System

Abstract

A system is provided.SOLUTION: A system comprising: means for obtaining voice data of a user; means for converting the obtained voice data to text data; means for analyzing the text data to detect a fraud pattern; means for generating an alert to notify the user if a fraud pattern is detected; means for sending the alert to a server; and means for the server to receive the alert and notify a registered third party.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] Currently, there are many cases of telephone fraud, such as refund scams, and the elderly are particularly vulnerable to these scams. Scammers use sophisticated methods to pose as local government officials or tax office employees to defraud the elderly of their money. Effective means for early detection and prevention of such fraudulent activities are needed, but existing countermeasures are insufficient, and many elderly people are still victims. The objective of this invention is to provide a system that analyzes audio data, detects potentially fraudulent conversations, and immediately issues an alert in order to prevent such fraudulent activities. [Means for solving the problem]

[0005] The present invention is a system that includes a means for acquiring user voice data, a means for converting the acquired voice data into text data, a means for analyzing the text data and detecting fraud patterns, a means for generating an alert and notifying the user when a fraud pattern is detected, a means for transmitting the alert to a server, and a means for the server to receive the alert and notify a registered third party. This system can prevent elderly people from falling victim to fraud by monitoring conversations that show signs of fraud in real time, detecting them early, and issuing an alert.

[0006] "User" refers to any individual who uses the System, including, in particular, those who are more vulnerable to fraud, such as seniors.

[0007] "Means for acquiring voice data" refers to a microphone or voice input device for recording the user's conversation.

[0008] "Means for converting into text data" refers to a voice recognition function that converts voice data into character data.

[0009] "Means for analyzing text data and detecting fraud patterns" refers to generative artificial intelligence (AI) for analyzing text data. This AI analyzes the data based on past fraud patterns.

[0010] "Means for generating alerts and notifying users" refers to the ability to issue a warning and notify users visually or audibly of potential fraud.

[0011] "Means for sending an alert to a server" refers to a communication function that sends the generated alert as data to a specified server.

[0012] "Server" means the computer system that receives and analyzes Alert Data, processes the Alert Data, and sends notifications to appropriate third parties.

[0013] "Registered third parties" refer to family members, police, and other related parties who are registered by the user in advance. These third parties are responsible for receiving alert notifications and responding promptly. [Brief explanation of the drawings]

[0014] [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

[0015] 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.

[0016] First, the terms used in the following description will be explained.

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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."

[0035] As an embodiment for carrying out the present invention, specific system operations and program processing will be described below.

[0036] System configuration

[0037] The system consists of a user-held device (a smartphone or dedicated device) and a central server. This device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence (such as ChatGPT), alert sending functionality, and communication functionality. The server has a database and handles alert notifications and data analysis.

[0038] Program processing

[0039] Acquiring voice data and converting it to text

[0040] Subject: Terminal

[0041] The device is always on and picks up the user's speech through a microphone. This voice data is sent to a speech recognition engine in real time and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve high-precision text conversion.

[0042] Text data analysis

[0043] Subject: Terminal

[0044] Generative AI (ChatGPT or equivalent technology) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[0045] Alert generation and user notification

[0046] Subject: Terminal

[0047] If a fraud pattern is detected, the device immediately generates an alert, which includes information about the potential fraud and specific actions the user should take, and is notified to the user through an audio message, a visual message, or both.

[0048] Alert transmission to the server and server response

[0049] Subject: Terminal, Server

[0050] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, push notification, or other methods.

[0051] Specific examples

[0052] Scenario 1: Refund fraud phone call

[0053] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0054] The device records the conversation and converts the audio data into text data.

[0055] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0056] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[0057] The terminal transmits the alert data to the server.

[0058] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0059] Scenario 2: Harmless phone call

[0060] A user is having a normal conversation with a friend.

[0061] The device records the conversation and converts the audio into text data.

[0062] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[0063] The device will not generate an alert and will notify the user as usual.

[0064] System convenience

[0065] The system provides users with a highly intuitive and easy-to-use interface. It specializes in detecting and preventing fraud targeting the elderly, allowing them to live their daily lives with peace of mind. It also notifies third parties, such as family members and the police, so they can respond quickly.

[0066] The above is an embodiment of the present invention. This system can provide an effective means for preventing fraud damage before it occurs.

[0067] The processing flow will be explained below.

[0068] Step 1:

[0069] Subject: Terminal

[0070] The device records the user's conversation in real time through a microphone and temporarily stores the recorded voice data in the device's memory.

[0071] Step 2:

[0072] Subject: Terminal

[0073] The device passes the recorded voice data to a speech recognition engine, which converts it into text data. The speech recognition engine performs noise filtering and voice clearing to convert the conversation into text with high accuracy.

[0074] Step 3:

[0075] Subject: Terminal

[0076] Generative AI (ChatGPT) analyzes text data and extracts key phrases and keywords, which are then compared with past fraud patterns.

[0077] Step 4:

[0078] Subject: Terminal

[0079] If a fraud pattern is detected, the device will immediately generate an alert that includes the likelihood of fraud and a recommended action, such as "This call may be fraudulent. Please hang up immediately."

[0080] Step 5:

[0081] Subject: Terminal

[0082] The device notifies the user of the generated alert either audibly or visually. In the case of an audio alert, the user is warned through the device's speaker, and in the case of a visual alert, a warning message is displayed on the screen.

[0083] Step 6:

[0084] Subject: Terminal

[0085] The terminal transmits alert data to the server via the network. The alert data includes the alert message as well as information about the user's current situation (e.g., location information).

[0086] Step 7:

[0087] Subject: Server

[0088] The server receives the alert data sent from the device, analyzes the received alert data, and determines which third parties (family members or the police) need to be notified.

[0089] Step 8:

[0090] Subject: Server

[0091] The server notifies pre-registered third parties of the alert data via email, SMS, push notification, etc. The notification content includes the alert message and current user information.

[0092] Step 9:

[0093] Subject: Family / Police

[0094] Family members and police receive notifications from the server and can respond quickly based on the content of the notification, for example, by contacting the user or heading to the scene.

[0095] The above is a description of the processing steps of the system and their specific operations. By explaining the specific processing performed at each step in detail, the method of implementing the present invention and the specific content of its operations will become clear.

[0096] Example 1

[0097] 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."

[0098] In recent years, telephone fraud has been on the rise, with elderly people being particularly targeted. Early detection and warnings are essential to prevent these types of fraud. However, current manual countermeasures tend to miss the timing, and it is difficult for elderly people to make the right decision. Therefore, there is a need to develop a system that automatically detects potential fraud using user voice data and issues an immediate warning.

[0099] 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.

[0100] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into character string data, means for analyzing the character string data and detecting fraudulent activity patterns, means for generating a warning and notifying the user when a fraudulent activity pattern is detected, means for transmitting the warning to the server, means for the server to receive the warning and notify a registered third party, means for converting voice data into character string data using a voice recognition engine, means for analyzing the character string data using a generative AI model and detecting fraudulent activity patterns, and means for notifying the user of the warning by voice message or visual message. This makes it possible to automatically detect fraud before a user becomes a victim of fraud and respond quickly.

[0101] "User" refers to an individual who uses the system.

[0102] "Voice data" refers to data that is a digital recording of a user's spoken words.

[0103] "Character string data" refers to data that has been converted from audio data into text format.

[0104] "Fraud patterns" refer to keywords or phrases associated with fraud or other illegal activity.

[0105] "Warning" refers to a message that notifies the user when the system detects possible fraud.

[0106] "Server" refers to a centralized device that manages databases and provides alert notifications.

[0107] "Third party" refers to an individual or organization other than the user (e.g., family members or the police).

[0108] A "voice recognition engine" refers to software that converts voice data into text data in real time.

[0109] A "generative AI model" refers to a model that uses artificial intelligence technology to perform specific tasks based on provided data.

[0110] "Voice message" refers to a warning that is given to the user in the form of a voice.

[0111] "Visual message" refers to a warning that is displayed on the user's terminal screen.

[0112] MODE FOR CARRYING OUT THE INVENTION

[0113] As an embodiment of the present invention, specific system operations and program processing will be described below.

[0114] System configuration

[0115] The system consists of a user-held device (e.g., a smartphone or dedicated device) and a central server. The device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence (e.g., ChatGPT), alert sending functionality, and communication functionality. The server has a database and handles alert notifications and information analysis.

[0116] Program processing

[0117] Acquiring voice data and converting it to text

[0118] Subject: Terminal

[0119] The device is always on and picks up the user's speech through a microphone. This speech data is sent in real time to a speech recognition engine (such as Google Cloud Speech-to-Text or Nuance's Dragon) and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve highly accurate text conversion.

[0120] Text data analysis

[0121] Subject: Terminal

[0122] Generative AI (such as OpenAI's ChatGPT) analyzes the converted text data in real time, extracting keywords and phrases related to fraud and matching them with an internal fraud pattern database that contains information on past fraud cases and indicators.

[0123] Alert generation and user notification

[0124] Subject: Terminal

[0125] If a fraud pattern is detected, the device immediately generates an alert that includes information about the potential fraud and specific actions the user should take. The alert may also include an audio and / or visual message to the user, such as "This call may be fraudulent. Please hang up immediately."

[0126] Alert transmission to the server and server response

[0127] Subject: Terminal and Server

[0128] The generated alert is sent from the device to a server. The server receives the alert and notifies pre-registered third parties, such as family members or the police. Notifications are sent via email, SMS, push notifications, etc. Furthermore, the notification also includes the user's current location information, allowing for a prompt response.

