification code or facial recognition via video call. After successfully verifying the user's identity, the server unlocks the bank account.

[0672] Specific examples

[0673] Scenario: Senior Citizen Refund Fraud

[0674] 1. Start a call

[0675] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[0676] 2. Collection and analysis of audio data

[0677] The device transmits the audio data of the fraudulent call to the server in real time.

[0678] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[0679] At the same time, the server uses an emotion engine to analyze Mr. A's emotional state and assess his level of stress and anxiety.

[0680] 3. Fraud Assessment and Notification

[0681] The server detects keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on the results of AI analysis and emotional data.

[0682] The server immediately notifies Mr. A and his family of the possible fraud.

[0683] 4. Temporarily locking your bank account

[0684] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[0685] 5. Identity Verification Procedures

[0686] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[0687] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[0688] After verification, the server contacts the bank and unlocks Mr. A's bank account.

[0689] This invention makes it possible to detect fraudulent phone calls suffered by the elderly and young in real time, and by utilizing user emotion data, it is possible to improve the accuracy of fraud detection and prevent damage by quickly notifying relevant parties.In addition, the safety of users' funds can be ensured by locking bank accounts and implementing identity verification procedures.

[0690] The processing flow will be explained below.

[0691] Step 1:

[0692] The terminal detects that the user has started a telephone call, and simultaneously starts collecting the call voice data.

[0693] Step 2:

[0694] The terminal transmits the collected call voice data to the server in real time.

[0695] Step 3:

[0696] The server passes the received voice data to a voice recognition engine and converts the voice data into text data.

[0697] Step 4:

[0698] The server passes the converted text data to an AI analysis module to detect abnormal patterns and fraud-related keywords within the text data.

[0699] Step 5:

[0700] At the same time, the server passes the voice data to the emotion engine, which analyzes the user's emotions by analyzing the tone, tempo, and speaking style of the voice to assess whether the user is feeling stressed or anxious.

[0701] Step 6:

[0702] The server makes a comprehensive judgment based on the results of the AI ​​analysis and the evaluation results of the emotion engine, and if there is a high possibility of fraud, it raises a fraud warning flag.

[0703] Step 7:

[0704] If a fraud alert flag is raised, the server will immediately notify the user and relevant parties via phone, SMS, email, or other means.

[0705] Step 8:

[0706] The server instructs the user's designated bank to temporarily lock the user's bank account.

[0707] Step 9:

[0708] The user receives the notification and starts the identity verification procedure according to the specified procedure. The terminal provides an interface for identity verification.

[0709] Step 10:

[0710] The device sends the verification code entered by the user and the facial recognition results obtained via video call to the server.

[0711] Step 11:

[0712] The server receives the authentication information and verifies that the user is a legitimate user.

[0713] Step 12:

[0714] If the verification is successful, the server instructs the bank to unlock the user's bank account.

[0715] This process allows for real-time detection of potentially fraudulent calls, and prompt notification and response. The emotion engine also enables advanced fraud detection that takes into account the user's emotional state.

[0716] Example 2

[0717] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0718] Conventional communication monitoring systems simply convert voice data into text data and detect fraud through simple keyword matching, but this method is insufficient to counter increasingly sophisticated fraud techniques. Furthermore, because the system ignores the user's emotions and psychological state, the accuracy of fraud detection can be reduced. Furthermore, it is difficult to respond quickly after fraud is detected, making it difficult to effectively prevent users from losing funds.

[0719] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0720] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, emotion recognition means for analyzing the call voice data and recognizing the user's emotion, means for notifying relevant parties when there is a high possibility of fraud, means for issuing instructions to temporarily lock the relevant bank account, and interface means for carrying out identity verification procedures. This enables highly sophisticated fraud to be detected with high accuracy, and by utilizing user emotion data, the accuracy of fraud detection is improved, enabling quick and effective response.

[0721] A "monitored communication line" is a communication infrastructure that is monitored and through which information is sent and received.

[0722] "Speech recognition technology" is a technology for converting voice data into text data, and typically uses an algorithm to extract linguistic information from voice.

[0723] "Voice data" refers to data that is a digital recording of the voice generated during a call.

[0724] "Text data" refers to the format of character information obtained after voice data is converted using voice recognition technology.

[0725] "Artificial intelligence means" means artificial intelligence-based techniques for analyzing the converted text data and assessing any abnormal patterns or potential fraud.

[0726] The "emotion recognition means" is a system that analyzes call voice data and evaluates the user's emotions and psychological state.

[0727] "Notification means" refers to methods or tools used to warn or inform relevant parties when a high possibility of fraud is determined.

[0728] A "bank account locking mechanism" is a system that executes instructions to temporarily freeze a user's bank account if fraud is suspected.

[0729] "Interface means" refers to input and output means for users and related parties to interact with the system, and includes, for example, a confirmation code input screen and a function for video calling.

[0730] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing user emotions. This system mainly consists of three components: a server, a terminal, and a user. Specific embodiments of the system are described below.

[0731] Audio data collection and analysis

[0732] The server constantly monitors the target communication line. Network monitoring software (e.g., Wireshark or PRTG) is used for monitoring. When a call is initiated, the device sends the call audio data to the server in real time. A streaming API (e.g., WebRTC) is used to collect the audio data, minimizing the delay in the audio data.

[0733] The server converts the received voice data into text data using the Google Cloud Speech-to-Text API. This conversion process is performed in parallel, and text is generated in real time.

[0734] The server then analyzes the converted text data using natural language processing (NLP) techniques such as GPT-3, which can identify unusual patterns and fraud-related keywords (e.g., "refund," "bank," "transfer," etc.).

[0735] Emotion recognition by emotion engine

[0736] The server then passes the call audio data to an emotion recognition engine (Emotion AI) to recognize the user's emotions. The emotion engine analyzes the tone and pace of the voice to assess whether the user is feeling stressed or anxious. For example, a trembling voice or a high-pitched tone can be detected as a sign of stress.

[0737] Fraud Assessment and Notification

[0738] The server combines the analysis results with sentiment data to assess the likelihood of fraud, and if it determines that fraud is likely, it raises a fraud warning flag.

[0739] If a potential fraud is detected, the server immediately notifies the user and relevant parties. The notification is sent in real time via SMS or email, allowing the user and relevant parties to take immediate action. For example, the notification may say, "A potential fraudulent call has been detected. Please take action immediately."

[0740] Temporary locking of bank accounts

[0741] After the fraud notification is complete, the server sends an instruction to the partner financial institution via API to temporarily lock the user's bank account, thereby preventing the outflow of funds.

[0742] Identity verification procedures

[0743] Upon receiving the notification, the user begins the identity verification process. The device provides facial recognition via a verification code input screen or video call. For example, the user receives a verification code via SMS and enters it on the device's input screen. Alternatively, facial recognition is performed via a video call to verify the user's identity. The server then contacts the financial institution and unlocks the bank account.

[0744] Specific examples

[0745] For example, consider the case of an elderly person named Mr. A who receives a call about a tax refund scam. The moment the call starts, the device sends a call start message to the server via the API, and captures and transmits voice data in real time.

[0746] The server converts the voice data into text using the Google Cloud Speech-to-Text API and analyzes the text using GPT-3. At the same time, it uses an emotion engine to analyze Mr. A's emotional state, detecting that he is experiencing increased stress and anxiety.

[0747] The server detects fraud-related keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on AI analysis and emotional data. It immediately notifies Mr. A and his family of the possibility of fraud and instructs the financial institution to temporarily lock Mr. A's bank account.

[0748] Upon receiving the notification, Mr. A or his family members can use the device interface to complete identity verification procedures, and once verification is complete, the bank account will be unlocked.

[0749] Prompt Sentence Examples

[0750] "They mentioned a refund, is this a scam?"

[0751] "How can I identify a fraudulent phone call?"

[0752] "Please let me know the analysis results of calls that are likely to be fraudulent."

[0753] The present invention makes it possible to detect fraudulent phone call damage in real time, improve the accuracy of fraud detection by utilizing user emotion data, and respond quickly.

[0754] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0755] Step 1:

[0756] The server constantly monitors the target communication line. The input is the connection information of the communication line, and the output is the call start trigger. Specifically, it uses network monitoring software (e.g., Wireshark or PRTG) to detect the start of a call in real time.

[0757] Step 2:

[0758] When a call is started, the device notifies the server of the call start information. The input is a call start event, and the output is a notification message to the server. Specifically, the device sends a call start message to the server through the API. This message contains the call identification information.

[0759] Step 3:

[0760] The device captures the call audio data in real time and sends it to the server. The input is the call audio data, and the output is a stream of audio data to the server. Specifically, it uses a streaming API (e.g., WebRTC) to capture the audio data and sends it to the server in an encrypted format.

[0761] Step 4:

[0762] The server converts the received voice data into text data using the Google Cloud Speech-to-Text API. The input is voice data and the output is text data. Specifically, the server sends voice data to the API and generates text data in real time. Parallel processing is used to improve conversion efficiency.

[0763] Step 5:

[0764] The server analyzes the converted text data using natural language processing technology such as GPT-3. The input is text data, and the output is the analysis results. Specifically, the text data is input into a generative AI model to detect fraud-related keywords such as "refund," "bank," and "transfer."

[0765] Step 6:

[0766] The server passes the call voice data to the Emotion AI engine, which analyzes the user's emotions. The input is voice data, and the output is emotional evaluation data. Specifically, it analyzes the voice tone and pace to evaluate whether the user is feeling stressed or anxious.

[0767] Step 7:

[0768] The server integrates the analysis results and emotion data to assess the likelihood of fraud. The input is the text analysis results and emotion evaluation data, and the output is a fraud warning flag. Specifically, it determines the likelihood of fraud based on abnormal patterns, fraud-related keywords, and user emotion data.

[0769] Step 8:

[0770] If it is determined that there is a high possibility of fraud, the server will notify the relevant parties. The input is a fraud warning flag and the output is a notification message. Specifically, a notification is sent to the relevant parties via SMS or email. The message sent will say something like, "A call with a high possibility of fraud has been detected. Please take immediate action."

[0771] Step 9:

[0772] After the fraud notification is complete, the server issues an instruction to the partner financial institution to temporarily lock the user's bank account. The input is the fraud warning flag and the user's bank information, and the output is a lock instruction to the bank. Specifically, the server sends the lock instruction to the bank via API to ensure the safety of the user's funds.

[0773] Step 10:

[0774] Upon receiving the notification, the user begins the identity verification procedure. The input is the notification message, and the output is the identity verification result. Specifically, the procedure is carried out using the verification code input screen or video call function provided by the device. After verification is complete, the server contacts the bank and unlocks the bank account.

[0775] (Application example 2)

[0776] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0777] In recent years, the number of victims of fraudulent phone calls targeting the elderly and young people has been increasing. These fraudulent acts have become more diverse, making it difficult to deal with them using conventional fraud detection methods alone. Furthermore, since the damage often occurs before the user even realizes it, new technologies are needed to respond to fraud early.

[0778] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties when there is a high possibility of fraud, means for issuing an instruction to temporarily lock the relevant trading account, and emotion recognition means for evaluating the user's emotional state. This enables highly accurate fraud detection and rapid response that takes the user's emotional state into account.

[0779] A "monitored communication line" is a line on a communication network that is continuously monitored to monitor activity on specific calls or data traffic.

[0780] "Voice recognition technology" is a technology that captures voice as a digital signal and converts it into text data or commands.

[0781] "Text data" is character information converted using speech recognition technology, and is the data that is the subject of analysis.

[0782] "Artificial intelligence means" are means that use technologies such as machine learning and deep learning to analyze data and detect patterns and anomalies.

[0783] "Means for notifying relevant parties" refers to means for conveying warnings and information to users and relevant third parties when possible fraud is detected.

[0784] "Means of issuing instructions to temporarily lock a trading account" refers to a means of instructing a bank or financial institution to freeze a trading account in order to prevent further damage when fraud is suspected.

[0785] "Emotion recognition means" refers to a technical means for assessing a user's emotional state from voice and other data, and determining the level of stress or anxiety.

[0786] The "interface means for carrying out identity verification procedures" refers to means including input devices and software for the user to verify their identity.

[0787] An "anomalous voice pattern" refers to a unique voice characteristic or word usage that differs from normal speech, and is a pattern that is an indicator of possible fraud.

[0788] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing the user's emotions. This system mainly consists of a server, a terminal, and a user, and operates as follows.

[0789] Audio data collection and analysis

[0790] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence means to detect abnormal patterns and fraud-related keywords. This identification is performed using a generative AI model.