[0129] Specific examples

[0130] Scenario 1: Refund fraud phone call

[0131] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0132] The device records the conversation and converts the audio data into text data.

[0133] Generative artificial intelligence analyzes string data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0134] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[0135] The terminal transmits the alert data to the server.

[0136] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0137] Scenario 2: Harmless phone call

[0138] A user is having a normal conversation with a friend.

[0139] The device records the conversation and converts the audio data into text data.

[0140] Generative AI analyzes the string data but does not detect any keywords or patterns specifically related to fraud.

[0141] The device will not generate an alert and will notify the user as usual.

[0142] Prompt Sentence Examples

[0143] A user receives a call from someone claiming to be a city hall employee saying, "You have a tax refund, so please go to an ATM." Analyze this conversation and determine whether it is a scam.

[0144] The above is an embodiment of the present invention, and the system allows users to automatically detect fraudulent activity before they become victims and take prompt action.

[0145] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0146] Step 1: Acquire audio data

[0147] Subject: Terminal

[0148] Input: Ambient audio environment

[0149] How it works: The device's microphone constantly monitors the surrounding audio. When the user starts speaking, the device captures the audio data and starts recording.

[0150] Output: Captured audio data

[0151] Step 2: Convert audio data to text

[0152] Subject: Terminal

[0153] Input: Captured audio data

[0154] How it works: The device sends voice data in real time to a speech recognition engine (such as Google Cloud Speech-to-Text or Nuance's Dragon), which performs noise filtering and voice clearing and converts the voice data into text data.

[0155] Output: Converted string data

[0156] Step 3: Analyzing the text data

[0157] Subject: Terminal

[0158] Input: Converted string data

[0159] How it works: Generative AI (e.g., ChatGPT) analyzes text data to detect keywords and phrases associated with fraud, which are then compared against an internal fraud pattern database.

[0160] Output: Analysis results (data on possible fraudulent activity)

[0161] Step 4: Alert Generation

[0162] Subject: Terminal

[0163] Input: Analysis results

[0164] Specific Actions: When a fraudulent pattern is detected, the device immediately generates a warning message. The warning message includes information about the potential fraud and specific actions the user should take. For example, a message could read, "This call may be fraudulent. Please hang up immediately." The message can be in both audio and visual formats.

[0165] Output: Generated warning message

[0166] Step 5: User Notification

[0167] Subject: Terminal

[0168] Input: The generated warning message

[0169] Specific operation: The generated warning message is notified to the user via audio and visual means. The user's device will display a text message stating "Possible fraud" along with a voice message saying "This call may be fraudulent. Please hang up immediately."

[0170] Output: The warning message

[0171] Step 6: Sending alerts to the server

[0172] Subject: Terminal

[0173] Input: The generated warning message

[0174] Specific operation: The device sends the generated warning message to a server via the Internet. Communication uses the TCP / IP protocol, and data is encrypted for security.

[0175] Output: Alert data sent

[0176] Step 7: Server Alert Processing and Notification

[0177] Subject: Server

[0178] Input: Alert data sent

[0179] Specific operation: The server analyzes the received alert and notifies pre-registered family members and police. The notification includes the alert message and the user's current location. Notifications can be sent via email, SMS, or push notification.

[0180] Output: Alert message and location information sent to a third party

[0181] (Application example 1)

[0182] 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."

[0183] There is a need for a means to reduce the risk of elderly people and users who are not familiar with technology being caught up in fraudulent phone calls and fraudulent conversations, and to enable them to live their daily lives safely. In particular, it is essential to provide a system that analyzes the content of calls received by users in real time, promptly issues a warning in the event of a possible fraud, and notifies the appropriate third party.

[0184] 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.

[0185] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data using automatic generation technology to detect fraud patterns, means for generating an alert and notifying the user by voice or visual means when a fraud pattern is detected, means for sending the alert to the server, and means for the server to receive the alert and notify a registered third party by email or SMS. This allows users to deal with fraud risks in real time and prevent damage before it occurs through prompt warnings and notifications.

[0186] "User" refers to a person who uses this system.

[0187] "Voice data" refers to information that is a digital recording of a user's voice.

[0188] "Text data" refers to data that has been analyzed and converted into a character string of characters from audio data.

[0189] "Automatic generation technology" refers to technology that uses generative artificial intelligence to generate or analyze information.

[0190] "Fraud patterns" refer to keywords and phrases that indicate the possibility of fraud, based on information about past fraud cases and indicators of fraud.

[0191] "Alert" refers to a warning message provided to a user when potential fraud is detected.

[0192] "Server" refers to a central computer system that has a database and handles alert notifications and data analysis.

[0193] "Registered third parties" refer to family members, police, or other relevant parties that the user has pre-defined.

[0194] "Email" refers to a means of sending and receiving digital messages over the Internet.

[0195] "SMS" refers to a means of sending short text messages using a mobile phone.

[0196] As a specific embodiment of the present invention, the operation of the system and program processing will be described below.

[0197] System configuration

[0198] This system consists of a device used by the user (a smartphone or dedicated device) and a central server. The device is equipped with a voice input device, voice recognition function, generative artificial intelligence using automatic generation technology, alert sending function, and communication function. The server has a database and is responsible for sending alerts and analyzing data.

[0199] Program processing

[0200] Acquiring voice data and converting it to text

[0201] The device is always on and picks up the user's speech via a microphone. This voice data is sent in real time to a speech recognition function and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve highly accurate text conversion. The Google Cloud Speech-to-Text API is used for speech recognition.

[0202] Text data analysis

[0203] Generative AI (ChatGPT or equivalent technology) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[0204] Alert generation and user notification

[0205] If a fraud pattern is detected, the device will immediately generate an alert, which will include information about the potential fraud and specific actions the user should take. The alert will be communicated to the user via audio and / or visual messages, and will be delivered via the device's on-screen and audio output capabilities.

[0206] Alert transmission to the server and server response

[0207] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, or other means.

[0208] Specific examples

[0209] Scenario 1: Refund fraud phone call

[0210] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0211] The device records the conversation and converts the audio data into text data.

[0212] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0213] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[0214] The terminal transmits the alert data to the server.

[0215] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0216] Scenario 2: Harmless phone call

[0217] A user is having a normal conversation with a friend.

[0218] The device records the conversation and converts the audio into text data.

[0219] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[0220] The device will not generate an alert and will notify the user as usual.

[0221] Example prompt sentences to use

[0222] Analyze the likelihood that the following statements are fraudulent.

[0223] The refund is in your account. Please process it at an ATM.

[0224] If you suspect fraud, please explain why.

[0225] This system allows users to deal with fraud risks in real time and prevent damage before it occurs through prompt warnings and notifications. Notifications are also sent to third parties such as family members and the police, allowing for a prompt response.

[0226] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0227] Step 1:

[0228] The device acquires the user's voice data. Specifically, it records the user's conversation in real time through the device's microphone and saves the voice data in digital format. The input is the user's voice, and the output is digital voice data.

[0229] Step 2:

[0230] The device sends the acquired voice data to a voice recognition engine and converts it into text data. Specifically, noise filtering and voice clearing are performed on the voice data to achieve highly accurate text conversion. The Google Cloud Speech-to-Text API is used for voice recognition. The input is digital voice data, and the output is text data.

[0231] Step 3:

[0232] The device uses automated generation technology to send text data to a generative artificial intelligence (AI) for analysis. Specifically, the AI ​​(ChatGPT or equivalent technology) analyzes the text data and extracts fraud-related keywords and phrases. The input is the text data, and the output is a list of fraud-related keywords and phrases.

[0233] Step 4:

[0234] The device detects fraud patterns based on the analysis results from the generative AI. Specifically, it compares the results with a fraud pattern database to determine whether keywords or phrases indicate the possibility of fraud. The input is a list of keywords or phrases, and the output is a judgment result regarding the possibility of fraud.

[0235] Step 5:

[0236] If a fraud pattern is detected, the terminal immediately generates an alert. Specifically, it creates an alert message indicating the possibility of fraud and instructing the user on specific actions to take. The input is the result of the fraud possibility judgment, and the output is the alert message.

[0237] Step 6:

[0238] The terminal notifies the user of the generated alert message by voice and visual means. Specifically, the terminal displays the alert message on the screen and plays an audio alert using the audio output function. The input is the alert message, and the output is the notification to the user.

[0239] Step 7:

[0240] The terminal sends alert data to a central server. Specifically, it sends a data packet containing the alert content and the user's current location information to the server. The input is the alert message and location information, and the output is the data transmission to the server.

[0241] Step 8:

[0242] The server notifies pre-registered third parties based on the received alert data. Specifically, it sends the alert content and location information to family members or the police via email or SMS. The input is the alert data, and the output is a notification to the third party.

[0243] 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.

[0244] As an embodiment for carrying out the present invention, the operation and program processing of a system incorporating an emotion engine will be described below.

[0245] System configuration

[0246] The system consists of a user-held device (a smartphone or dedicated device), a central server, and an emotion engine. The device is equipped with a voice input device, voice recognition functionality, generative AI (such as ChatGPT), the emotion engine, an alert function, and communication functionality. The server has a database and handles alert notifications and data analysis.

[0247] Program processing

[0248] Acquiring voice data and converting it to text

[0249] Subject: Terminal

[0250] The device is always on and picks up the user's speech through a microphone. This voice data is sent to a speech recognition engine in real time and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve high-precision text conversion.