[0791] Emotion recognition by emotion engine

[0792] The server then passes the call audio data to an emotion recognition engine to recognize the user's emotions. This emotion engine assesses whether the user is experiencing stress or anxiety. The user's emotion data is then used to augment the assessment of the likelihood of fraud.

[0793] Fraud Assessment and Notification

[0794] If the analysis determines that there is a high possibility of fraud, the server will raise a fraud warning flag and immediately notify the user and relevant parties via phone, SMS, email, etc. to quickly warn them.

[0795] Temporary locking of trading accounts

[0796] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's trading account, preventing the user's funds from being leaked to the fraudsters.

[0797] Identity verification procedures

[0798] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device will provide a verification code entry screen or face recognition via video call. After successful identity verification, the server will unlock the trading account.

[0799] Specific examples

[0800] Scenario: Senior Citizen Refund Fraud

[0801] 1. Start a call

[0802] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[0803] 2. Collection and analysis of audio data

[0804] The device transmits the audio data of the fraudulent call to the server in real time.

[0805] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[0806] At the same time, the server uses an emotion engine to analyze Mr. A's emotional state and assess his level of stress and anxiety.

[0807] 3. Fraud Assessment and Notification

[0808] The server detects keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on the results of AI analysis and emotional data.

[0809] The server immediately notifies Mr. A and his family of the possible fraud.

[0810] 4. Temporary Locking of Trading Account

[0811] The server instructs the partner bank to temporarily lock Mr. A's trading account.

[0812] 5. Identity Verification Procedures

[0813] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[0814] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[0815] After confirmation, the server contacts the bank and unlocks Mr. A's trading account.

[0816] This technology can detect fraudulent phone calls in real time, and by utilizing user emotional data, it is possible to improve the accuracy of fraud detection and prevent damage by quickly notifying relevant parties.In addition, the safety of users' funds can be ensured by locking their trading accounts and implementing identity verification procedures.

[0817] An example of a prompt to be input to the generative AI model is as follows:

[0818] "Develop a system that analyzes human emotions and detects fraudulent phone calls. Design an application that detects calls containing keywords such as 'refund,' 'bank,' and 'transfer,' as well as the user's feelings of stress and anxiety, and displays a warning if there is a high possibility of fraud."

[0819] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0820] Step 1:

[0821] The terminal starts collecting the call voice data when the call starts. The terminal transmits the collected voice data to the server in real time. The input is the call voice data, and the output is the transmitted voice data.

[0822] Step 2:

[0823] The server converts the received voice data into text data using voice recognition technology. Specifically, the server converts the voice data into text data using a voice recognition engine (e.g., Google Speech-to-Text API). The input is voice data, and the output is text data.

[0824] Step 3:

[0825] The server analyzes the converted text data using artificial intelligence means to detect fraud-related keywords and abnormal patterns. Specifically, the server uses natural language processing (NLP) techniques and deep learning models to extract fraud-related keywords (e.g., "refund," "bank," "transfer," etc.) from the text data. The input is the text data, and the output is the analysis results.

[0826] Step 4:

[0827] At the same time, the server passes the call voice data to an emotion recognition engine to evaluate the user's emotional state. Specifically, the server uses a voice emotion recognition model (e.g., IBM Watson Tone Analyzer) to evaluate emotional states such as stress and anxiety from the voice data. The input is voice data, and the output is emotional state data.

[0828] Step 5:

[0829] The server comprehensively evaluates the text data analysis results and the emotional state data to determine the likelihood of fraud. If the likelihood of fraud is high, the server raises a fraud warning flag. In this step, the text analysis results and the emotional state data are used as inputs, and a fraud warning flag is generated as output.

[0830] Step 6:

[0831] The server notifies users and other interested parties when a fraud alert flag is raised. Specifically, the server uses a notification method (e.g., phone call, SMS, email) to send a message informing them of a possible fraud. The input is the fraud alert flag, and the output is the notification message sent.

[0832] Step 7:

[0833] After the server is notified of the fraud, it instructs the affiliated financial institution to temporarily lock the user's trading account. The input is a fraud warning flag, and the output is an instruction to lock the trading account.

[0834] Step 8:

[0835] The user receives the fraud notification and starts the identity verification procedure using the device. Specifically, the device provides a verification code input screen and facial recognition functionality via video call. The input is the verification code and facial recognition data, and the output is the identity verification result.

[0836] Step 9:

[0837] After the identity verification procedure is successful, the server contacts the financial institution and unlocks the trading account. The input is the identity verification result, and the output is the instruction to unlock the trading account.

[0838] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0839] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0840] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0841] [Third embodiment]

[0842] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0843] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0844] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0845] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0846] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0847] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0848] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0849] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0850] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[0851] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0852] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0853] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[0854] The present invention is a system that monitors communication lines to be monitored, detects calls that are likely to be fraudulent, and prevents damage before they occur. This system is mainly composed of a server, a terminal, and a user, and operates as follows.

[0855] Audio data collection and analysis

[0856] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence means to detect abnormal patterns and fraud-related keywords in the text.

[0857] Fraud Assessment and Notification

[0858] After conversion and analysis, the server evaluates the likelihood of fraud. If it determines that there is a high likelihood of fraud, the server notifies the relevant user and relevant parties. The notification is sent via phone, SMS, email, etc., to provide a comprehensive warning.

[0859] Temporary locking of bank accounts

[0860] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from being leaked to the fraudsters.

[0861] Identity verification procedures

[0862] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a verification code entry screen or facial recognition via video call. After successful identity verification, the server unlocks the bank account.

[0863] Specific examples

[0864] Scenario: Senior Citizen Refund Fraud

[0865] 1. Start a call

[0866] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[0867] 2. Collection and analysis of audio data

[0868] The device transmits the audio data of the fraudulent call to the server in real time.

[0869] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[0870] 3. Fraud Assessment and Notification

[0871] The server detects keywords such as "refund," "bank," and "transfer" and determines that there is a high possibility of fraud.

[0872] The server immediately notifies Mr. A and his family of the possible fraud.

[0873] 4. Temporarily locking your bank account

[0874] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[0875] 5. Identity Verification Procedures

[0876] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[0877] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[0878] After verification, the server contacts the bank and unlocks Mr. A's bank account.

[0879] This invention makes it possible to detect fraudulent phone calls to elderly and young people in real time, and to prevent damage by quickly notifying the relevant parties. It also ensures the safety of users' funds by locking their bank accounts and implementing identity verification procedures.

[0880] The processing flow will be explained below.

[0881] Step 1:

[0882] The terminal detects that the user has started a telephone call, and simultaneously starts collecting the call voice data.

[0883] Step 2:

[0884] The terminal transmits the collected call voice data to the server in real time.

[0885] Step 3:

[0886] The server passes the received voice data to a voice recognition engine and converts the voice data into text data.

[0887] Step 4:

[0888] The server passes the converted text data to an AI analysis module to detect abnormal patterns and fraud-related keywords within the text data.

[0889] Step 5:

[0890] The server will raise a fraud warning flag if AI analysis determines that fraud is likely.

[0891] Step 6:

[0892] If a fraud alert flag is raised, the server will immediately notify the user and relevant parties via phone, SMS, email, or other means.

[0893] Step 7:

[0894] The server instructs the user's designated bank to temporarily lock the user's bank account.

[0895] Step 8:

[0896] The user receives the notification and starts the identity verification procedure according to the specified procedure. The terminal provides an interface for identity verification.

[0897] Step 9:

[0898] The device sends the verification code entered by the user and the facial recognition results obtained via video call to the server.

[0899] Step 10:

[0900] The server receives the authentication information and verifies that the user is a legitimate user.

[0901] Step 11:

[0902] If the verification is successful, the server instructs the bank to unlock the user's bank account.

[0903] This series of processes allows for the real-time detection of potentially fraudulent calls and prompt notification and response.

[0904] Example 1

[0905] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0906] In recent years, fraud methods using telecommunications lines have become more sophisticated and diverse. Refund fraud and bank transfer fraud targeting the elderly in particular have been on the rise, causing serious damage. Conventional crime prevention methods have struggled to detect fraudulent activity in real time and prevent damage before it occurs. There is also a lack of methods for quickly and accurately assessing the possibility of fraud and notifying relevant parties. In addition, to prevent fraud once it has occurred, there is a need for a function to quickly lock financial institution accounts and a method for users to easily complete identity verification procedures.

[0907] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0908] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties if a fraudulent call is highly likely, means for issuing instructions to temporarily lock the relevant financial institution's account, means for unlocking the account by providing personal identification information, means for transmitting voice data in real time when the call is made, means for detecting fraud keywords based on the analysis results, and means for transmitting notifications using multiple communication means. This makes it possible to detect potentially fraudulent calls in real time and prevent damage before it occurs. Furthermore, the relevant financial institution's account can be immediately locked and safely unlocked through personal identification procedures.

[0909] A "monitored communication line" is a communication network that is constantly monitored to detect potentially fraudulent calls.

[0910] "Speech recognition technology" is a technology used to convert voice data into text data.

[0911] "Voice data" means voice information, in analog or digital form, captured during a telephone call.

[0912] "Text data" is character information converted from voice data using voice recognition technology.

[0913] "Artificial intelligence means" are machine learning or deep learning algorithms used to analyze text data and detect specific patterns or keywords.

[0914] "Fraudulent keywords" are words or phrases that may be associated with fraudulent activity.

[0915] "Assessing the likelihood of fraud" means determining the risk of fraud based on pattern and keyword analysis within text data.

[0916] "Notifying" means sending warnings and information to users and other interested parties when fraud is likely.

[0917] "Interested Persons" are people or entities that should receive information about fraud risks.

[0918] A "financial institution account," or simply an "account," is an account managed by a financial institution such as a bank that records a user's deposits and funds.

[0919] "Temporarily locking" means suspending activities such as withdrawals from a financial institution account for a certain period of time.

[0920] "Personal identification information" refers to information or data used to prove the identity of a user.

[0921] "Unlocking" means lifting the suspension of an account at a financial institution after identity verification has been completed.

[0922] "Sending in real time when a call is made" means sending voice data to the server immediately at the moment the call is started.

[0923] "Detecting fraud keywords based on the analysis results" means that the artificial intelligence means analyzes the text data and identifies specific words related to fraudulent activities.

[0924] "Sending notifications using multiple communication methods" means sending notifications using different media such as telephone, SMS, and email.

[0925] The present invention relates to a system for monitoring communication lines to detect calls that are likely to be fraudulent and prevent damage before they occur. This system uses a server, a terminal, and a user as its main components.

[0926] The server uses high-performance communication monitoring software (such as WireShark or Snort) to constantly monitor the communication line. When a call is initiated, the device sends the call audio data to the server in real time. The audio data is encoded using VoIP technology (such as the SIP protocol) and sent via a secure communication protocol such as HTTPS.

[0927] The server converts the received voice data into text using the Google Cloud Speech-to-Text API. It then uses speech recognition technology to obtain accurate text data, and analyzes the converted text data using generative AI models such as BERT and GPT-3. The analysis aims to detect abnormal patterns and fraud-related keywords in the text (e.g., "refund," "bank," and "transfer").

[0928] If a fraud risk is determined to be high, the server will notify the relevant parties (e.g., the user and their family) via SMS or phone using the Twilio API. It is also possible to notify them by email using the SendGrid API. This notification allows users to quickly be informed of the risk of fraud.

[0929] After sending the fraud notification, the server uses the API of the partner financial institution (e.g., Plaid API) to issue a command to temporarily lock the user's financial institution account, thereby preventing the user's funds from being stolen due to fraudulent activity.

[0930] When the user receives the fraud notification, they use the provided interface (e.g., a verification code entry screen or video call) to verify their identity. The device provides the interface for identity verification through the API of Jumio or IDNow. The user completes identity verification by entering a verification code or by performing facial recognition via a video call. Once verification is complete, the server again uses the financial institution's API to send instructions to unlock the bank account.

[0931] Specific examples

[0932] Scenario: Senior Citizen Refund Fraud

[0933] 1. Start a call

[0934] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[0935] 2. Collection and analysis of audio data

[0936] The device (smartphone) transmits the audio data of the fraudulent call to the server in real time.

[0937] The server converts the received voice data into text data using Google Cloud Speech-to-Text.

[0938] The converted text data is analyzed using the BERT model to detect keywords such as "refund," "bank," and "transfer."

[0939] 3. Fraud Assessment and Notification

[0940] Based on the keywords detected, the server determines with a high probability that the request is fraudulent.

[0941] The server uses the Twilio API to notify Mr. A and his family via SMS of the possible fraud.

[0942] 4. Temporary locking of financial institution accounts

[0943] The server sends an instruction to the partner bank via API to temporarily lock Mr. A's financial institution account.

[0944] 5. Identity Verification Procedures

[0945] The user, Mr. A, or his / her family member clicks on the link provided in the notification to begin the identity verification process.

[0946] Enter the verification code in the interface provided by your device or use facial recognition via video call.

[0947] After the verification is complete, the server again uses the API to send instructions to unlock the bank account.

[0948] Example prompts for generative AI models

[0949] Using the following system scenario, please explain in natural language how a program to detect potentially fraudulent calls works:

[0950] system:

[0951] The server monitors the communication line and analyzes the voice data.

[0952] The terminal transmits the call voice data to the server.

[0953] The server uses artificial intelligence to analyze the converted text data and assess the likelihood of fraud.

[0954] If fraud is likely, the server will notify the user and lock the financial institution account.

[0955] The user goes through the identity verification procedure and the server unlocks the device.

[0956] Example: Senior Citizen Refund Fraud Scenario

[0957] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0958] Specific explanation of the program based on processing steps

[0959] Step 1:

[0960] Call detection

[0961] The terminal monitors the target communication line and detects when a call has started. As input, it receives a call signal from the communication line and analyzes this signal to detect the call start trigger. The output is event information that a call has started, which starts the subsequent processing.

[0962] Specific operation: When elderly person A answers the phone, the device's monitoring app detects the start of the call and immediately begins collecting voice data.

[0963] Step 2:

[0964] Collection and transmission of voice data

[0965] The device receives call start event information and collects real-time call audio data, which is then encoded using VoIP technology and sent to the server via a secure communication protocol (HTTPS).

[0966] As input, it receives real-time audio data acquired by the terminal, and as output, it transmits encoded audio data to the server.

[0967] Specific operation: While Person A continues the conversation, the device captures audio data and transmits it to the server through a secure channel.

[0968] Step 3:

[0969] Converting audio data to text

[0970] The server receives the voice data sent from the device and converts it into text data using the Google Cloud Speech-to-Text API. The input is the encoded voice data, and the output is the converted text data.

[0971] Specific operation: The server calls the Google Cloud Speech-to-Text API to convert the audio data into text data.

[0972] Step 4:

[0973] Text data analysis

[0974] The server receives text data from the Google Cloud Speech-to-Text API and analyzes it using artificial intelligence models such as BERT and GPT-3. The analysis detects keywords and anomalous patterns that may be indicative of fraud. The output is information about the detected fraud-related keywords and anomalous patterns.

[0975] Specific operation: The server inputs the text data into an artificial intelligence model and detects keywords such as "refund," "bank," and "transfer."

[0976] Step 5:

[0977] Fraud Assessment and Notification

[0978] The server evaluates the likelihood of fraud based on the results of text data analysis. It receives keyword detection results as input, scores the likelihood of fraud, and outputs a notification based on the evaluation results. It notifies users and relevant parties via SMS, phone, or email using the Twilio API, or the SendGrid API.

[0979] Specific operation: If the analysis results indicate a high risk of fraud, the server uses the Twilio API to send an SMS warning to Mr. A and his family.

[0980] Step 6:

[0981] Temporary locking of financial institution accounts

[0982] Based on the fraud risk assessment results, the server uses the API (such as Plaid API) of the partner financial institution to send an instruction to temporarily lock the user's financial institution account. The input receives the fraud risk assessment result, and the output is the completion of the lock instruction.

[0983] Specific operation: If it is determined that there is a high possibility of fraud, the server will generate an API request to immediately temporarily lock Mr. A's financial institution account.

[0984] Step 7:

[0985] Identity verification procedures

[0986] The user follows the notification to complete the identity verification procedure. Using the provided interface (verification code entry screen or video call), the user sends the verification information from the device to the server. The server receives the verification code entered by the user and facial recognition data as input, and the output is the completion of identity verification.

[0987] Specific actions: Person A or his / her family follows the instructions in the notification, enters the verification code on the device's browser, and sends it to the server.

[0988] Step 8:

[0989] Unlock

[0990] After the identity verification is successful, the server again uses the financial institution's API to send an unlock instruction. The input is the successful identity verification information, and the output is the unlock of the financial institution account.

[0991] Specific operation: Once identity verification is complete, the server contacts the financial institution via API and unlocks Mr. A's account.

[0992] This process flow allows for real-time detection of potentially fraudulent calls, preventing damage before it occurs, and providing rapid response to protect users' financial assets.

[0993] (Application example 1)

[0994] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0995] In modern society, voice-based fraud is on the rise, and fraud victimization of elderly people is becoming particularly serious. There is also a need for a method to monitor fraudulent call content in real time and respond quickly, but current technology makes this difficult to achieve. With current systems, it takes time for users receiving fraudulent calls to realize they are being scammed, making it difficult to prevent fraud before it occurs. To solve this problem, a system is needed that can assess the possibility of fraud in real time and respond quickly.

[0996] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0997] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties when there is a high possibility of fraud, means for issuing an instruction to temporarily lock the relevant bank account, means for notifying the user of a fraud warning in real time based on the results of the voice recognition and analysis, communication means for monitoring voice call data that is likely to be fraudulent and sending the notification, means for evaluating abnormal voice patterns using a generative AI model, and means for generating prompt sentences to be input to the generative AI model. This makes it possible to monitor the communication lines to be monitored in real time, quickly evaluate and respond to the possibility of fraud, and immediately notify the user and relevant parties.

[0998] 1. "Monitored communication lines" refer to lines that carry communications such as telephone and internet that are subject to monitoring.

[0999] 2. "Monitoring means" means devices or programs that constantly monitor communication lines and collect data.

[1000] 3. "Speech recognition technology" refers to technology that converts voice data into text data.

[1001] 4. "Voice data" refers to sound information collected through telephone conversations, recordings, etc.

[1002] 5. "Text data" means data that has been converted from audio data into text information.

[1003] 6. "Artificial intelligence means" means systems or software that use machine learning or natural language processing to perform specific tasks.

[1004] 7. "Assessing the likelihood of fraud" means analyzing voice and text data to determine whether there is a possibility of fraud.

[1005] 8. "Means of notification" means the means (e.g., telephone, SMS, email, etc.) used to send warnings or information to interested parties.

[1006] 9. "Means for issuing instructions to temporarily lock a bank account" means a means for sending an instruction to an affiliated financial institution to temporarily suspend the use of a user's account.

[1007] 10. "Means for providing real-time fraud alerts to users based on speech recognition and analysis results" means means for instantly sending alerts to users based on analysis results from speech recognition technology and artificial intelligence means.

[1008] 11. "Communication means for monitoring voice call data that may be fraudulent and sending notifications" refers to communication devices or programs that constantly monitor call data that may be fraudulent and notify the results.

[1009] 12. "Means for assessing anomalous speech patterns using a generative AI model" means means for identifying and assessing anomalous speech patterns using a generative AI model.

[1010] 13. "Means for generating prompt sentences to be input to a generative AI model" means means for creating initial input sentences to provide appropriate analytical instructions to a generative AI model.

[1011] The system for implementing this invention mainly comprises a server, a terminal, and a user. The server constantly monitors the communication line to be monitored and receives call voice data transmitted from the terminal in real time. The server converts the received voice data into text data using speech recognition technology. The converted text data is then analyzed using artificial intelligence to detect abnormal patterns and fraud-related keywords in the text.

[1012] Audio data collection and analysis

[1013] The server constantly monitors the target communication line. The device sends the call audio data to the server in real time when the call starts. The server converts this audio data into text data using speech recognition technology (e.g., Google Speech Recognition API). This text data is then analyzed using artificial intelligence means (e.g., machine learning algorithms).

[1014] Fraud Assessment and Notification

[1015] Once the conversion and analysis are complete, the server evaluates the likelihood of fraud. If it determines that there is a high likelihood of fraud, the server notifies the user and relevant parties. Notifications are sent with high speed via phone, SMS, email, etc. The server also issues instructions to temporarily lock the relevant bank account.

[1016] Temporary locking of bank accounts

[1017] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from being leaked to the fraudsters.

[1018] Identity verification procedures

[1019] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a verification code entry screen or facial recognition via video call. After successful identity verification, the server unlocks the bank account.

[1020] Evaluation by generative AI model

[1021] Based on the results of speech recognition and analysis, the server evaluates abnormal speech patterns using a generative AI model, to which an appropriate prompt is input. For example, the following prompt might be used:

[1022] Prompt text: Convert the voice data into text and detect the keywords "refund," "bank," or "transfer" to determine the possibility of fraud. If there are signs of fraud, send a warning to the user via email.

[1023] Specific examples

[1024] Below are some specific scenarios where this invention can be used:

[1025] Preventing refund fraud among the elderly

[1026] 1. Start a call:

[1027] An elderly person, Mr. A, receives a fraudulent phone call and is asked for "information regarding receiving a refund."

[1028] 2. Audio data collection and analysis:

[1029] The device transmits the voice data of the fraudulent call in real time to a server, which uses voice recognition technology to convert the voice data into text data, which is then analyzed using artificial intelligence means.

[1030] 3. Potential Fraud Assessment and Notification:

[1031] The server detects keywords such as "refund," "bank," and "transfer," and determines that there is a high possibility of fraud. The server immediately notifies Mr. A and his family of the possibility of fraud.

[1032] 4. Temporarily locking your bank account:

[1033] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[1034] 5. Identity Verification Procedure:

[1035] The user, Mr. A, or his / her family member, confirms the notification and begins the identity verification process. They enter a verification code using the interface provided by the device, and the server receives the information. After verification, the server contacts the bank and unlocks Mr. A's bank account.

[1036] As described above, this invention allows for real-time monitoring of potentially fraudulent calls and prompt response. This makes it possible to prevent fraudulent calls from affecting both the elderly and young. Furthermore, the safety of users' funds can be ensured by locking their bank accounts and implementing identity verification procedures.

[1037] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1038] Step 1:

[1039] Initiating a call and collecting audio data

[1040] When a terminal initiates a user's call, voice data is collected in real time. This voice data is sent by the terminal to a server. The input is the user's call voice, and the output is the voice data sent to the server. The specific operations include the terminal capturing the voice of the call and sending the data to the server.

[1041] Step 2:

[1042] Converting audio data to text

[1043] The server converts the received voice data into text data using voice recognition technology (e.g., Google Speech Recognition API). The input is the voice data sent from the device, and the output is text data. Specific operations include calling the voice recognition API and converting the voice data into text data.

[1044] Step 3:

[1045] Text data analysis

[1046] The server analyzes the text data using artificial intelligence means (e.g., machine learning algorithms). The input is the text data and the output is the analysis results. Specific operations include using machine learning models to detect anomalous patterns and fraud-related keywords in the text data.

[1047] Step 4:

[1048] Fraud likelihood assessment

[1049] The server evaluates the possibility of fraud based on the analysis results. The input is the analysis results, and the output is the fraud possibility evaluation result. Specific operations include scoring the possibility of fraud based on the analysis results and determining whether the score exceeds a threshold.

[1050] Step 5:

[1051] Sending notifications

[1052] If it determines that there is a high possibility of fraud, the server notifies the user and relevant parties. The input is the fraud possibility assessment result, and the output is a notification message. Specific operations include sending warnings to the user and relevant parties via email, SMS, phone, etc.

[1053] Step 6:

[1054] Temporary locking of bank accounts

[1055] The server issues an instruction to temporarily lock the relevant bank account. The input is the fraud probability assessment result and user information, and the output is a lock instruction to be sent to the bank. Specific operations include sending a request to temporarily lock the account to the partner bank's API.

[1056] Step 7:

[1057] Start the identity verification process

[1058] Upon receiving the notification, the user begins the identity verification process. The device then provides a screen for entering a verification code or facial recognition via video call. The input is the notification message and user information, and the output is progress information on identity verification. Specific actions include the user entering a verification code or performing facial recognition via video call.

[1059] Step 8:

[1060] Unlocking a bank account

[1061] After successful identity verification, the server issues a command to unlock the bank account. The input is the successful identity verification information, and the output is an unlock command sent to the bank. Specific operations include sending an unlock request to the partner bank's API.

[1062] Step 9:

[1063] Analyzing abnormal voice patterns using AI models

[1064] The server analyzes abnormal voice patterns using a generative AI model. The input is the voice recognition and analysis results, and the output is the evaluation result of the abnormal voice pattern. Specific operations include inputting appropriate prompt sentences to the generative AI model and evaluating the abnormal voice pattern.

[1065] Step 10:

[1066] Generate prompt statement

[1067] The server generates prompt sentences to be input to the generative AI model. The inputs are the AI ​​model specifications and the information to be analyzed, and the output is the prompt sentence. Specific operations include creating prompt sentences appropriate to the situation to be analyzed (e.g., "Convert the voice data into text and detect any of the keywords 'refund,' 'bank,' or 'transfer' to determine the possibility of fraud. If there are signs of fraud, send a warning to the user via email.").