[0251] Text data analysis

[0252] Subject: Terminal

[0253] The generative AI (ChatGPT) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[0254] Acquiring and analyzing emotion data

[0255] Subject: Terminal

[0256] The emotion engine extracts emotional information from the user's voice and input data. The device measures the user's level of stress and anxiety, and combines this with fraud patterns to determine the level of risk. The emotion engine also analyzes the emotional data in real time and responds according to the user's emotional state.

[0257] Alert generation and user notification

[0258] Subject: Terminal

[0259] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the device immediately generates an alert. This alert includes information about the likelihood of fraud, specific recommended actions, and a counseling message based on the user's emotional state. For example, "This call may be fraudulent. Please hang up immediately. Also, you appear to be emotionally upset. Please remain calm."

[0260] Alert transmission to the server and server response

[0261] Subject: Terminal, Server

[0262] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, push notification, or other methods.

[0263] Specific examples

[0264] Scenario 1: Refund fraud phone call

[0265] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0266] The device records the conversation and converts the audio data into text data.

[0267] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0268] The emotion engine detects high stress levels from the user's tone of voice and language.

[0269] The device determines that the call is likely to be fraudulent and generates an alert that reads, "This call may be fraudulent. Please hang up immediately. Also, you appear to be upset, so please remain calm and take appropriate action," and notifies the user via audio and visual means.

[0270] The terminal transmits the alert data to the server.

[0271] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0272] Scenario 2: Regular phone call

[0273] A user is having a normal conversation with a friend.

[0274] The device records the conversation and converts the audio into text data.

[0275] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[0276] The emotion engine determines that the user's emotional state is normal.

[0277] The device will not generate an alert and will notify the user as usual.

[0278] System convenience and scalability

[0279] This system provides users with a highly intuitive and easy-to-use interface. It specializes in detecting and preventing fraud targeting the elderly, allowing them to live their daily lives with peace of mind. It also notifies third parties, such as family members and the police, enabling a prompt response. The introduction of an emotion engine provides a comprehensive crime prevention system that also responds to the user's emotional state.

[0280] The above is an embodiment of the present invention. This system can provide an effective means for preventing fraud damage before it occurs.

[0281] The processing flow will be explained below.

[0282] Step 1:

[0283] Subject: Terminal

[0284] The device records the user's conversation in real time through a microphone and temporarily stores the recorded voice data in the device's memory.

[0285] Step 2:

[0286] Subject: Terminal

[0287] The device passes the recorded voice data to a speech recognition engine for conversion to text data, which performs noise filtering and voice clearing to achieve highly accurate text conversion.

[0288] Step 3:

[0289] Subject: Terminal

[0290] The generative AI (ChatGPT) analyzes text data and extracts key phrases and keywords, which are then compared against an internal fraud pattern database.

[0291] Step 4:

[0292] Subject: Terminal

[0293] The emotion engine analyzes the voice data and extracts emotional information from the user's tone of voice and vocabulary, including levels of stress and anxiety.

[0294] Step 5:

[0295] Subject: Terminal

[0296] The emotion engine combines the extracted emotion information with the results of text analysis to determine the likelihood of fraud and the user's emotional state. If a fraud pattern is detected and the user is determined to be in a high-stress state, the device will immediately generate an alert.

[0297] Step 6:

[0298] Subject: Terminal

[0299] The device then notifies the user of the generated alerts, which may include information about potential fraud, specific recommended actions, and counseling messages based on the user's emotional state. Notifications may be audio or visual.

[0300] Step 7:

[0301] Subject: Terminal

[0302] The terminal transmits alert data to the server via the network. The alert data includes the alert message as well as the user's location information.

[0303] Step 8:

[0304] Subject: Server

[0305] The server receives the alert data sent from the device, analyzes the received alert data, and determines which third parties (family members or the police) need to be notified.

[0306] Step 9:

[0307] Subject: Server

[0308] The server notifies pre-registered third parties of the alert data via email, SMS, push notification, etc. The notification content includes the alert message and current user information.

[0309] Step 10:

[0310] Subject: Family / Police

[0311] Family members and police receive notifications from the server and can respond quickly based on the content of the notification, for example, by contacting the user or heading to the scene.

[0312] The above are the specific processing steps of the system that combines the emotion engine. By explaining the specific processing performed at each step in detail, the method of implementing the present invention and the specific content of its operation will become clear.

[0313] Example 2

[0314] 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."

[0315] In recent years, fraudulent phone scams have become more sophisticated, with an increasing number of victims targeting elderly people in particular. However, conventional fraud prevention systems only detect fraudulent patterns and are unable to respond in a way that takes into account the user's emotional state. As a result, even if users suspect a fraud, they are often unable to take appropriate action due to stress and anxiety. This results in the problem of not being able to prevent fraud victims from occurring.

[0316] 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.

[0317] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data using generative artificial intelligence to detect fraud patterns, means for acquiring and analyzing user emotional data, means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable, means for transmitting the alert to the server, and means for the server to receive the alert and notify a registered third party. This improves the accuracy of fraud detection and enables appropriate responses according to the user's emotional state.

[0318] "User" refers to an individual or end user who uses the System.

[0319] "Voice data" refers to data that is a digital recording of a user's voice.

[0320] "Text data" refers to data that has been converted from voice data into a string of characters using a voice recognition function.

[0321] "Generative artificial intelligence" refers to an artificial intelligence model that learns from large amounts of data and analyzes and generates natural language.

[0322] "Fraud patterns" refer to patterns that include specific keywords or phrases defined based on past fraud cases and indicators of fraud.

[0323] "Emotion data" refers to data related to the user's psychological state or emotions extracted from the user's voice or input data.

[0324] "Alert" means a notification generated when potential fraud is detected.

[0325] "Server" refers to a central management system that manages data and has notification functions.

[0326] "Registered third parties" refers to family members, friends, police, and other related parties who have been pre-registered as recipients of alerts from the server.

[0327] "Location information" refers to data that identifies a user's current geographic location.

[0328] As an embodiment for carrying out the present invention, the operation and program processing of a system incorporating an emotion engine will be described below.

[0329] System configuration

[0330] The system consists of a user-held device (a smartphone or dedicated device), a central server, and an emotion engine. The device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence, the emotion engine, an alert function, and communication functionality. The server has a database and can send alerts and perform data analysis.

[0331] Acquiring voice data and converting it to text

[0332] Subject: Terminal

[0333] The device is always on and picks up the user's speech via a microphone. The speech data is sent to a speech recognition engine such as the Google Speech-to-Text API, where it undergoes noise filtering and voice clearing before being converted into highly accurate text data. For example, a speech such as "Hello, this is City Hall" can be instantly converted into text.

[0334] Text data analysis

[0335] Subject: Terminal

[0336] The converted text data is analyzed by generative artificial intelligence. The device uses this text data to extract keywords and phrases related to fraud. For example, it checks against a database to see if it contains keywords such as "refund," "ATM," or "account information." This database contains information on past fraud cases and signs of fraud, enabling highly accurate fraud detection.

[0337] Acquiring and analyzing emotion data

[0338] Subject: Terminal

[0339] The emotion engine extracts emotional information from the user's voice and input data. The device uses the AWS Comprehend API, for example, to analyze emotions in real time from the user's tone of voice and vocabulary. For example, it can detect the tone of voice when the user is feeling doubtful or stressed and determine that the user is in a high stress state.

[0340] Alert generation and user notification

[0341] Subject: Terminal

[0342] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the device will immediately generate an alert. For example, if keywords such as "refund," "ATM," and "account information" are detected, the device will generate a message to the user via audio and visual notification, such as "This call may be fraudulent. Please hang up immediately. Also, you appear to be emotionally upset, so please remain calm."

[0343] Alert transmission to the server and server response

[0344] Subject: Terminal, Server

[0345] The generated alert is sent from the device to a server. The server receives this alert and sends a notification to a pre-registered third party (family member or police). For example, the server uses the Twilio API to send an emergency message (SMS or email). Push notifications can also be sent using Firebase, and notifications will be sent including the user's current location information.

[0346] Specific examples

[0347] Scenario 1: Refund fraud phone call

[0348] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0349] The device records the conversation and converts the audio data into text using the Google Speech-to-Text API.

[0350] Generative artificial intelligence analyzes this text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0351] The emotion engine detects high stress levels from the user's tone of voice and language.

[0352] The device will determine that the call is likely to be fraudulent and will send an alert to the user saying, "This call may be fraudulent. Please hang up immediately. Also, you seem upset, so please remain calm and take action."

[0353] The terminal transmits the alert data to the server.

[0354] The server receives the alert and uses the Twilio API to send notifications to family members and police, including the user's current location.

[0355] Scenario 2: Regular phone call

[0356] A user is having a normal conversation with a friend.

[0357] The device records the conversation and converts the audio into text using the Google Speech-to-Text API.

[0358] Generative AI analyzes text data but does not detect any keywords or patterns related to fraud.

[0359] The emotion engine determines that the user's emotional state is normal.

[0360] The device will not generate an alert and will notify the user as usual.

[0361] Examples of prompt statements

[0362] "Simulate the system's response when a user receives a fraudulent phone call."

[0363] "Please explain how the emotion engine and alert system work when making a regular phone call with a friend."

[0364] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0365] Step 1:

[0366] Subject: Terminal

[0367] The device constantly captures the user's voice data through the microphone. This voice data is input and sent in real time to a voice recognition engine such as the Google Speech-to-Text API. The voice recognition engine analyzes the voice while performing noise filtering and voice clearing, and converts it into a string of characters (text data). The output text data is a specific sentence such as "Hello, this is City Hall."