[1068] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1069] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing the user's emotions. This system mainly consists of a server, a terminal, and a user, and operates as follows.

[1070] Audio data collection and analysis

[1071] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence tools to detect abnormal patterns and fraud-related keywords.

[1072] Emotion recognition by emotion engine

[1073] The server then passes the call audio data to an emotion engine to recognize the user's emotions. This emotion engine assesses whether the user is experiencing stress or anxiety. The user's emotion data is then used to augment the assessment of the likelihood of fraud.

[1074] Fraud Assessment and Notification

[1075] If the analysis determines that there is a high possibility of fraud, the server will raise a fraud warning flag and immediately notify the user and relevant parties via phone, SMS, email, etc. to quickly warn them.

[1076] Temporary locking of bank accounts

[1077] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from reaching the fraudsters.

[1078] Identity verification procedures

[1079] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a screen for entering a verification code or facial recognition via video call. After successfully verifying the user's identity, the server unlocks the bank account.

[1080] Specific examples

[1081] Scenario: Senior Citizen Refund Fraud

[1082] 1. Start a call

[1083] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[1084] 2. Collection and analysis of audio data

[1085] The device transmits the audio data of the fraudulent call to the server in real time.

[1086] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[1087] At the same time, the server uses an emotion engine to analyze Mr. A's emotional state and assess his level of stress and anxiety.

[1088] 3. Fraud Assessment and Notification

[1089] The server detects keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on the results of AI analysis and emotional data.

[1090] The server immediately notifies Mr. A and his family of the possible fraud.

[1091] 4. Temporarily locking your bank account

[1092] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[1093] 5. Identity Verification Procedures

[1094] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[1095] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[1096] After verification, the server contacts the bank and unlocks Mr. A's bank account.

[1097] This invention makes it possible to detect fraudulent phone calls suffered by the elderly and young in real time, and by utilizing user emotion data, it is possible to improve the accuracy of fraud detection and prevent damage by quickly notifying relevant parties.In addition, the safety of users' funds can be ensured by locking bank accounts and implementing identity verification procedures.

[1098] The processing flow will be explained below.

[1099] Step 1:

[1100] The terminal detects that the user has started a telephone call, and simultaneously starts collecting the call voice data.

[1101] Step 2:

[1102] The terminal transmits the collected call voice data to the server in real time.

[1103] Step 3:

[1104] The server passes the received voice data to a voice recognition engine and converts the voice data into text data.

[1105] Step 4:

[1106] The server passes the converted text data to an AI analysis module to detect abnormal patterns and fraud-related keywords within the text data.

[1107] Step 5:

[1108] At the same time, the server passes the voice data to the emotion engine, which analyzes the user's emotions by analyzing the tone, tempo, and speaking style of the voice to assess whether the user is feeling stressed or anxious.

[1109] Step 6:

[1110] The server makes a comprehensive judgment based on the results of the AI ​​analysis and the evaluation results of the emotion engine, and if there is a high possibility of fraud, it raises a fraud warning flag.

[1111] Step 7:

[1112] If a fraud alert flag is raised, the server will immediately notify the user and relevant parties via phone, SMS, email, or other means.

[1113] Step 8:

[1114] The server instructs the user's designated bank to temporarily lock the user's bank account.

[1115] Step 9:

[1116] The user receives the notification and starts the identity verification procedure according to the specified procedure. The terminal provides an interface for identity verification.

[1117] Step 10:

[1118] The device sends the verification code entered by the user and the facial recognition results obtained via video call to the server.

[1119] Step 11:

[1120] The server receives the authentication information and verifies that the user is a legitimate user.

[1121] Step 12:

[1122] If the verification is successful, the server instructs the bank to unlock the user's bank account.

[1123] This process allows for real-time detection of potentially fraudulent calls, and prompt notification and response. The emotion engine also enables advanced fraud detection that takes into account the user's emotional state.

[1124] Example 2

[1125] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1126] Conventional communication monitoring systems simply convert voice data into text data and detect fraud through simple keyword matching, but this method is insufficient to counter increasingly sophisticated fraud techniques. Furthermore, because the system ignores the user's emotions and psychological state, the accuracy of fraud detection can be reduced. Furthermore, it is difficult to respond quickly after fraud is detected, making it difficult to effectively prevent users from losing funds.

[1127] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1128] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, emotion recognition means for analyzing the call voice data and recognizing the user's emotion, means for notifying relevant parties when there is a high possibility of fraud, means for issuing instructions to temporarily lock the relevant bank account, and interface means for carrying out identity verification procedures. This enables highly sophisticated fraud to be detected with high accuracy, and by utilizing user emotion data, the accuracy of fraud detection is improved, enabling quick and effective response.

[1129] A "monitored communication line" is a communication infrastructure that is monitored and through which information is sent and received.

[1130] "Speech recognition technology" is a technology for converting voice data into text data, and typically uses an algorithm to extract linguistic information from voice.

[1131] "Voice data" refers to data that is a digital recording of the voice generated during a call.

[1132] "Text data" refers to the format of character information obtained after voice data is converted using voice recognition technology.

[1133] "Artificial intelligence means" means artificial intelligence-based techniques for analyzing the converted text data and assessing any abnormal patterns or potential fraud.

[1134] The "emotion recognition means" is a system that analyzes call voice data and evaluates the user's emotions and psychological state.

[1135] "Notification means" refers to methods or tools used to warn or inform relevant parties when a high possibility of fraud is determined.

[1136] A "bank account locking mechanism" is a system that executes instructions to temporarily freeze a user's bank account if fraud is suspected.

[1137] "Interface means" refers to input and output means for users and related parties to interact with the system, and includes, for example, a confirmation code input screen and a function for video calling.

[1138] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing user emotions. This system mainly consists of three components: a server, a terminal, and a user. Specific embodiments of the system are described below.

[1139] Audio data collection and analysis

[1140] The server constantly monitors the target communication line. Network monitoring software (e.g., Wireshark or PRTG) is used for monitoring. When a call is initiated, the device sends the call audio data to the server in real time. A streaming API (e.g., WebRTC) is used to collect the audio data, minimizing the delay in the audio data.

[1141] The server converts the received voice data into text data using the Google Cloud Speech-to-Text API. This conversion process is performed in parallel, and text is generated in real time.

[1142] The server then analyzes the converted text data using natural language processing (NLP) techniques such as GPT-3, which can identify unusual patterns and fraud-related keywords (e.g., "refund," "bank," "transfer," etc.).

[1143] Emotion recognition by emotion engine

[1144] The server then passes the call audio data to an emotion recognition engine (Emotion AI) to recognize the user's emotions. The emotion engine analyzes the tone and pace of the voice to assess whether the user is feeling stressed or anxious. For example, a trembling voice or a high-pitched tone can be detected as a sign of stress.

[1145] Fraud Assessment and Notification

[1146] The server combines the analysis results with sentiment data to assess the likelihood of fraud, and if it determines that fraud is likely, it raises a fraud warning flag.

[1147] If a potential fraud is detected, the server immediately notifies the user and relevant parties. The notification is sent in real time via SMS or email, allowing the user and relevant parties to take immediate action. For example, the notification may say, "A potential fraudulent call has been detected. Please take action immediately."

[1148] Temporary locking of bank accounts

[1149] After the fraud notification is complete, the server sends an instruction to the partner financial institution via API to temporarily lock the user's bank account, thereby preventing the outflow of funds.

[1150] Identity verification procedures

[1151] Upon receiving the notification, the user begins the identity verification process. The device provides facial recognition via a verification code input screen or video call. For example, the user receives a verification code via SMS and enters it on the device's input screen. Alternatively, facial recognition is performed via a video call to verify the user's identity. The server then contacts the financial institution and unlocks the bank account.

[1152] Specific examples

[1153] For example, consider the case of an elderly person named Mr. A who receives a call about a tax refund scam. The moment the call starts, the device sends a call start message to the server via the API, and captures and transmits voice data in real time.

[1154] The server converts the voice data into text using the Google Cloud Speech-to-Text API and analyzes the text using GPT-3. At the same time, it uses an emotion engine to analyze Mr. A's emotional state, detecting that he is experiencing increased stress and anxiety.

[1155] The server detects fraud-related keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on AI analysis and emotional data. It immediately notifies Mr. A and his family of the possibility of fraud and instructs the financial institution to temporarily lock Mr. A's bank account.

[1156] Upon receiving the notification, Mr. A or his family members can use the device interface to complete identity verification procedures, and once verification is complete, the bank account will be unlocked.

[1157] Prompt Sentence Examples

[1158] "They mentioned a refund, is this a scam?"

[1159] "How can I identify a fraudulent phone call?"

[1160] "Please let me know the analysis results of calls that are likely to be fraudulent."

[1161] The present invention makes it possible to detect fraudulent phone call damage in real time, improve the accuracy of fraud detection by utilizing user emotion data, and respond quickly.

[1162] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1163] Step 1:

[1164] The server constantly monitors the target communication line. The input is the connection information of the communication line, and the output is the call start trigger. Specifically, it uses network monitoring software (e.g., Wireshark or PRTG) to detect the start of a call in real time.

[1165] Step 2:

[1166] When a call is started, the device notifies the server of the call start information. The input is a call start event, and the output is a notification message to the server. Specifically, the device sends a call start message to the server through the API. This message contains the call identification information.

[1167] Step 3:

[1168] The device captures the call audio data in real time and sends it to the server. The input is the call audio data, and the output is a stream of audio data to the server. Specifically, it uses a streaming API (e.g., WebRTC) to capture the audio data and sends it to the server in an encrypted format.

[1169] Step 4:

[1170] The server converts the received voice data into text data using the Google Cloud Speech-to-Text API. The input is voice data and the output is text data. Specifically, the server sends voice data to the API and generates text data in real time. Parallel processing is used to improve conversion efficiency.

[1171] Step 5:

[1172] The server analyzes the converted text data using natural language processing technology such as GPT-3. The input is text data, and the output is the analysis results. Specifically, the text data is input into a generative AI model to detect fraud-related keywords such as "refund," "bank," and "transfer."

[1173] Step 6:

[1174] The server passes the call voice data to the Emotion AI engine, which analyzes the user's emotions. The input is voice data, and the output is emotional evaluation data. Specifically, it analyzes the voice tone and pace to evaluate whether the user is feeling stressed or anxious.

[1175] Step 7:

[1176] The server integrates the analysis results and emotion data to assess the likelihood of fraud. The input is the text analysis results and emotion evaluation data, and the output is a fraud warning flag. Specifically, it determines the likelihood of fraud based on abnormal patterns, fraud-related keywords, and user emotion data.

[1177] Step 8:

[1178] If it is determined that there is a high possibility of fraud, the server will notify the relevant parties. The input is a fraud warning flag and the output is a notification message. Specifically, a notification is sent to the relevant parties via SMS or email. The message sent will say something like, "A call with a high possibility of fraud has been detected. Please take immediate action."

[1179] Step 9:

[1180] After the fraud notification is complete, the server issues an instruction to the partner financial institution to temporarily lock the user's bank account. The input is the fraud warning flag and the user's bank information, and the output is a lock instruction to the bank. Specifically, the server sends the lock instruction to the bank via API to ensure the safety of the user's funds.

[1181] Step 10:

[1182] Upon receiving the notification, the user begins the identity verification procedure. The input is the notification message, and the output is the identity verification result. Specifically, the procedure is carried out using the verification code input screen or video call function provided by the device. After verification is complete, the server contacts the bank and unlocks the bank account.

[1183] (Application example 2)

[1184] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1185] In recent years, the number of victims of fraudulent phone calls targeting the elderly and young people has been increasing. These fraudulent acts have become more diverse, making it difficult to deal with them using conventional fraud detection methods alone. Furthermore, since the damage often occurs before the user even realizes it, new technologies are needed to respond to fraud early.

[1186] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties when there is a high possibility of fraud, means for issuing an instruction to temporarily lock the relevant trading account, and emotion recognition means for evaluating the user's emotional state. This enables highly accurate fraud detection and rapid response that takes the user's emotional state into account.

[1187] A "monitored communication line" is a line on a communication network that is continuously monitored to monitor activity on specific calls or data traffic.

[1188] "Voice recognition technology" is a technology that captures voice as a digital signal and converts it into text data or commands.

[1189] "Text data" is character information converted using speech recognition technology, and is the data that is the subject of analysis.

[1190] "Artificial intelligence means" are means that use technologies such as machine learning and deep learning to analyze data and detect patterns and anomalies.

[1191] "Means for notifying relevant parties" refers to means for conveying warnings and information to users and relevant third parties when possible fraud is detected.

[1192] "Means of issuing instructions to temporarily lock a trading account" refers to a means of instructing a bank or financial institution to freeze a trading account in order to prevent further damage when fraud is suspected.

[1193] "Emotion recognition means" refers to a technical means for assessing a user's emotional state from voice and other data, and determining the level of stress or anxiety.

[1194] The "interface means for carrying out identity verification procedures" refers to means including input devices and software for the user to verify their identity.

[1195] An "anomalous voice pattern" refers to a unique voice characteristic or word usage that differs from normal speech, and is a pattern that is an indicator of possible fraud.

[1196] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing the user's emotions. This system mainly consists of a server, a terminal, and a user, and operates as follows.

[1197] Audio data collection and analysis

[1198] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence means to detect abnormal patterns and fraud-related keywords. This identification is performed using a generative AI model.

[1199] Emotion recognition by emotion engine

[1200] The server then passes the call audio data to an emotion recognition engine to recognize the user's emotions. This emotion engine assesses whether the user is experiencing stress or anxiety. The user's emotion data is then used to augment the assessment of the likelihood of fraud.

[1201] Fraud Assessment and Notification

[1202] If the analysis determines that there is a high possibility of fraud, the server will raise a fraud warning flag and immediately notify the user and relevant parties via phone, SMS, email, etc. to quickly warn them.

[1203] Temporary locking of trading accounts

[1204] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's trading account, preventing the user's funds from being leaked to the fraudsters.

[1205] Identity verification procedures

[1206] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device will provide a verification code entry screen or face recognition via video call. After successful identity verification, the server will unlock the trading account.

[1207] Specific examples

[1208] Scenario: Senior Citizen Refund Fraud

[1209] 1. Start a call

[1210] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[1211] 2. Collection and analysis of audio data

[1212] The device transmits the audio data of the fraudulent call to the server in real time.

[1213] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[1214] At the same time, the server uses an emotion engine to analyze Mr. A's emotional state and assess his level of stress and anxiety.

[1215] 3. Fraud Assessment and Notification

[1216] The server detects keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on the results of AI analysis and emotional data.

[1217] The server immediately notifies Mr. A and his family of the possible fraud.

[1218] 4. Temporary Locking of Trading Account

[1219] The server instructs the partner bank to temporarily lock Mr. A's trading account.

[1220] 5. Identity Verification Procedures

[1221] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[1222] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[1223] After confirmation, the server contacts the bank and unlocks Mr. A's trading account.

[1224] This technology can detect fraudulent phone calls in real time, and by utilizing user emotional data, it is possible to improve the accuracy of fraud detection and prevent damage by quickly notifying relevant parties.In addition, the safety of users' funds can be ensured by locking their trading accounts and implementing identity verification procedures.

[1225] An example of a prompt to be input to the generative AI model is as follows:

[1226] "Develop a system that analyzes human emotions and detects fraudulent phone calls. Design an application that detects calls containing keywords such as 'refund,' 'bank,' and 'transfer,' as well as the user's feelings of stress and anxiety, and displays a warning if there is a high possibility of fraud."

[1227] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1228] Step 1:

[1229] The terminal starts collecting the call voice data when the call starts. The terminal transmits the collected voice data to the server in real time. The input is the call voice data, and the output is the transmitted voice data.

[1230] Step 2:

[1231] The server converts the received voice data into text data using voice recognition technology. Specifically, the server converts the voice data into text data using a voice recognition engine (e.g., Google Speech-to-Text API). The input is voice data, and the output is text data.

[1232] Step 3:

[1233] The server analyzes the converted text data using artificial intelligence means to detect fraud-related keywords and abnormal patterns. Specifically, the server uses natural language processing (NLP) techniques and deep learning models to extract fraud-related keywords (e.g., "refund," "bank," "transfer," etc.) from the text data. The input is the text data, and the output is the analysis results.

[1234] Step 4:

[1235] At the same time, the server passes the call voice data to an emotion recognition engine to evaluate the user's emotional state. Specifically, the server uses a voice emotion recognition model (e.g., IBM Watson Tone Analyzer) to evaluate emotional states such as stress and anxiety from the voice data. The input is voice data, and the output is emotional state data.

[1236] Step 5:

[1237] The server comprehensively evaluates the text data analysis results and the emotional state data to determine the likelihood of fraud. If the likelihood of fraud is high, the server raises a fraud warning flag. In this step, the text analysis results and the emotional state data are used as inputs, and a fraud warning flag is generated as output.

[1238] Step 6:

[1239] The server notifies users and other interested parties when a fraud alert flag is raised. Specifically, the server uses a notification method (e.g., phone call, SMS, email) to send a message informing them of a possible fraud. The input is the fraud alert flag, and the output is the notification message sent.

[1240] Step 7:

[1241] After the server is notified of the fraud, it instructs the affiliated financial institution to temporarily lock the user's trading account. The input is a fraud warning flag, and the output is an instruction to lock the trading account.

[1242] Step 8:

[1243] The user receives the fraud notification and starts the identity verification procedure using the device. Specifically, the device provides a verification code input screen and facial recognition functionality via video call. The input is the verification code and facial recognition data, and the output is the identity verification result.

[1244] Step 9:

[1245] After the identity verification procedure is successful, the server contacts the financial institution and unlocks the trading account. The input is the identity verification result, and the output is the instruction to unlock the trading account.

[1246] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1247] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1248] 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 the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1249] [Fourth embodiment]

[1250] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1251] 7, a 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.

[1252] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1253] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1254] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1255] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1256] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1257] The control object 443 includes a display device, LEDs in the eyes, and motors for driving 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 emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1258] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1259] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[1260] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1261] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1262] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1263] The present invention is a system that monitors communication lines to be monitored, detects calls that are likely to be fraudulent, and prevents damage before they occur. This system is mainly composed of a server, a terminal, and a user, and operates as follows.

[1264] Audio data collection and analysis

[1265] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence means to detect abnormal patterns and fraud-related keywords in the text.

[1266] Fraud Assessment and Notification

[1267] After conversion and analysis, the server evaluates the likelihood of fraud. If it determines that there is a high likelihood of fraud, the server notifies the relevant user and relevant parties. The notification is sent via phone, SMS, email, etc., to provide a comprehensive warning.

[1268] Temporary locking of bank accounts

[1269] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from being leaked to the fraudsters.

[1270] Identity verification procedures

[1271] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a verification code entry screen or facial recognition via video call. After successful identity verification, the server unlocks the bank account.

[1272] Specific examples

[1273] Scenario: Senior Citizen Refund Fraud

[1274] 1. Start a call

[1275] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[1276] 2. Collection and analysis of audio data

[1277] The device transmits the audio data of the fraudulent call to the server in real time.

[1278] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[1279] 3. Fraud Assessment and Notification

[1280] The server detects keywords such as "refund," "bank," and "transfer" and determines that there is a high possibility of fraud.

[1281] The server immediately notifies Mr. A and his family of the possible fraud.

[1282] 4. Temporarily locking your bank account

[1283] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[1284] 5. Identity Verification Procedures

[1285] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[1286] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[1287] After verification, the server contacts the bank and unlocks Mr. A's bank account.

[1288] This invention makes it possible to detect fraudulent phone calls to elderly and young people in real time, and to prevent damage by quickly notifying the relevant parties. It also ensures the safety of users' funds by locking their bank accounts and implementing identity verification procedures.

[1289] The processing flow will be explained below.

[1290] Step 1:

[1291] The terminal detects that the user has started a telephone call, and simultaneously starts collecting the call voice data.

[1292] Step 2:

[1293] The terminal transmits the collected call voice data to the server in real time.

[1294] Step 3:

[1295] The server passes the received voice data to a voice recognition engine and converts the voice data into text data.

[1296] Step 4:

[1297] The server passes the converted text data to an AI analysis module to detect abnormal patterns and fraud-related keywords within the text data.

[1298] Step 5:

[1299] The server will raise a fraud warning flag if AI analysis determines that fraud is likely.

[1300] Step 6:

[1301] If a fraud alert flag is raised, the server will immediately notify the user and relevant parties via phone, SMS, email, or other means.

[1302] Step 7:

[1303] The server instructs the user's designated bank to temporarily lock the user's bank account.

[1304] Step 8:

[1305] The user receives the notification and starts the identity verification procedure according to the specified procedure. The terminal provides an interface for identity verification.

[1306] Step 9:

[1307] The device sends the verification code entered by the user and the facial recognition results obtained via video call to the server.

[1308] Step 10:

[1309] The server receives the authentication information and verifies that the user is a legitimate user.

[1310] Step 11:

[1311] If the verification is successful, the server instructs the bank to unlock the user's bank account.

[1312] This series of processes allows for the real-time detection of potentially fraudulent calls and prompt notification and response.

[1313] Example 1

[1314] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1315] In recent years, fraud methods using telecommunications lines have become more sophisticated and diverse. Refund fraud and bank transfer fraud targeting the elderly in particular have been on the rise, causing serious damage. Conventional crime prevention methods have struggled to detect fraudulent activity in real time and prevent damage before it occurs. There is also a lack of methods for quickly and accurately assessing the possibility of fraud and notifying relevant parties. In addition, to prevent fraud once it has occurred, there is a need for a function to quickly lock financial institution accounts and a method for users to easily complete identity verification procedures.

[1316] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1317] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties if a fraudulent call is highly likely, means for issuing instructions to temporarily lock the relevant financial institution's account, means for unlocking the account by providing personal identification information, means for transmitting voice data in real time when the call is made, means for detecting fraud keywords based on the analysis results, and means for transmitting notifications using multiple communication means. This makes it possible to detect potentially fraudulent calls in real time and prevent damage before it occurs. Furthermore, the relevant financial institution's account can be immediately locked and safely unlocked through personal identification procedures.

[1318] A "monitored communication line" is a communication network that is constantly monitored to detect potentially fraudulent calls.

[1319] "Speech recognition technology" is a technology used to convert voice data into text data.

[1320] "Voice data" means voice information, in analog or digital form, captured during a telephone call.

[1321] "Text data" is character information converted from voice data using voice recognition technology.

[1322] "Artificial intelligence means" are machine learning or deep learning algorithms used to analyze text data and detect specific patterns or keywords.

[1323] "Fraudulent keywords" are words or phrases that may be associated with fraudulent activity.

[1324] "Assessing the likelihood of fraud" means determining the risk of fraud based on pattern and keyword analysis within text data.

[1325] "Notifying" means sending warnings and information to users and other interested parties when fraud is likely.

[1326] "Interested Persons" are people or entities that should receive information about fraud risks.

[1327] A "financial institution account," or simply an "account," is an account managed by a financial institution such as a bank that records a user's deposits and funds.

[1328] "Temporarily locking" means suspending activities such as withdrawals from a financial institution account for a certain period of time.

[1329] "Personal identification information" refers to information or data used to prove the identity of a user.

[1330] "Unlocking" means lifting the suspension of an account at a financial institution after identity verification has been completed.

[1331] "Sending in real time when a call is made" means sending voice data to the server immediately at the moment the call is started.

[1332] "Detecting fraud keywords based on the analysis results" means that the artificial intelligence means analyzes the text data and identifies specific words related to fraudulent activities.

[1333] "Sending notifications using multiple communication methods" means sending notifications using different media such as telephone, SMS, and email.

[1334] The present invention relates to a system for monitoring communication lines to detect calls that are likely to be fraudulent and prevent damage before they occur. This system uses a server, a terminal, and a user as its main components.

[1335] The server uses high-performance communication monitoring software (such as WireShark or Snort) to constantly monitor the communication line. When a call is initiated, the device sends the call audio data to the server in real time. The audio data is encoded using VoIP technology (such as the SIP protocol) and sent via a secure communication protocol such as HTTPS.

[1336] The server converts the received voice data into text using the Google Cloud Speech-to-Text API. It then uses speech recognition technology to obtain accurate text data, and analyzes the converted text data using generative AI models such as BERT and GPT-3. The analysis aims to detect abnormal patterns and fraud-related keywords in the text (e.g., "refund," "bank," and "transfer").

[1337] If a fraud risk is determined to be high, the server will notify the relevant parties (e.g., the user and their family) via SMS or phone using the Twilio API. It is also possible to notify them by email using the SendGrid API. This notification allows users to quickly be informed of the risk of fraud.

[1338] After sending the fraud notification, the server uses the API of the partner financial institution (e.g., Plaid API) to issue a command to temporarily lock the user's financial institution account, thereby preventing the user's funds from being stolen due to fraudulent activity.