[0368] Step 2:

[0369] Subject: Terminal

[0370] The device inputs the converted text data into a generative AI (e.g., ChatGPT) and begins analysis. During the analysis process, fraud-related keywords and phrases (e.g., "refund," "ATM," and "account information") are extracted from the text data. The generative AI model compares the data with an internal fraud pattern database to determine the likelihood of fraud. The output is a judgment indicating whether a fraud pattern has been detected.

[0371] Step 3:

[0372] Subject: Terminal

[0373] The device inputs the user's voice and text data into an emotion engine (e.g., AWS Comprehend) and analyzes the emotion data. Specifically, it measures the user's stress and anxiety levels from their voice and vocabulary. The emotion engine performs tone analysis and determines if a high level of stress or unstable emotions is detected. The analysis results are output as emotion data.

[0374] Step 4:

[0375] Subject: Terminal

[0376] The device combines the fraud pattern judgment data detected in step 2 with the emotion data acquired in step 3 to determine the overall risk level. An alert is generated based on this judgment result. The alert contains specific messages such as, "This call may be fraudulent. Please hang up immediately. Also, you seem upset, so please remain calm and take action." The generated alert message is output.

[0377] Step 5:

[0378] Subject: Terminal

[0379] The device sends the generated alert to the server. The sent data also includes the user's location information. The alert message and location information are collected as input data and sent to the server via an appropriate communication protocol. The output data is a notification packet containing the alert message and location information.

[0380] Step 6:

[0381] Subject: Server

[0382] The server processes the alert notification received from the device. It sends an emergency message to pre-registered third parties (e.g., family members or police) using the Twilio API or Firebase push notification. The notification message contains the alert content and the user's current location information. The output data is the emergency notification message sent to the third party.

[0383] (Application example 2)

[0384] 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."

[0385] In modern society, fraud methods have become more sophisticated, making elderly people especially vulnerable to fraud. Fraud methods are so ingenious that simply detecting fraud patterns is not enough. To prevent fraud before it happens, a comprehensive crime prevention system that also takes into account the user's emotional state is necessary. However, current technology has not yet realized a system that can analyze the user's emotional state in real time and notify them of any signs of fraud.

[0386] 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 acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data and detecting fraud patterns, means for analyzing the user's emotional state, means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable, means for transmitting the alert to the server, and means for the server to receive the alert and notify a registered third party. This enables a comprehensive security system that takes into account the user's emotional state as well as the detection of fraud patterns.

[0387] "User voice data" is a digital signal containing the voice uttered by the user.

[0388] The "means for converting acquired voice data into text data" refers to a software and hardware system that utilizes voice recognition technology to convert voice data into a corresponding text format.

[0389] "Means for analyzing text data and detecting fraud patterns" refers to algorithms that analyze text data and identify signs and keywords of fraud based on past fraud cases, etc.

[0390] The "means for analyzing the user's emotional state" is an emotion analysis engine that analyzes the user's emotions such as stress, anxiety, and fear in real time from voice and text.

[0391] "Means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable" refers to a system that generates a warning message and notifies the user when it is determined that there is a high possibility of fraud and the user is in an unstable emotional state.

[0392] The "means for transmitting an alert to a server" is a communication system that securely transmits the generated alert message to a server over the Internet.

[0393] "Means for the server to receive alerts and notify registered third parties" refers to a function that analyzes alerts received on the server and notifies pre-set contacts (for example, family members or the police) via email, SMS, etc.

[0394] System configuration

[0395] The present invention is a system for acquiring and analyzing user voice data, which is composed of the following main components:

[0396] A means of obtaining user voice data

[0397] A means of converting voice data into text data using voice recognition technology

[0398] A method for analyzing text data and detecting fraud patterns using generative artificial intelligence

[0399] A sentiment analysis engine for analyzing users' emotional state in real time

[0400] A means of generating alerts and notifying users

[0401] A means of sending alerts to the server

[0402] A means for the server to receive alerts and notify registered third parties

[0403] Program processing

[0404] The system uses voice recognition technology to capture the user's voice data, then analyzes the text data using generative artificial intelligence models (such as ChatGPT and BERT), detects fraud patterns from the analyzed text data, and evaluates the user's emotional state using a sentiment analysis engine.

[0405] Hardware and Software Details

[0406] Hardware: Smartphone, microphone

[0407] Software: Python, speech_recognition library, transformers library, requests library

[0408] The smartphone's microphone is used to continuously record the user's voice, which is then converted into text in real time by a speech recognition engine. The converted text is then analyzed by generative artificial intelligence to detect fraud patterns, keywords, and phrases. At the same time, a sentiment analysis engine evaluates the user's emotional state from the text and voice to determine whether they are experiencing increased stress or anxiety.

[0409] Alerting and Notifications

[0410] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the system immediately generates an alert and notifies the user. The alert includes the possibility of fraud, specific recommended actions, and a counseling message based on the user's emotional state. The generated alert is sent to the server, which then notifies registered third parties.

[0411] Specific examples

[0412] Scenario 1: Refund fraud phone call

[0413] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0414] The device records the conversation and converts the audio data into text data.

[0415] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0416] The emotion engine detects high stress levels from the user's tone of voice and language.

[0417] The device determines that the call is likely to be fraudulent and generates an alert that reads, "This call may be fraudulent. Please hang up immediately. Also, you appear to be upset, so please remain calm and take appropriate action," and notifies the user via audio and visual means.

[0418] The terminal transmits the alert data to the server.

[0419] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0420] Prompt Sentence Examples

[0421] "This call is from City Hall to inform you about your refund. Please go to an ATM and follow the instructions provided."

[0422] This invention allows for a comprehensive security system that takes into account the emotional state of the user as well as detecting fraud patterns.

[0423] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0424] Step 1:

[0425] The terminal acquires the user's voice data. The user's voice is constantly recorded through a microphone and the voice data is acquired. The input is voice data, and the output is also voice data.

[0426] Step 2:

[0427] The device converts the acquired voice data into text data. It uses a voice recognition engine (e.g., Google Recognition API) to convert the voice data into text format. The input is voice data, and the output is text data.

[0428] Step 3:

[0429] The device analyzes text data using generative artificial intelligence (e.g., ChatGPT or BERT) to detect fraud-related keywords and phrases. The input is text data, and the output is the detection results of fraud patterns.

[0430] Step 4:

[0431] The device uses an emotion analysis engine to analyze the user's emotional state. It evaluates the user's emotions from voice and text data and determines the level of stress and anxiety. The input is voice and text data, and the output is the evaluation result of the emotional state.

[0432] Step 5:

[0433] If a fraud pattern is detected and the user's emotional state is unstable, the device generates an alert and notifies the user. Specifically, it generates a message such as "There is a possibility of fraud. Please hang up the phone immediately" and notifies the user visually or audibly. The input is the fraud pattern detection result and the emotional state evaluation result, and the output is the alert message.

[0434] Step 6:

[0435] The terminal sends the generated alert to the server. The alert message is sent to the server via the Internet and recorded in the database. The input is the alert message, and the output is the alert sending completion status.

[0436] Step 7:

[0437] The server analyzes the received alert and notifies pre-registered contacts (family, police, etc.) via email, SMS, push notification, etc. The input is the alert message, and the output is the notification message.

[0438] Specific working example:

[0439] For example, suppose the user speaks the following prompt:

[0440] "This call is from City Hall to inform you about your refund. Please go to an ATM and follow the instructions provided."

[0441] In this case, the system detected a fraud pattern and the user's emotional state indicated high stress, so it generated an alert and notified the user, "This may be a scam. Please hang up immediately." Similar alert messages were also sent to family members and the police.

[0442] 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.

[0443] 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.

[0444] 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.

[0445] [Second embodiment]

[0446] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0447] 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.

[0448] 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).

[0449] 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.

[0450] 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.

[0451] 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).

[0452] 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.

[0453] 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.

[0454] 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.

[0455] 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.

[0456] 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.

[0457] 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."

[0458] As an embodiment for carrying out the present invention, specific system operations and program processing will be described below.

[0459] System configuration

[0460] The system consists of a user-held device (a smartphone or dedicated device) and a central server. This device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence (such as ChatGPT), alert sending functionality, and communication functionality. The server has a database and handles alert notifications and data analysis.

[0461] Program processing

[0462] Acquiring voice data and converting it to text

[0463] Subject: Terminal

[0464] The device is always on and picks up the user's speech through a microphone. This voice data is sent to a speech recognition engine in real time and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve high-precision text conversion.

[0465] Text data analysis

[0466] Subject: Terminal

[0467] Generative AI (ChatGPT or equivalent technology) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[0468] Alert generation and user notification

[0469] Subject: Terminal

[0470] If a fraud pattern is detected, the device immediately generates an alert, which includes information about the potential fraud and specific actions the user should take, and is notified to the user through an audio message, a visual message, or both.

[0471] Alert transmission to the server and server response

[0472] Subject: Terminal, Server

[0473] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, push notification, or other methods.

[0474] Specific examples

[0475] Scenario 1: Refund fraud phone call

[0476] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0477] The device records the conversation and converts the audio data into text data.

[0478] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0479] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[0480] The terminal transmits the alert data to the server.

[0481] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0482] Scenario 2: Harmless phone call

[0483] A user is having a normal conversation with a friend.

[0484] The device records the conversation and converts the audio into text data.

[0485] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[0486] The device will not generate an alert and will notify the user as usual.

[0487] System convenience

[0488] The system provides users with a highly intuitive and easy-to-use interface. It specializes in detecting and preventing fraud targeting the elderly, allowing them to live their daily lives with peace of mind. It also notifies third parties, such as family members and the police, so they can respond quickly.

[0489] The above is an embodiment of the present invention. This system can provide an effective means for preventing fraud damage before it occurs.

[0490] The processing flow will be explained below.

[0491] Step 1:

[0492] Subject: Terminal

[0493] The device records the user's conversation in real time through a microphone and temporarily stores the recorded voice data in the device's memory.

[0494] Step 2:

[0495] Subject: Terminal

[0496] The device passes the recorded voice data to a speech recognition engine, which converts it into text data. The speech recognition engine performs noise filtering and voice clearing to convert the conversation into text with high accuracy.

[0497] Step 3:

[0498] Subject: Terminal

[0499] Generative AI (ChatGPT) analyzes text data and extracts key phrases and keywords, which are then compared with past fraud patterns.

[0500] Step 4:

[0501] Subject: Terminal

[0502] If a fraud pattern is detected, the device will immediately generate an alert that includes the likelihood of fraud and a recommended action, such as "This call may be fraudulent. Please hang up immediately."

[0503] Step 5:

[0504] Subject: Terminal

[0505] The device notifies the user of the generated alert either audibly or visually. In the case of an audio alert, the user is warned through the device's speaker, and in the case of a visual alert, a warning message is displayed on the screen.

[0506] Step 6:

[0507] Subject: Terminal

[0508] The terminal transmits alert data to the server via the network. The alert data includes the alert message as well as information about the user's current situation (e.g., location information).

[0509] Step 7:

[0510] Subject: Server

[0511] The server receives the alert data sent from the device, analyzes the received alert data, and determines which third parties (family members or the police) need to be notified.

[0512] Step 8:

[0513] Subject: Server

[0514] The server notifies pre-registered third parties of the alert data via email, SMS, push notification, etc. The notification content includes the alert message and current user information.

[0515] Step 9:

[0516] Subject: Family / Police

[0517] Family members and police receive notifications from the server and can respond quickly based on the content of the notification, for example, by contacting the user or heading to the scene.

[0518] The above is a description of the processing steps of the system and their specific operations. By explaining the specific processing performed at each step in detail, the method of implementing the present invention and the specific content of its operations will become clear.

[0519] Example 1

[0520] 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."

[0521] In recent years, telephone fraud has been on the rise, with elderly people being particularly targeted. Early detection and warnings are essential to prevent these types of fraud. However, current manual countermeasures tend to miss the timing, and it is difficult for elderly people to make the right decision. Therefore, there is a need to develop a system that automatically detects potential fraud using user voice data and issues an immediate warning.

[0522] 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.

[0523] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into character string data, means for analyzing the character string data and detecting fraudulent activity patterns, means for generating a warning and notifying the user when a fraudulent activity pattern is detected, means for transmitting the warning to the server, means for the server to receive the warning and notify a registered third party, means for converting voice data into character string data using a voice recognition engine, means for analyzing the character string data using a generative AI model and detecting fraudulent activity patterns, and means for notifying the user of the warning by voice message or visual message. This makes it possible to automatically detect fraud before a user becomes a victim of fraud and respond quickly.

[0524] "User" refers to an individual who uses the system.

[0525] "Voice data" refers to data that is a digital recording of a user's spoken words.

[0526] "Character string data" refers to data that has been converted from audio data into text format.

[0527] "Fraud patterns" refer to keywords or phrases associated with fraud or other illegal activity.

[0528] "Warning" refers to a message that notifies the user when the system detects possible fraud.

[0529] "Server" refers to a centralized device that manages databases and provides alert notifications.

[0530] "Third party" refers to an individual or organization other than the user (e.g., family members or the police).

[0531] A "voice recognition engine" refers to software that converts voice data into text data in real time.

[0532] A "generative AI model" refers to a model that uses artificial intelligence technology to perform specific tasks based on provided data.

[0533] "Voice message" refers to a warning that is given to the user in the form of a voice.

[0534] "Visual message" refers to a warning that is displayed on the user's terminal screen.

[0535] MODE FOR CARRYING OUT THE INVENTION

[0536] As an embodiment of the present invention, specific system operations and program processing will be described below.

[0537] System configuration

[0538] The system consists of a user-held device (e.g., a smartphone or dedicated device) and a central server. The device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence (e.g., ChatGPT), alert sending functionality, and communication functionality. The server has a database and handles alert notifications and information analysis.

[0539] Program processing

[0540] Acquiring voice data and converting it to text

[0541] Subject: Terminal

[0542] The device is always on and picks up the user's speech through a microphone. This speech data is sent in real time to a speech recognition engine (such as Google Cloud Speech-to-Text or Nuance's Dragon) and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve highly accurate text conversion.

[0543] Text data analysis

[0544] Subject: Terminal

[0545] Generative AI (such as OpenAI's ChatGPT) analyzes the converted text data in real time, extracting keywords and phrases related to fraud and matching them with an internal fraud pattern database that contains information on past fraud cases and indicators.

[0546] Alert generation and user notification

[0547] Subject: Terminal

[0548] If a fraud pattern is detected, the device immediately generates an alert that includes information about the potential fraud and specific actions the user should take. The alert may also include an audio and / or visual message to the user, such as "This call may be fraudulent. Please hang up immediately."

[0549] Alert transmission to the server and server response

[0550] Subject: Terminal and Server

[0551] The generated alert is sent from the device to a server. The server receives the alert and notifies pre-registered third parties, such as family members or the police. Notifications are sent via email, SMS, push notifications, etc. Furthermore, the notification also includes the user's current location information, allowing for a prompt response.

[0552] Specific examples

[0553] Scenario 1: Refund fraud phone call

[0554] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0555] The device records the conversation and converts the audio data into text data.

[0556] Generative artificial intelligence analyzes string data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0557] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[0558] The terminal transmits the alert data to the server.

[0559] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0560] Scenario 2: Harmless phone call

[0561] A user is having a normal conversation with a friend.

[0562] The device records the conversation and converts the audio data into text data.

[0563] Generative AI analyzes the string data but does not detect any keywords or patterns specifically related to fraud.

[0564] The device will not generate an alert and will notify the user as usual.

[0565] Prompt Sentence Examples

[0566] A user receives a call from someone claiming to be a city hall employee saying, "You have a tax refund, so please go to an ATM." Analyze this conversation and determine whether it is a scam.

[0567] The above is an embodiment of the present invention, and the system allows users to automatically detect fraudulent activity before they become victims and take prompt action.

[0568] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0569] Step 1: Acquire audio data

[0570] Subject: Terminal

[0571] Input: Ambient audio environment

[0572] How it works: The device's microphone constantly monitors the surrounding audio. When the user starts speaking, the device captures the audio data and starts recording.

[0573] Output: Captured audio data

[0574] Step 2: Convert audio data to text

[0575] Subject: Terminal

[0576] Input: Captured audio data

[0577] How it works: The device sends voice data in real time to a speech recognition engine (such as Google Cloud Speech-to-Text or Nuance's Dragon), which performs noise filtering and voice clearing and converts the voice data into text data.

[0578] Output: Converted string data

[0579] Step 3: Analyzing the text data

[0580] Subject: Terminal

[0581] Input: Converted string data

[0582] How it works: Generative AI (e.g., ChatGPT) analyzes text data to detect keywords and phrases associated with fraud, which are then compared against an internal fraud pattern database.

[0583] Output: Analysis results (data on possible fraudulent activity)

[0584] Step 4: Alert Generation

[0585] Subject: Terminal

[0586] Input: Analysis results

[0587] Specific Actions: When a fraudulent pattern is detected, the device immediately generates a warning message. The warning message includes information about the potential fraud and specific actions the user should take. For example, a message could read, "This call may be fraudulent. Please hang up immediately." The message can be in both audio and visual formats.

[0588] Output: Generated warning message

[0589] Step 5: User Notification

[0590] Subject: Terminal

[0591] Input: The generated warning message

[0592] Specific operation: The generated warning message is notified to the user via audio and visual means. The user's device will display a text message stating "Possible fraud" along with a voice message saying "This call may be fraudulent. Please hang up immediately."

[0593] Output: The warning message

[0594] Step 6: Sending alerts to the server

[0595] Subject: Terminal

[0596] Input: The generated warning message

[0597] Specific operation: The device sends the generated warning message to a server via the Internet. Communication uses the TCP / IP protocol, and data is encrypted for security.

[0598] Output: Alert data sent

[0599] Step 7: Server Alert Processing and Notification

[0600] Subject: Server

[0601] Input: Alert data sent

[0602] Specific operation: The server analyzes the received alert and notifies pre-registered family members and police. The notification includes the alert message and the user's current location. Notifications can be sent via email, SMS, or push notification.

[0603] Output: Alert message and location information sent to a third party

[0604] (Application example 1)

[0605] 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."

[0606] There is a need for a means to reduce the risk of elderly people and users who are not familiar with technology being caught up in fraudulent phone calls and fraudulent conversations, and to enable them to live their daily lives safely. In particular, it is essential to provide a system that analyzes the content of calls received by users in real time, promptly issues a warning in the event of a possible fraud, and notifies the appropriate third party.

[0607] 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.