[1339] When the user receives the fraud notification, they use the provided interface (e.g., a verification code entry screen or video call) to verify their identity. The device provides the interface for identity verification through the API of Jumio or IDNow. The user completes identity verification by entering a verification code or by performing facial recognition via a video call. Once verification is complete, the server again uses the financial institution's API to send instructions to unlock the bank account.

[1340] Specific examples

[1341] Scenario: Senior Citizen Refund Fraud

[1342] 1. Start a call

[1343] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[1344] 2. Collection and analysis of audio data

[1345] The device (smartphone) transmits the audio data of the fraudulent call to the server in real time.

[1346] The server converts the received voice data into text data using Google Cloud Speech-to-Text.

[1347] The converted text data is analyzed using the BERT model to detect keywords such as "refund," "bank," and "transfer."

[1348] 3. Fraud Assessment and Notification

[1349] Based on the keywords detected, the server determines with a high probability that the request is fraudulent.

[1350] The server uses the Twilio API to notify Mr. A and his family via SMS of the possible fraud.

[1351] 4. Temporary locking of financial institution accounts

[1352] The server sends an instruction to the partner bank via API to temporarily lock Mr. A's financial institution account.

[1353] 5. Identity Verification Procedures

[1354] The user, Mr. A, or his / her family member clicks on the link provided in the notification to begin the identity verification process.

[1355] Enter the verification code in the interface provided by your device or use facial recognition via video call.

[1356] After the verification is complete, the server again uses the API to send instructions to unlock the bank account.

[1357] Example prompts for generative AI models

[1358] Using the following system scenario, please explain in natural language how a program to detect potentially fraudulent calls works:

[1359] system:

[1360] The server monitors the communication line and analyzes the voice data.

[1361] The terminal transmits the call voice data to the server.

[1362] The server uses artificial intelligence to analyze the converted text data and assess the likelihood of fraud.

[1363] If fraud is likely, the server will notify the user and lock the financial institution account.

[1364] The user goes through the identity verification procedure and the server unlocks the device.

[1365] Example: Senior Citizen Refund Fraud Scenario

[1366] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1367] Specific explanation of the program based on processing steps

[1368] Step 1:

[1369] Call detection

[1370] The terminal monitors the target communication line and detects when a call has started. As input, it receives a call signal from the communication line and analyzes this signal to detect the call start trigger. The output is event information that a call has started, which starts the subsequent processing.

[1371] Specific operation: When elderly person A answers the phone, the device's monitoring app detects the start of the call and immediately begins collecting voice data.

[1372] Step 2:

[1373] Collection and transmission of voice data

[1374] The device receives call start event information and collects real-time call audio data, which is then encoded using VoIP technology and sent to the server via a secure communication protocol (HTTPS).

[1375] As input, it receives real-time audio data acquired by the terminal, and as output, it transmits encoded audio data to the server.

[1376] Specific operation: While Person A continues the conversation, the device captures audio data and transmits it to the server through a secure channel.

[1377] Step 3:

[1378] Converting audio data to text

[1379] The server receives the voice data sent from the device and converts it into text data using the Google Cloud Speech-to-Text API. The input is the encoded voice data, and the output is the converted text data.

[1380] Specific operation: The server calls the Google Cloud Speech-to-Text API to convert the audio data into text data.

[1381] Step 4:

[1382] Text data analysis

[1383] The server receives text data from the Google Cloud Speech-to-Text API and analyzes it using artificial intelligence models such as BERT and GPT-3. The analysis detects keywords and anomalous patterns that may be indicative of fraud. The output is information about the detected fraud-related keywords and anomalous patterns.

[1384] Specific operation: The server inputs the text data into an artificial intelligence model and detects keywords such as "refund," "bank," and "transfer."

[1385] Step 5:

[1386] Fraud Assessment and Notification

[1387] The server evaluates the likelihood of fraud based on the results of text data analysis. It receives keyword detection results as input, scores the likelihood of fraud, and outputs a notification based on the evaluation results. It notifies users and relevant parties via SMS, phone, or email using the Twilio API, or the SendGrid API.

[1388] Specific operation: If the analysis results indicate a high risk of fraud, the server uses the Twilio API to send an SMS warning to Mr. A and his family.

[1389] Step 6:

[1390] Temporary locking of financial institution accounts

[1391] Based on the fraud risk assessment results, the server uses the API (such as Plaid API) of the partner financial institution to send an instruction to temporarily lock the user's financial institution account. The input receives the fraud risk assessment result, and the output is the completion of the lock instruction.

[1392] Specific operation: If it is determined that there is a high possibility of fraud, the server will generate an API request to immediately temporarily lock Mr. A's financial institution account.

[1393] Step 7:

[1394] Identity verification procedures

[1395] The user follows the notification to complete the identity verification procedure. Using the provided interface (verification code entry screen or video call), the user sends the verification information from the device to the server. The server receives the verification code entered by the user and facial recognition data as input, and the output is the completion of identity verification.

[1396] Specific actions: Person A or his / her family follows the instructions in the notification, enters the verification code on the device's browser, and sends it to the server.

[1397] Step 8:

[1398] Unlock

[1399] After the identity verification is successful, the server again uses the financial institution's API to send an unlock instruction. The input is the successful identity verification information, and the output is the unlock of the financial institution account.

[1400] Specific operation: Once identity verification is complete, the server contacts the financial institution via API and unlocks Mr. A's account.

[1401] This process flow allows for real-time detection of potentially fraudulent calls, preventing damage before it occurs, and providing rapid response to protect users' financial assets.

[1402] (Application example 1)

[1403] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1404] In modern society, voice-based fraud is on the rise, and fraud victimization of elderly people is becoming particularly serious. There is also a need for a method to monitor fraudulent call content in real time and respond quickly, but current technology makes this difficult to achieve. With current systems, it takes time for users receiving fraudulent calls to realize they are being scammed, making it difficult to prevent fraud before it occurs. To solve this problem, a system is needed that can assess the possibility of fraud in real time and respond quickly.

[1405] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1406] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties when there is a high possibility of fraud, means for issuing an instruction to temporarily lock the relevant bank account, means for notifying the user of a fraud warning in real time based on the results of the voice recognition and analysis, communication means for monitoring voice call data that is likely to be fraudulent and sending the notification, means for evaluating abnormal voice patterns using a generative AI model, and means for generating prompt sentences to be input to the generative AI model. This makes it possible to monitor the communication lines to be monitored in real time, quickly evaluate and respond to the possibility of fraud, and immediately notify the user and relevant parties.

[1407] 1. "Monitored communication lines" refer to lines that carry communications such as telephone and internet that are subject to monitoring.

[1408] 2. "Monitoring means" means devices or programs that constantly monitor communication lines and collect data.

[1409] 3. "Speech recognition technology" refers to technology that converts voice data into text data.

[1410] 4. "Voice data" refers to sound information collected through telephone conversations, recordings, etc.

[1411] 5. "Text data" means data that has been converted from audio data into text information.

[1412] 6. "Artificial intelligence means" means systems or software that use machine learning or natural language processing to perform specific tasks.

[1413] 7. "Assessing the likelihood of fraud" means analyzing voice and text data to determine whether there is a possibility of fraud.

[1414] 8. "Means of notification" means the means (e.g., telephone, SMS, email, etc.) used to send warnings or information to interested parties.

[1415] 9. "Means for issuing instructions to temporarily lock a bank account" means a means for sending an instruction to an affiliated financial institution to temporarily suspend the use of a user's account.

[1416] 10. "Means for providing real-time fraud alerts to users based on speech recognition and analysis results" means means for instantly sending alerts to users based on analysis results from speech recognition technology and artificial intelligence means.

[1417] 11. "Communication means for monitoring voice call data that may be fraudulent and sending notifications" refers to communication devices or programs that constantly monitor call data that may be fraudulent and notify the results.

[1418] 12. "Means for assessing anomalous speech patterns using a generative AI model" means means for identifying and assessing anomalous speech patterns using a generative AI model.

[1419] 13. "Means for generating prompt sentences to be input to a generative AI model" means means for creating initial input sentences to provide appropriate analytical instructions to a generative AI model.

[1420] The system for implementing this invention mainly comprises a server, a terminal, and a user. The server constantly monitors the communication line to be monitored and receives call voice data transmitted from the terminal in real time. The server converts the received voice data into text data using speech recognition technology. The converted text data is then analyzed using artificial intelligence to detect abnormal patterns and fraud-related keywords in the text.

[1421] Audio data collection and analysis

[1422] The server constantly monitors the target communication line. The device sends the call audio data to the server in real time when the call starts. The server converts this audio data into text data using speech recognition technology (e.g., Google Speech Recognition API). This text data is then analyzed using artificial intelligence means (e.g., machine learning algorithms).

[1423] Fraud Assessment and Notification

[1424] Once the conversion and analysis are complete, the server evaluates the likelihood of fraud. If it determines that there is a high likelihood of fraud, the server notifies the user and relevant parties. Notifications are sent with high speed via phone, SMS, email, etc. The server also issues instructions to temporarily lock the relevant bank account.

[1425] Temporary locking of bank accounts

[1426] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from being leaked to the fraudsters.

[1427] Identity verification procedures

[1428] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a verification code entry screen or facial recognition via video call. After successful identity verification, the server unlocks the bank account.

[1429] Evaluation by generative AI model

[1430] Based on the results of speech recognition and analysis, the server evaluates abnormal speech patterns using a generative AI model, to which an appropriate prompt is input. For example, the following prompt might be used:

[1431] Prompt text: Convert the voice data into text and detect the keywords "refund," "bank," or "transfer" to determine the possibility of fraud. If there are signs of fraud, send a warning to the user via email.

[1432] Specific examples

[1433] Below are some specific scenarios where this invention can be used:

[1434] Preventing refund fraud among the elderly

[1435] 1. Start a call:

[1436] An elderly person, Mr. A, receives a fraudulent phone call and is asked for "information regarding receiving a refund."

[1437] 2. Audio data collection and analysis:

[1438] The device transmits the voice data of the fraudulent call in real time to a server, which uses voice recognition technology to convert the voice data into text data, which is then analyzed using artificial intelligence means.

[1439] 3. Potential Fraud Assessment and Notification:

[1440] The server detects keywords such as "refund," "bank," and "transfer," and determines that there is a high possibility of fraud. The server immediately notifies Mr. A and his family of the possibility of fraud.

[1441] 4. Temporarily locking your bank account:

[1442] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[1443] 5. Identity Verification Procedure:

[1444] The user, Mr. A, or his / her family member, confirms the notification and begins the identity verification process. They enter a verification code using the interface provided by the device, and the server receives the information. After verification, the server contacts the bank and unlocks Mr. A's bank account.

[1445] As described above, this invention allows for real-time monitoring of potentially fraudulent calls and prompt response. This makes it possible to prevent fraudulent calls from affecting both the elderly and young. Furthermore, the safety of users' funds can be ensured by locking their bank accounts and implementing identity verification procedures.

[1446] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1447] Step 1:

[1448] Initiating a call and collecting audio data

[1449] When a terminal initiates a user's call, voice data is collected in real time. This voice data is sent by the terminal to a server. The input is the user's call voice, and the output is the voice data sent to the server. The specific operations include the terminal capturing the voice of the call and sending the data to the server.

[1450] Step 2:

[1451] Converting audio data to text

[1452] The server converts the received voice data into text data using voice recognition technology (e.g., Google Speech Recognition API). The input is the voice data sent from the device, and the output is text data. Specific operations include calling the voice recognition API and converting the voice data into text data.

[1453] Step 3:

[1454] Text data analysis

[1455] The server analyzes the text data using artificial intelligence means (e.g., machine learning algorithms). The input is the text data and the output is the analysis results. Specific operations include using machine learning models to detect anomalous patterns and fraud-related keywords in the text data.

[1456] Step 4:

[1457] Fraud likelihood assessment

[1458] The server evaluates the possibility of fraud based on the analysis results. The input is the analysis results, and the output is the fraud possibility evaluation result. Specific operations include scoring the possibility of fraud based on the analysis results and determining whether the score exceeds a threshold.

[1459] Step 5:

[1460] Sending notifications

[1461] If it determines that there is a high possibility of fraud, the server notifies the user and relevant parties. The input is the fraud possibility assessment result, and the output is a notification message. Specific operations include sending warnings to the user and relevant parties via email, SMS, phone, etc.

[1462] Step 6:

[1463] Temporary locking of bank accounts

[1464] The server issues an instruction to temporarily lock the relevant bank account. The input is the fraud probability assessment result and user information, and the output is a lock instruction to be sent to the bank. Specific operations include sending a request to temporarily lock the account to the partner bank's API.