[0608] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data using automatic generation technology to detect fraud patterns, means for generating an alert and notifying the user by voice or visual means when a fraud pattern is detected, means for sending the alert to the server, and means for the server to receive the alert and notify a registered third party by email or SMS. This allows users to deal with fraud risks in real time and prevent damage before it occurs through prompt warnings and notifications.

[0609] "User" refers to a person who uses this system.

[0610] "Voice data" refers to information that is a digital recording of a user's voice.

[0611] "Text data" refers to data that has been analyzed and converted into a character string of characters from audio data.

[0612] "Automatic generation technology" refers to technology that uses generative artificial intelligence to generate or analyze information.

[0613] "Fraud patterns" refer to keywords and phrases that indicate the possibility of fraud, based on information about past fraud cases and indicators of fraud.

[0614] "Alert" refers to a warning message provided to a user when potential fraud is detected.

[0615] "Server" refers to a central computer system that has a database and handles alert notifications and data analysis.

[0616] "Registered third parties" refer to family members, police, or other relevant parties that the user has pre-defined.

[0617] "Email" refers to a means of sending and receiving digital messages over the Internet.

[0618] "SMS" refers to a means of sending short text messages using a mobile phone.

[0619] As a specific embodiment of the present invention, the operation of the system and program processing will be described below.

[0620] System configuration

[0621] This system consists of a device used by the user (a smartphone or dedicated device) and a central server. The device is equipped with a voice input device, voice recognition function, generative artificial intelligence using automatic generation technology, alert sending function, and communication function. The server has a database and is responsible for sending alerts and analyzing data.

[0622] Program processing

[0623] Acquiring voice data and converting it to text

[0624] The device is always on and picks up the user's speech via a microphone. This voice data is sent in real time to a speech recognition function and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve highly accurate text conversion. The Google Cloud Speech-to-Text API is used for speech recognition.

[0625] Text data analysis

[0626] Generative AI (ChatGPT or equivalent technology) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[0627] Alert generation and user notification

[0628] If a fraud pattern is detected, the device will immediately generate an alert, which will include information about the potential fraud and specific actions the user should take. The alert will be communicated to the user via audio and / or visual messages, and will be delivered via the device's on-screen and audio output capabilities.

[0629] Alert transmission to the server and server response

[0630] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, or other means.

[0631] Specific examples

[0632] Scenario 1: Refund fraud phone call

[0633] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0634] The device records the conversation and converts the audio data into text data.

[0635] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0636] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[0637] The terminal transmits the alert data to the server.

[0638] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0639] Scenario 2: Harmless phone call

[0640] A user is having a normal conversation with a friend.

[0641] The device records the conversation and converts the audio into text data.

[0642] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[0643] The device will not generate an alert and will notify the user as usual.

[0644] Example prompt sentences to use

[0645] Analyze the likelihood that the following statements are fraudulent.

[0646] The refund is in your account. Please process it at an ATM.

[0647] If you suspect fraud, please explain why.

[0648] This system allows users to deal with fraud risks in real time and prevent damage before it occurs through prompt warnings and notifications. Notifications are also sent to third parties such as family members and the police, allowing for a prompt response.

[0649] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0650] Step 1:

[0651] The device acquires the user's voice data. Specifically, it records the user's conversation in real time through the device's microphone and saves the voice data in digital format. The input is the user's voice, and the output is digital voice data.

[0652] Step 2:

[0653] The device sends the acquired voice data to a voice recognition engine and converts it into text data. Specifically, noise filtering and voice clearing are performed on the voice data to achieve highly accurate text conversion. The Google Cloud Speech-to-Text API is used for voice recognition. The input is digital voice data, and the output is text data.

[0654] Step 3:

[0655] The device uses automated generation technology to send text data to a generative artificial intelligence (AI) for analysis. Specifically, the AI ​​(ChatGPT or equivalent technology) analyzes the text data and extracts fraud-related keywords and phrases. The input is the text data, and the output is a list of fraud-related keywords and phrases.

[0656] Step 4:

[0657] The device detects fraud patterns based on the analysis results from the generative AI. Specifically, it compares the results with a fraud pattern database to determine whether keywords or phrases indicate the possibility of fraud. The input is a list of keywords or phrases, and the output is a judgment result regarding the possibility of fraud.

[0658] Step 5:

[0659] If a fraud pattern is detected, the terminal immediately generates an alert. Specifically, it creates an alert message indicating the possibility of fraud and instructing the user on specific actions to take. The input is the result of the fraud possibility judgment, and the output is the alert message.

[0660] Step 6:

[0661] The terminal notifies the user of the generated alert message by voice and visual means. Specifically, the terminal displays the alert message on the screen and plays an audio alert using the audio output function. The input is the alert message, and the output is the notification to the user.

[0662] Step 7:

[0663] The terminal sends alert data to a central server. Specifically, it sends a data packet containing the alert content and the user's current location information to the server. The input is the alert message and location information, and the output is the data transmission to the server.

[0664] Step 8:

[0665] The server notifies pre-registered third parties based on the received alert data. Specifically, it sends the alert content and location information to family members or the police via email or SMS. The input is the alert data, and the output is a notification to the third party.

[0666] 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.

[0667] As an embodiment for carrying out the present invention, the operation and program processing of a system incorporating an emotion engine will be described below.

[0668] System configuration

[0669] The system consists of a user-held device (a smartphone or dedicated device), a central server, and an emotion engine. The device is equipped with a voice input device, voice recognition functionality, generative AI (such as ChatGPT), the emotion engine, an alert function, and communication functionality. The server has a database and handles alert notifications and data analysis.

[0670] Program processing

[0671] Acquiring voice data and converting it to text

[0672] Subject: Terminal

[0673] The device is always on and picks up the user's speech through a microphone. This voice data is sent to a speech recognition engine in real time and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve high-precision text conversion.

[0674] Text data analysis

[0675] Subject: Terminal

[0676] The generative AI (ChatGPT) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[0677] Acquiring and analyzing emotion data

[0678] Subject: Terminal

[0679] The emotion engine extracts emotional information from the user's voice and input data. The device measures the user's level of stress and anxiety, and combines this with fraud patterns to determine the level of risk. The emotion engine also analyzes the emotional data in real time and responds according to the user's emotional state.

[0680] Alert generation and user notification

[0681] Subject: Terminal

[0682] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the device immediately generates an alert. This alert includes information about the likelihood of fraud, specific recommended actions, and a counseling message based on the user's emotional state. For example, "This call may be fraudulent. Please hang up immediately. Also, you appear to be emotionally upset. Please remain calm."

[0683] Alert transmission to the server and server response

[0684] Subject: Terminal, Server

[0685] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, push notification, or other methods.

[0686] Specific examples

[0687] Scenario 1: Refund fraud phone call

[0688] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0689] The device records the conversation and converts the audio data into text data.

[0690] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0691] The emotion engine detects high stress levels from the user's tone of voice and language.

[0692] The device determines that the call is likely to be fraudulent and generates an alert that reads, "This call may be fraudulent. Please hang up immediately. Also, you appear to be upset, so please remain calm and take appropriate action," and notifies the user via audio and visual means.

[0693] The terminal transmits the alert data to the server.

[0694] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0695] Scenario 2: Regular phone call

[0696] A user is having a normal conversation with a friend.

[0697] The device records the conversation and converts the audio into text data.

[0698] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[0699] The emotion engine determines that the user's emotional state is normal.

[0700] The device will not generate an alert and will notify the user as usual.

[0701] System convenience and scalability

[0702] This system provides users with a highly intuitive and easy-to-use interface. It specializes in detecting and preventing fraud targeting the elderly, allowing them to live their daily lives with peace of mind. It also notifies third parties, such as family members and the police, enabling a prompt response. The introduction of an emotion engine provides a comprehensive crime prevention system that also responds to the user's emotional state.

[0703] The above is an embodiment of the present invention. This system can provide an effective means for preventing fraud damage before it occurs.

[0704] The processing flow will be explained below.

[0705] Step 1:

[0706] Subject: Terminal

[0707] The device records the user's conversation in real time through a microphone and temporarily stores the recorded voice data in the device's memory.

[0708] Step 2:

[0709] Subject: Terminal

[0710] The device passes the recorded voice data to a speech recognition engine for conversion to text data, which performs noise filtering and voice clearing to achieve highly accurate text conversion.

[0711] Step 3:

[0712] Subject: Terminal

[0713] The generative AI (ChatGPT) analyzes text data and extracts key phrases and keywords, which are then compared against an internal fraud pattern database.

[0714] Step 4:

[0715] Subject: Terminal

[0716] The emotion engine analyzes the voice data and extracts emotional information from the user's tone of voice and vocabulary, including levels of stress and anxiety.

[0717] Step 5:

[0718] Subject: Terminal

[0719] The emotion engine combines the extracted emotion information with the results of text analysis to determine the likelihood of fraud and the user's emotional state. If a fraud pattern is detected and the user is determined to be in a high-stress state, the device will immediately generate an alert.

[0720] Step 6:

[0721] Subject: Terminal

[0722] The device then notifies the user of the generated alerts, which may include information about potential fraud, specific recommended actions, and counseling messages based on the user's emotional state. Notifications may be audio or visual.

[0723] Step 7:

[0724] Subject: Terminal

[0725] The terminal transmits alert data to the server via the network. The alert data includes the alert message as well as the user's location information.

[0726] Step 8:

[0727] Subject: Server

[0728] The server receives the alert data sent from the device, analyzes the received alert data, and determines which third parties (family members or the police) need to be notified.