[1465] Step 7:

[1466] Start the identity verification process

[1467] Upon receiving the notification, the user begins the identity verification process. The device then provides a screen for entering a verification code or facial recognition via video call. The input is the notification message and user information, and the output is progress information on identity verification. Specific actions include the user entering a verification code or performing facial recognition via video call.

[1468] Step 8:

[1469] Unlocking a bank account

[1470] After successful identity verification, the server issues a command to unlock the bank account. The input is the successful identity verification information, and the output is an unlock command sent to the bank. Specific operations include sending an unlock request to the partner bank's API.

[1471] Step 9:

[1472] Analyzing abnormal voice patterns using AI models

[1473] The server analyzes abnormal voice patterns using a generative AI model. The input is the voice recognition and analysis results, and the output is the evaluation result of the abnormal voice pattern. Specific operations include inputting appropriate prompt sentences to the generative AI model and evaluating the abnormal voice pattern.

[1474] Step 10:

[1475] Generate prompt statement

[1476] The server generates prompt sentences to be input to the generative AI model. The inputs are the AI ​​model specifications and the information to be analyzed, and the output is the prompt sentence. Specific operations include creating prompt sentences appropriate to the situation to be analyzed (e.g., "Convert the voice data into text and detect any of the keywords 'refund,' 'bank,' or 'transfer' to determine the possibility of fraud. If there are signs of fraud, send a warning to the user via email.").

[1477] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1478] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing the user's emotions. This system mainly consists of a server, a terminal, and a user, and operates as follows.

[1479] Audio data collection and analysis

[1480] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence tools to detect abnormal patterns and fraud-related keywords.

[1481] Emotion recognition by emotion engine

[1482] The server then passes the call audio data to an emotion engine to recognize the user's emotions. This emotion engine assesses whether the user is experiencing stress or anxiety. The user's emotion data is then used to augment the assessment of the likelihood of fraud.

[1483] Fraud Assessment and Notification

[1484] If the analysis determines that there is a high possibility of fraud, the server will raise a fraud warning flag and immediately notify the user and relevant parties via phone, SMS, email, etc. to quickly warn them.

[1485] Temporary locking of bank accounts

[1486] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's bank account, preventing the user's funds from reaching the fraudsters.

[1487] Identity verification procedures

[1488] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device then provides a screen for entering a verification code or facial recognition via video call. After successfully verifying the user's identity, the server unlocks the bank account.

[1489] Specific examples

[1490] Scenario: Senior Citizen Refund Fraud

[1491] 1. Start a call

[1492] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[1493] 2. Collection and analysis of audio data

[1494] The device transmits the audio data of the fraudulent call to the server in real time.

[1495] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[1496] At the same time, the server uses an emotion engine to analyze Mr. A's emotional state and assess his level of stress and anxiety.

[1497] 3. Fraud Assessment and Notification

[1498] The server detects keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on the results of AI analysis and emotional data.

[1499] The server immediately notifies Mr. A and his family of the possible fraud.

[1500] 4. Temporarily locking your bank account

[1501] The server instructs the partner bank to temporarily lock Mr. A's bank account.

[1502] 5. Identity Verification Procedures

[1503] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[1504] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[1505] After verification, the server contacts the bank and unlocks Mr. A's bank account.

[1506] This invention makes it possible to detect fraudulent phone calls suffered by the elderly and young in real time, and by utilizing user emotion data, it is possible to improve the accuracy of fraud detection and prevent damage by quickly notifying relevant parties.In addition, the safety of users' funds can be ensured by locking bank accounts and implementing identity verification procedures.

[1507] The processing flow will be explained below.

[1508] Step 1:

[1509] The terminal detects that the user has started a telephone call, and simultaneously starts collecting the call voice data.

[1510] Step 2:

[1511] The terminal transmits the collected call voice data to the server in real time.

[1512] Step 3:

[1513] The server passes the received voice data to a voice recognition engine and converts the voice data into text data.

[1514] Step 4:

[1515] The server passes the converted text data to an AI analysis module to detect abnormal patterns and fraud-related keywords within the text data.

[1516] Step 5:

[1517] At the same time, the server passes the voice data to the emotion engine, which analyzes the user's emotions by analyzing the tone, tempo, and speaking style of the voice to assess whether the user is feeling stressed or anxious.

[1518] Step 6:

[1519] The server makes a comprehensive judgment based on the results of the AI ​​analysis and the evaluation results of the emotion engine, and if there is a high possibility of fraud, it raises a fraud warning flag.

[1520] Step 7:

[1521] If a fraud alert flag is raised, the server will immediately notify the user and relevant parties via phone, SMS, email, or other means.

[1522] Step 8:

[1523] The server instructs the user's designated bank to temporarily lock the user's bank account.

[1524] Step 9:

[1525] The user receives the notification and starts the identity verification procedure according to the specified procedure. The terminal provides an interface for identity verification.

[1526] Step 10:

[1527] The device sends the verification code entered by the user and the facial recognition results obtained via video call to the server.

[1528] Step 11:

[1529] The server receives the authentication information and verifies that the user is a legitimate user.

[1530] Step 12:

[1531] If the verification is successful, the server instructs the bank to unlock the user's bank account.

[1532] This process allows for real-time detection of potentially fraudulent calls, and prompt notification and response. The emotion engine also enables advanced fraud detection that takes into account the user's emotional state.

[1533] Example 2

[1534] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1535] Conventional communication monitoring systems simply convert voice data into text data and detect fraud through simple keyword matching, but this method is insufficient to counter increasingly sophisticated fraud techniques. Furthermore, because the system ignores the user's emotions and psychological state, the accuracy of fraud detection can be reduced. Furthermore, it is difficult to respond quickly after fraud is detected, making it difficult to effectively prevent users from losing funds.

[1536] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1537] In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, emotion recognition means for analyzing the call voice data and recognizing the user's emotion, means for notifying relevant parties when there is a high possibility of fraud, means for issuing instructions to temporarily lock the relevant bank account, and interface means for carrying out identity verification procedures. This enables highly sophisticated fraud to be detected with high accuracy, and by utilizing user emotion data, the accuracy of fraud detection is improved, enabling quick and effective response.

[1538] A "monitored communication line" is a communication infrastructure that is monitored and through which information is sent and received.

[1539] "Speech recognition technology" is a technology for converting voice data into text data, and typically uses an algorithm to extract linguistic information from voice.

[1540] "Voice data" refers to data that is a digital recording of the voice generated during a call.

[1541] "Text data" refers to the format of character information obtained after voice data is converted using voice recognition technology.

[1542] "Artificial intelligence means" means artificial intelligence-based techniques for analyzing the converted text data and assessing any abnormal patterns or potential fraud.

[1543] The "emotion recognition means" is a system that analyzes call voice data and evaluates the user's emotions and psychological state.

[1544] "Notification means" refers to methods or tools used to warn or inform relevant parties when a high possibility of fraud is determined.

[1545] A "bank account locking mechanism" is a system that executes instructions to temporarily freeze a user's bank account if fraud is suspected.

[1546] "Interface means" refers to input and output means for users and related parties to interact with the system, and includes, for example, a confirmation code input screen and a function for video calling.

[1547] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing user emotions. This system mainly consists of three components: a server, a terminal, and a user. Specific embodiments of the system are described below.

[1548] Audio data collection and analysis

[1549] The server constantly monitors the target communication line. Network monitoring software (e.g., Wireshark or PRTG) is used for monitoring. When a call is initiated, the device sends the call audio data to the server in real time. A streaming API (e.g., WebRTC) is used to collect the audio data, minimizing the delay in the audio data.

[1550] The server converts the received voice data into text data using the Google Cloud Speech-to-Text API. This conversion process is performed in parallel, and text is generated in real time.

[1551] The server then analyzes the converted text data using natural language processing (NLP) techniques such as GPT-3, which can identify unusual patterns and fraud-related keywords (e.g., "refund," "bank," "transfer," etc.).

[1552] Emotion recognition by emotion engine

[1553] The server then passes the call audio data to an emotion recognition engine (Emotion AI) to recognize the user's emotions. The emotion engine analyzes the tone and pace of the voice to assess whether the user is feeling stressed or anxious. For example, a trembling voice or a high-pitched tone can be detected as a sign of stress.

[1554] Fraud Assessment and Notification

[1555] The server combines the analysis results with sentiment data to assess the likelihood of fraud, and if it determines that fraud is likely, it raises a fraud warning flag.

[1556] If a potential fraud is detected, the server immediately notifies the user and relevant parties. The notification is sent in real time via SMS or email, allowing the user and relevant parties to take immediate action. For example, the notification may say, "A potential fraudulent call has been detected. Please take action immediately."

[1557] Temporary locking of bank accounts

[1558] After the fraud notification is complete, the server sends an instruction to the partner financial institution via API to temporarily lock the user's bank account, thereby preventing the outflow of funds.

[1559] Identity verification procedures

[1560] Upon receiving the notification, the user begins the identity verification process. The device provides facial recognition via a verification code input screen or video call. For example, the user receives a verification code via SMS and enters it on the device's input screen. Alternatively, facial recognition is performed via a video call to verify the user's identity. The server then contacts the financial institution and unlocks the bank account.

[1561] Specific examples

[1562] For example, consider the case of an elderly person named Mr. A who receives a call about a tax refund scam. The moment the call starts, the device sends a call start message to the server via the API, and captures and transmits voice data in real time.

[1563] The server converts the voice data into text using the Google Cloud Speech-to-Text API and analyzes the text using GPT-3. At the same time, it uses an emotion engine to analyze Mr. A's emotional state, detecting that he is experiencing increased stress and anxiety.

[1564] The server detects fraud-related keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on AI analysis and emotional data. It immediately notifies Mr. A and his family of the possibility of fraud and instructs the financial institution to temporarily lock Mr. A's bank account.

[1565] Upon receiving the notification, Mr. A or his family members can use the device interface to complete identity verification procedures, and once verification is complete, the bank account will be unlocked.

[1566] Prompt Sentence Examples

[1567] "They mentioned a refund, is this a scam?"

[1568] "How can I identify a fraudulent phone call?"

[1569] "Please let me know the analysis results of calls that are likely to be fraudulent."

[1570] The present invention makes it possible to detect fraudulent phone call damage in real time, improve the accuracy of fraud detection by utilizing user emotion data, and respond quickly.

[1571] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1572] Step 1:

[1573] The server constantly monitors the target communication line. The input is the connection information of the communication line, and the output is the call start trigger. Specifically, it uses network monitoring software (e.g., Wireshark or PRTG) to detect the start of a call in real time.

[1574] Step 2:

[1575] When a call is started, the device notifies the server of the call start information. The input is a call start event, and the output is a notification message to the server. Specifically, the device sends a call start message to the server through the API. This message contains the call identification information.

[1576] Step 3:

[1577] The device captures the call audio data in real time and sends it to the server. The input is the call audio data, and the output is a stream of audio data to the server. Specifically, it uses a streaming API (e.g., WebRTC) to capture the audio data and sends it to the server in an encrypted format.

[1578] Step 4:

[1579] The server converts the received voice data into text data using the Google Cloud Speech-to-Text API. The input is voice data and the output is text data. Specifically, the server sends voice data to the API and generates text data in real time. Parallel processing is used to improve conversion efficiency.

[1580] Step 5:

[1581] The server analyzes the converted text data using natural language processing technology such as GPT-3. The input is text data, and the output is the analysis results. Specifically, the text data is input into a generative AI model to detect fraud-related keywords such as "refund," "bank," and "transfer."

[1582] Step 6:

[1583] The server passes the call voice data to the Emotion AI engine, which analyzes the user's emotions. The input is voice data, and the output is emotional evaluation data. Specifically, it analyzes the voice tone and pace to evaluate whether the user is feeling stressed or anxious.

[1584] Step 7:

[1585] The server integrates the analysis results and emotion data to assess the likelihood of fraud. The input is the text analysis results and emotion evaluation data, and the output is a fraud warning flag. Specifically, it determines the likelihood of fraud based on abnormal patterns, fraud-related keywords, and user emotion data.

[1586] Step 8:

[1587] If it is determined that there is a high possibility of fraud, the server will notify the relevant parties. The input is a fraud warning flag and the output is a notification message. Specifically, a notification is sent to the relevant parties via SMS or email. The message sent will say something like, "A call with a high possibility of fraud has been detected. Please take immediate action."

[1588] Step 9:

[1589] After the fraud notification is complete, the server issues an instruction to the partner financial institution to temporarily lock the user's bank account. The input is the fraud warning flag and the user's bank information, and the output is a lock instruction to the bank. Specifically, the server sends the lock instruction to the bank via API to ensure the safety of the user's funds.