[0729] Step 9:

[0730] Subject: Server

[0731] The server notifies pre-registered third parties of the alert data via email, SMS, push notification, etc. The notification content includes the alert message and current user information.

[0732] Step 10:

[0733] Subject: Family / Police

[0734] Family members and police receive notifications from the server and can respond quickly based on the content of the notification, for example, by contacting the user or heading to the scene.

[0735] The above are the specific processing steps of the system that combines the emotion engine. By explaining the specific processing performed at each step in detail, the method of implementing the present invention and the specific content of its operation will become clear.

[0736] Example 2

[0737] 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."

[0738] In recent years, fraudulent phone scams have become more sophisticated, with an increasing number of victims targeting elderly people in particular. However, conventional fraud prevention systems only detect fraudulent patterns and are unable to respond in a way that takes into account the user's emotional state. As a result, even if users suspect a fraud, they are often unable to take appropriate action due to stress and anxiety. This results in the problem of not being able to prevent fraud victims from occurring.

[0739] 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.

[0740] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data using generative artificial intelligence to detect fraud patterns, means for acquiring and analyzing user emotional data, means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable, means for transmitting the alert to the server, and means for the server to receive the alert and notify a registered third party. This improves the accuracy of fraud detection and enables appropriate responses according to the user's emotional state.

[0741] "User" refers to an individual or end user who uses the System.

[0742] "Voice data" refers to data that is a digital recording of a user's voice.

[0743] "Text data" refers to data that has been converted from voice data into a string of characters using a voice recognition function.

[0744] "Generative artificial intelligence" refers to an artificial intelligence model that learns from large amounts of data and analyzes and generates natural language.

[0745] "Fraud patterns" refer to patterns that include specific keywords or phrases defined based on past fraud cases and indicators of fraud.

[0746] "Emotion data" refers to data related to the user's psychological state or emotions extracted from the user's voice or input data.

[0747] "Alert" means a notification generated when potential fraud is detected.

[0748] "Server" refers to a central management system that manages data and has notification functions.

[0749] "Registered third parties" refers to family members, friends, police, and other related parties who have been pre-registered as recipients of alerts from the server.

[0750] "Location information" refers to data that identifies a user's current geographic location.

[0751] As an embodiment for carrying out the present invention, the operation and program processing of a system incorporating an emotion engine will be described below.

[0752] System configuration

[0753] The system consists of a user-held device (a smartphone or dedicated device), a central server, and an emotion engine. The device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence, the emotion engine, an alert function, and communication functionality. The server has a database and can send alerts and perform data analysis.

[0754] Acquiring voice data and converting it to text

[0755] Subject: Terminal

[0756] The device is always on and picks up the user's speech via a microphone. The speech data is sent to a speech recognition engine such as the Google Speech-to-Text API, where it undergoes noise filtering and voice clearing before being converted into highly accurate text data. For example, a speech such as "Hello, this is City Hall" can be instantly converted into text.

[0757] Text data analysis

[0758] Subject: Terminal

[0759] The converted text data is analyzed by generative artificial intelligence. The device uses this text data to extract keywords and phrases related to fraud. For example, it checks against a database to see if it contains keywords such as "refund," "ATM," or "account information." This database contains information on past fraud cases and signs of fraud, enabling highly accurate fraud detection.

[0760] Acquiring and analyzing emotion data

[0761] Subject: Terminal

[0762] The emotion engine extracts emotional information from the user's voice and input data. The device uses the AWS Comprehend API, for example, to analyze emotions in real time from the user's tone of voice and vocabulary. For example, it can detect the tone of voice when the user is feeling doubtful or stressed and determine that the user is in a high stress state.

[0763] Alert generation and user notification

[0764] Subject: Terminal

[0765] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the device will immediately generate an alert. For example, if keywords such as "refund," "ATM," and "account information" are detected, the device will generate a message to the user via audio and visual notification, such as "This call may be fraudulent. Please hang up immediately. Also, you appear to be emotionally upset, so please remain calm."

[0766] Alert transmission to the server and server response

[0767] Subject: Terminal, Server

[0768] The generated alert is sent from the device to a server. The server receives this alert and sends a notification to a pre-registered third party (family member or police). For example, the server uses the Twilio API to send an emergency message (SMS or email). Push notifications can also be sent using Firebase, and notifications will be sent including the user's current location information.

[0769] Specific examples

[0770] Scenario 1: Refund fraud phone call

[0771] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0772] The device records the conversation and converts the audio data into text using the Google Speech-to-Text API.

[0773] Generative artificial intelligence analyzes this text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0774] The emotion engine detects high stress levels from the user's tone of voice and language.

[0775] The device will determine that the call is likely to be fraudulent and will send an alert to the user saying, "This call may be fraudulent. Please hang up immediately. Also, you seem upset, so please remain calm and take action."

[0776] The terminal transmits the alert data to the server.

[0777] The server receives the alert and uses the Twilio API to send notifications to family members and police, including the user's current location.

[0778] Scenario 2: Regular phone call

[0779] A user is having a normal conversation with a friend.

[0780] The device records the conversation and converts the audio into text using the Google Speech-to-Text API.

[0781] Generative AI analyzes text data but does not detect any keywords or patterns related to fraud.

[0782] The emotion engine determines that the user's emotional state is normal.

[0783] The device will not generate an alert and will notify the user as usual.

[0784] Examples of prompt statements

[0785] "Simulate the system's response when a user receives a fraudulent phone call."

[0786] "Please explain how the emotion engine and alert system work when making a regular phone call with a friend."

[0787] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0788] Step 1:

[0789] Subject: Terminal

[0790] The device constantly captures the user's voice data through the microphone. This voice data is input and sent in real time to a voice recognition engine such as the Google Speech-to-Text API. The voice recognition engine analyzes the voice while performing noise filtering and voice clearing, and converts it into a string of characters (text data). The output text data is a specific sentence such as "Hello, this is City Hall."

[0791] Step 2:

[0792] Subject: Terminal

[0793] The device inputs the converted text data into a generative AI (e.g., ChatGPT) and begins analysis. During the analysis process, fraud-related keywords and phrases (e.g., "refund," "ATM," and "account information") are extracted from the text data. The generative AI model compares the data with an internal fraud pattern database to determine the likelihood of fraud. The output is a judgment indicating whether a fraud pattern has been detected.

[0794] Step 3:

[0795] Subject: Terminal

[0796] The device inputs the user's voice and text data into an emotion engine (e.g., AWS Comprehend) and analyzes the emotion data. Specifically, it measures the user's stress and anxiety levels from their voice and vocabulary. The emotion engine performs tone analysis and determines if a high level of stress or unstable emotions is detected. The analysis results are output as emotion data.

[0797] Step 4:

[0798] Subject: Terminal

[0799] The device combines the fraud pattern judgment data detected in step 2 with the emotion data acquired in step 3 to determine the overall risk level. An alert is generated based on this judgment result. The alert contains specific messages such as, "This call may be fraudulent. Please hang up immediately. Also, you seem upset, so please remain calm and take action." The generated alert message is output.

[0800] Step 5:

[0801] Subject: Terminal

[0802] The device sends the generated alert to the server. The sent data also includes the user's location information. The alert message and location information are collected as input data and sent to the server via an appropriate communication protocol. The output data is a notification packet containing the alert message and location information.

[0803] Step 6:

[0804] Subject: Server

[0805] The server processes the alert notification received from the device. It sends an emergency message to pre-registered third parties (e.g., family members or police) using the Twilio API or Firebase push notification. The notification message contains the alert content and the user's current location information. The output data is the emergency notification message sent to the third party.

[0806] (Application example 2)

[0807] 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."

[0808] In modern society, fraud methods have become more sophisticated, making elderly people especially vulnerable to fraud. Fraud methods are so ingenious that simply detecting fraud patterns is not enough. To prevent fraud before it happens, a comprehensive crime prevention system that also takes into account the user's emotional state is necessary. However, current technology has not yet realized a system that can analyze the user's emotional state in real time and notify them of any signs of fraud.

[0809] 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 acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data and detecting fraud patterns, means for analyzing the user's emotional state, means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable, means for transmitting the alert to the server, and means for the server to receive the alert and notify a registered third party. This enables a comprehensive security system that takes into account the user's emotional state as well as the detection of fraud patterns.

[0810] "User voice data" is a digital signal containing the voice uttered by the user.

[0811] The "means for converting acquired voice data into text data" refers to a software and hardware system that utilizes voice recognition technology to convert voice data into a corresponding text format.

[0812] "Means for analyzing text data and detecting fraud patterns" refers to algorithms that analyze text data and identify signs and keywords of fraud based on past fraud cases, etc.

[0813] The "means for analyzing the user's emotional state" is an emotion analysis engine that analyzes the user's emotions such as stress, anxiety, and fear in real time from voice and text.

[0814] "Means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable" refers to a system that generates a warning message and notifies the user when it is determined that there is a high possibility of fraud and the user is in an unstable emotional state.

[0815] The "means for transmitting an alert to a server" is a communication system that securely transmits the generated alert message to a server over the Internet.

[0816] "Means for the server to receive alerts and notify registered third parties" refers to a function that analyzes alerts received on the server and notifies pre-set contacts (for example, family members or the police) via email, SMS, etc.