[1590] Step 10:

[1591] Upon receiving the notification, the user begins the identity verification procedure. The input is the notification message, and the output is the identity verification result. Specifically, the procedure is carried out using the verification code input screen or video call function provided by the device. After verification is complete, the server contacts the bank and unlocks the bank account.

[1592] (Application example 2)

[1593] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1594] In recent years, the number of victims of fraudulent phone calls targeting the elderly and young people has been increasing. These fraudulent acts have become more diverse, making it difficult to deal with them using conventional fraud detection methods alone. Furthermore, since the damage often occurs before the user even realizes it, new technologies are needed to respond to fraud early.

[1595] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for monitoring the communication lines to be monitored, means for converting voice data into text data using voice recognition technology, artificial intelligence means for analyzing the converted text data and evaluating the possibility of fraud, means for notifying relevant parties when there is a high possibility of fraud, means for issuing an instruction to temporarily lock the relevant trading account, and emotion recognition means for evaluating the user's emotional state. This enables highly accurate fraud detection and rapid response that takes the user's emotional state into account.

[1596] A "monitored communication line" is a line on a communication network that is continuously monitored to monitor activity on specific calls or data traffic.

[1597] "Voice recognition technology" is a technology that captures voice as a digital signal and converts it into text data or commands.

[1598] "Text data" is character information converted using speech recognition technology, and is the data that is the subject of analysis.

[1599] "Artificial intelligence means" are means that use technologies such as machine learning and deep learning to analyze data and detect patterns and anomalies.

[1600] "Means for notifying relevant parties" refers to means for conveying warnings and information to users and relevant third parties when possible fraud is detected.

[1601] "Means of issuing instructions to temporarily lock a trading account" refers to a means of instructing a bank or financial institution to freeze a trading account in order to prevent further damage when fraud is suspected.

[1602] "Emotion recognition means" refers to a technical means for assessing a user's emotional state from voice and other data, and determining the level of stress or anxiety.

[1603] The "interface means for carrying out identity verification procedures" refers to means including input devices and software for the user to verify their identity.

[1604] An "anomalous voice pattern" refers to a unique voice characteristic or word usage that differs from normal speech, and is a pattern that is an indicator of possible fraud.

[1605] The present invention is a system that monitors communication lines to be monitored, detects calls with a high probability of fraud, and further improves the accuracy of fraud detection by recognizing the user's emotions. This system mainly consists of a server, a terminal, and a user, and operates as follows.

[1606] Audio data collection and analysis

[1607] The server constantly monitors the target communication line. When a call is initiated, the device transmits the call audio data to the server in real time. The server converts the received audio data into text data using speech recognition technology. The server then analyzes the converted text data using artificial intelligence means to detect abnormal patterns and fraud-related keywords. This identification is performed using a generative AI model.

[1608] Emotion recognition by emotion engine

[1609] The server then passes the call audio data to an emotion recognition engine to recognize the user's emotions. This emotion engine assesses whether the user is experiencing stress or anxiety. The user's emotion data is then used to augment the assessment of the likelihood of fraud.

[1610] Fraud Assessment and Notification

[1611] If the analysis determines that there is a high possibility of fraud, the server will raise a fraud warning flag and immediately notify the user and relevant parties via phone, SMS, email, etc. to quickly warn them.

[1612] Temporary locking of trading accounts

[1613] After being notified of the fraud, the server instructs its partner banks to temporarily lock the user's trading account, preventing the user's funds from being leaked to the fraudsters.

[1614] Identity verification procedures

[1615] After receiving the fraud notification, the user begins the identity verification process using the provided method. The device will provide a verification code entry screen or face recognition via video call. After successful identity verification, the server will unlock the trading account.

[1616] Specific examples

[1617] Scenario: Senior Citizen Refund Fraud

[1618] 1. Start a call

[1619] Elderly person A receives a fraudulent phone call about receiving a tax refund.

[1620] 2. Collection and analysis of audio data

[1621] The device transmits the audio data of the fraudulent call to the server in real time.

[1622] The server converts the voice data into text data using voice recognition technology, and then analyzes the text data using artificial intelligence.

[1623] At the same time, the server uses an emotion engine to analyze Mr. A's emotional state and assess his level of stress and anxiety.

[1624] 3. Fraud Assessment and Notification

[1625] The server detects keywords such as "refund," "bank," and "transfer," and determines the likelihood of fraud based on the results of AI analysis and emotional data.

[1626] The server immediately notifies Mr. A and his family of the possible fraud.

[1627] 4. Temporary Locking of Trading Account

[1628] The server instructs the partner bank to temporarily lock Mr. A's trading account.

[1629] 5. Identity Verification Procedures

[1630] The user, Mr. A, or his family member will confirm the notification and begin the identity verification procedure.

[1631] The verification code is entered using the interface provided by the device, and the server receives and verifies the information.

[1632] After confirmation, the server contacts the bank and unlocks Mr. A's trading account.

[1633] This technology can detect fraudulent phone calls in real time, and by utilizing user emotional data, it is possible to improve the accuracy of fraud detection and prevent damage by quickly notifying relevant parties.In addition, the safety of users' funds can be ensured by locking their trading accounts and implementing identity verification procedures.

[1634] An example of a prompt to be input to the generative AI model is as follows:

[1635] "Develop a system that analyzes human emotions and detects fraudulent phone calls. Design an application that detects calls containing keywords such as 'refund,' 'bank,' and 'transfer,' as well as the user's feelings of stress and anxiety, and displays a warning if there is a high possibility of fraud."

[1636] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1637] Step 1:

[1638] The terminal starts collecting the call voice data when the call starts. The terminal transmits the collected voice data to the server in real time. The input is the call voice data, and the output is the transmitted voice data.

[1639] Step 2:

[1640] The server converts the received voice data into text data using voice recognition technology. Specifically, the server converts the voice data into text data using a voice recognition engine (e.g., Google Speech-to-Text API). The input is voice data, and the output is text data.

[1641] Step 3:

[1642] The server analyzes the converted text data using artificial intelligence means to detect fraud-related keywords and abnormal patterns. Specifically, the server uses natural language processing (NLP) techniques and deep learning models to extract fraud-related keywords (e.g., "refund," "bank," "transfer," etc.) from the text data. The input is the text data, and the output is the analysis results.

[1643] Step 4:

[1644] At the same time, the server passes the call voice data to an emotion recognition engine to evaluate the user's emotional state. Specifically, the server uses a voice emotion recognition model (e.g., IBM Watson Tone Analyzer) to evaluate emotional states such as stress and anxiety from the voice data. The input is voice data, and the output is emotional state data.

[1645] Step 5:

[1646] The server comprehensively evaluates the text data analysis results and the emotional state data to determine the likelihood of fraud. If the likelihood of fraud is high, the server raises a fraud warning flag. In this step, the text analysis results and the emotional state data are used as inputs, and a fraud warning flag is generated as output.

[1647] Step 6:

[1648] The server notifies users and other interested parties when a fraud alert flag is raised. Specifically, the server uses a notification method (e.g., phone call, SMS, email) to send a message informing them of a possible fraud. The input is the fraud alert flag, and the output is the notification message sent.

[1649] Step 7:

[1650] After the server is notified of the fraud, it instructs the affiliated financial institution to temporarily lock the user's trading account. The input is a fraud warning flag, and the output is an instruction to lock the trading account.

[1651] Step 8:

[1652] The user receives the fraud notification and starts the identity verification procedure using the device. Specifically, the device provides a verification code input screen and facial recognition functionality via video call. The input is the verification code and facial recognition data, and the output is the identity verification result.

[1653] Step 9:

[1654] After the identity verification procedure is successful, the server contacts the financial institution and unlocks the trading account. The input is the identity verification result, and the output is the instruction to unlock the trading account.

[1655] 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 control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1656] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1657] 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 the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1658] The emotion identification model 59 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 an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1659] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1660] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1661] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1662] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1663] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs 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 a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1664] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1665] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1666] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1667] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1668] 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.

[1669] It is not necessary to store all 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 all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1670] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1671] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with 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). Also, the hardware resource that executes the specific processing may be a single processor.

[1672] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1673] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1674] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1675] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1676] The following is further disclosed regarding the above embodiment.

[1677] (Claim 1)

[1678] a means for monitoring the monitored communication line;

[1679] A means for converting voice data into text data using voice recognition technology;

[1680] artificial intelligence means for analyzing the converted text data and assessing the likelihood of fraud;

[1681] A means of notifying relevant parties in cases where fraud is likely;

[1682] The system includes a means for issuing instructions to temporarily lock the applicable bank account.

[1683] (Claim 2)

[1684] 10. The system of claim 1, further comprising an interface means for performing an identity verification procedure.

[1685] (Claim 3)

[1686] 10. The system of claim 1, further comprising means for analyzing, in real time, abnormal audio patterns detected on the monitored communication line.

[1687] "Example 1"

[1688] (Claim 1)

[1689] a means for monitoring the monitored communication line;

[1690] A means for converting voice data into text data using voice recognition technology;

[1691] artificial intelligence means for analyzing the converted text data and assessing the likelihood of fraud;

[1692] A means of notifying relevant parties in cases where fraud is likely;

[1693] A means for issuing instructions to temporarily lock the account at the relevant financial institution;

[1694] A means for unlocking the device by providing identity verification information; and

[1695] means for transmitting voice data in real time as the call occurs;

[1696] A means for detecting fraudulent keywords based on the analysis results;

[1697] means for transmitting the notification using multiple communication means;

[1698] A system including:

[1699] (Claim 2)

[1700] 10. The system of claim 1, further comprising an interface means for performing an identity verification procedure.

[1701] (Claim 3)

[1702] 10. The system of claim 1, further comprising means for analyzing, in real time, abnormal audio patterns detected on the monitored communication line.

[1703] "Application Example 1"

[1704] (Claim 1)

[1705] a means for monitoring the monitored communication line;

[1706] A means for converting voice data into text data using voice recognition technology;

[1707] artificial intelligence means for analyzing the converted text data and assessing the likelihood of fraud;

[1708] A means of notifying relevant parties in cases where fraud is likely;

[1709] a means of issuing instructions to temporarily lock the relevant bank account;

[1710] A means for notifying a user of fraud warnings in real time based on the results of speech recognition and analysis;

[1711] A means of communication to monitor and notify potential fraudulent voice call data;

[1712] a means for evaluating anomalous voice patterns using a generative AI model; and

[1713] A system including means for generating prompt sentences to be input to a generative AI model.

[1714] (Claim 2)

[1715] 10. The system of claim 1, further comprising an interface means for performing an identity verification procedure.

[1716] (Claim 3)

[1717] 10. The system of claim 1, further comprising means for analyzing, in real time, abnormal audio patterns detected on the monitored communication line.

[1718] "Example 2: Combining Emotion Engines"

[1719] (Claim 1)

[1720] a means for monitoring the monitored communication line;

[1721] A means for converting voice data into text data using voice recognition technology;

[1722] artificial intelligence means for analyzing the converted text data and assessing the likelihood of fraud;

[1723] emotion recognition means for analyzing call voice data and recognizing the emotion of a user;

[1724] A means of notifying relevant parties in cases where fraud is likely;

[1725] The system includes a means for issuing instructions to temporarily lock the applicable bank account.

[1726] (Claim 2)

[1727] 10. The system of claim 1, further comprising an interface means for performing an identity verification procedure.

[1728] (Claim 3)

[1729] 10. The system of claim 1, further comprising means for analyzing, in real time, abnormal audio patterns detected on the monitored communication line.

[1730] "Application example 2 when combining emotion engines"

[1731] (Claim 1)

[1732] a means for monitoring the monitored communication line;

[1733] A means for converting voice data into text data using voice recognition technology;

[1734] artificial intelligence means for analyzing the converted text data and assessing the likelihood of fraud;

[1735] A means of notifying relevant parties in cases where fraud is likely;

[1736] A means for issuing instructions to temporarily lock the relevant trading account;

[1737] A system including an emotion recognition means for assessing the emotional state of a user.

[1738] (Claim 2)

[1739] 10. The system of claim 1, further comprising an interface means for performing an identity verification procedure.

[1740] (Claim 3)

[1741] 10. The system of claim 1, further comprising means for analyzing, in real time, abnormal audio patterns detected on the monitored communication line. [Explanation of symbols]

[1742] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a means for monitoring the monitored communication line; A means for converting voice data into text data using voice recognition technology; artificial intelligence means for analyzing the converted text data and assessing the likelihood of fraud; A means of notifying relevant parties in cases where fraud is likely; The system includes a means for issuing instructions to temporarily lock the applicable bank account.

2. 10. The system of claim 1, further comprising an interface means for performing an identity verification procedure.

3. 10. The system of claim 1, further comprising means for analyzing, in real time, abnormal audio patterns detected on the monitored communication line.

Images