[0817] System configuration

[0818] The present invention is a system for acquiring and analyzing user voice data, which is composed of the following main components:

[0819] A means of obtaining user voice data

[0820] A means of converting voice data into text data using voice recognition technology

[0821] A method for analyzing text data and detecting fraud patterns using generative artificial intelligence

[0822] A sentiment analysis engine for analyzing users' emotional state in real time

[0823] A means of generating alerts and notifying users

[0824] A means of sending alerts to the server

[0825] A means for the server to receive alerts and notify registered third parties

[0826] Program processing

[0827] The system uses voice recognition technology to capture the user's voice data, then analyzes the text data using generative artificial intelligence models (such as ChatGPT and BERT), detects fraud patterns from the analyzed text data, and evaluates the user's emotional state using a sentiment analysis engine.

[0828] Hardware and Software Details

[0829] Hardware: Smartphone, microphone

[0830] Software: Python, speech_recognition library, transformers library, requests library

[0831] The smartphone's microphone is used to continuously record the user's voice, which is then converted into text in real time by a speech recognition engine. The converted text is then analyzed by generative artificial intelligence to detect fraud patterns, keywords, and phrases. At the same time, a sentiment analysis engine evaluates the user's emotional state from the text and voice to determine whether they are experiencing increased stress or anxiety.

[0832] Alerting and Notifications

[0833] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the system immediately generates an alert and notifies the user. The alert includes the possibility of fraud, specific recommended actions, and a counseling message based on the user's emotional state. The generated alert is sent to the server, which then notifies registered third parties.

[0834] Specific examples

[0835] Scenario 1: Refund fraud phone call

[0836] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0837] The device records the conversation and converts the audio data into text data.

[0838] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0839] The emotion engine detects high stress levels from the user's tone of voice and language.

[0840] The device determines that the call is likely to be fraudulent and generates an alert that reads, "This call may be fraudulent. Please hang up immediately. Also, you appear to be upset, so please remain calm and take appropriate action," and notifies the user via audio and visual means.

[0841] The terminal transmits the alert data to the server.

[0842] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0843] Prompt Sentence Examples

[0844] "This call is from City Hall to inform you about your refund. Please go to an ATM and follow the instructions provided."

[0845] This invention allows for a comprehensive security system that takes into account the emotional state of the user as well as detecting fraud patterns.

[0846] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0847] Step 1:

[0848] The terminal acquires the user's voice data. The user's voice is constantly recorded through a microphone and the voice data is acquired. The input is voice data, and the output is also voice data.

[0849] Step 2:

[0850] The device converts the acquired voice data into text data. It uses a voice recognition engine (e.g., Google Recognition API) to convert the voice data into text format. The input is voice data, and the output is text data.

[0851] Step 3:

[0852] The device analyzes text data using generative artificial intelligence (e.g., ChatGPT or BERT) to detect fraud-related keywords and phrases. The input is text data, and the output is the detection results of fraud patterns.

[0853] Step 4:

[0854] The device uses an emotion analysis engine to analyze the user's emotional state. It evaluates the user's emotions from voice and text data and determines the level of stress and anxiety. The input is voice and text data, and the output is the evaluation result of the emotional state.

[0855] Step 5:

[0856] If a fraud pattern is detected and the user's emotional state is unstable, the device generates an alert and notifies the user. Specifically, it generates a message such as "There is a possibility of fraud. Please hang up the phone immediately" and notifies the user visually or audibly. The input is the fraud pattern detection result and the emotional state evaluation result, and the output is the alert message.

[0857] Step 6:

[0858] The terminal sends the generated alert to the server. The alert message is sent to the server via the Internet and recorded in the database. The input is the alert message, and the output is the alert sending completion status.

[0859] Step 7:

[0860] The server analyzes the received alert and notifies pre-registered contacts (family, police, etc.) via email, SMS, push notification, etc. The input is the alert message, and the output is the notification message.

[0861] Specific working example:

[0862] For example, suppose the user speaks the following prompt:

[0863] "This call is from City Hall to inform you about your refund. Please go to an ATM and follow the instructions provided."

[0864] In this case, the system detected a fraud pattern and the user's emotional state indicated high stress, so it generated an alert and notified the user, "This may be a scam. Please hang up immediately." Similar alert messages were also sent to family members and the police.

[0865] 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.

[0866] 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.

[0867] 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.

[0868] [Third embodiment]

[0869] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0870] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0871] 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).

[0872] 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.

[0873] 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.

[0874] 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).

[0875] 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.

[0876] 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.

[0877] 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.

[0878] 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.

[0879] 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.

[0880] 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."

[0881] As an embodiment for carrying out the present invention, specific system operations and program processing will be described below.

[0882] System configuration

[0883] The system consists of a user-held device (a smartphone or dedicated device) and a central server. This device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence (such as ChatGPT), alert sending functionality, and communication functionality. The server has a database and handles alert notifications and data analysis.

[0884] Program processing

[0885] Acquiring voice data and converting it to text

[0886] Subject: Terminal

[0887] The device is always on and picks up the user's speech through a microphone. This voice data is sent to a speech recognition engine in real time and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve high-precision text conversion.

[0888] Text data analysis

[0889] Subject: Terminal

[0890] Generative AI (ChatGPT or equivalent technology) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[0891] Alert generation and user notification

[0892] Subject: Terminal

[0893] If a fraud pattern is detected, the device immediately generates an alert, which includes information about the potential fraud and specific actions the user should take, and is notified to the user through an audio message, a visual message, or both.

[0894] Alert transmission to the server and server response

[0895] Subject: Terminal, Server

[0896] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, push notification, or other methods.

[0897] Specific examples

[0898] Scenario 1: Refund fraud phone call

[0899] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0900] The device records the conversation and converts the audio data into text data.

[0901] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0902] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[0903] The terminal transmits the alert data to the server.

[0904] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0905] Scenario 2: Harmless phone call

[0906] A user is having a normal conversation with a friend.

[0907] The device records the conversation and converts the audio into text data.

[0908] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[0909] The device will not generate an alert and will notify the user as usual.

[0910] System convenience

[0911] The system provides users with a highly intuitive and easy-to-use interface. It specializes in detecting and preventing fraud targeting the elderly, allowing them to live their daily lives with peace of mind. It also notifies third parties, such as family members and the police, so they can respond quickly.

[0912] The above is an embodiment of the present invention. This system can provide an effective means for preventing fraud damage before it occurs.

[0913] The processing flow will be explained below.

[0914] Step 1:

[0915] Subject: Terminal

[0916] The device records the user's conversation in real time through a microphone and temporarily stores the recorded voice data in the device's memory.

[0917] Step 2:

[0918] Subject: Terminal

[0919] The device passes the recorded voice data to a speech recognition engine, which converts it into text data. The speech recognition engine performs noise filtering and voice clearing to convert the conversation into text with high accuracy.

[0920] Step 3:

[0921] Subject: Terminal

[0922] Generative AI (ChatGPT) analyzes text data and extracts key phrases and keywords, which are then compared with past fraud patterns.

[0923] Step 4:

[0924] Subject: Terminal

[0925] If a fraud pattern is detected, the device will immediately generate an alert that includes the likelihood of fraud and a recommended action, such as "This call may be fraudulent. Please hang up immediately."

[0926] Step 5:

[0927] Subject: Terminal

[0928] The device notifies the user of the generated alert either audibly or visually. In the case of an audio alert, the user is warned through the device's speaker, and in the case of a visual alert, a warning message is displayed on the screen.

[0929] Step 6:

[0930] Subject: Terminal

[0931] The terminal transmits alert data to the server via the network. The alert data includes the alert message as well as information about the user's current situation (e.g., location information).

[0932] Step 7:

[0933] Subject: Server

[0934] The server receives the alert data sent from the device, analyzes the received alert data, and determines which third parties (family members or the police) need to be notified.

[0935] Step 8:

[0936] Subject: Server

[0937] The server notifies pre-registered third parties of the alert data via email, SMS, push notification, etc. The notification content includes the alert message and current user information.

[0938] Step 9:

[0939] Subject: Family / Police

[0940] Family members and police receive notifications from the server and can respond quickly based on the content of the notification, for example, by contacting the user or heading to the scene.

[0941] The above is a description of the processing steps of the system and their specific operations. By explaining the specific processing performed at each step in detail, the method of implementing the present invention and the specific content of its operations will become clear.

[0942] Example 1

[0943] 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."

[0944] In recent years, telephone fraud has been on the rise, with elderly people being particularly targeted. Early detection and warnings are essential to prevent these types of fraud. However, current manual countermeasures tend to miss the timing, and it is difficult for elderly people to make the right decision. Therefore, there is a need to develop a system that automatically detects potential fraud using user voice data and issues an immediate warning.

[0945] 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.

[0946] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into character string data, means for analyzing the character string data and detecting fraudulent activity patterns, means for generating a warning and notifying the user when a fraudulent activity pattern is detected, means for transmitting the warning to the server, means for the server to receive the warning and notify a registered third party, means for converting voice data into character string data using a voice recognition engine, means for analyzing the character string data using a generative AI model and detecting fraudulent activity patterns, and means for notifying the user of the warning by voice message or visual message. This makes it possible to automatically detect fraud before a user becomes a victim of fraud and respond quickly.

[0947] "User" refers to an individual who uses the system.

[0948] "Voice data" refers to data that is a digital recording of a user's spoken words.

[0949] "Character string data" refers to data that has been converted from audio data into text format.

[0950] "Fraud patterns" refer to keywords or phrases associated with fraud or other illegal activity.

[0951] "Warning" refers to a message that notifies the user when the system detects possible fraud.

[0952] "Server" refers to a centralized device that manages databases and provides alert notifications.

[0953] "Third party" refers to an individual or organization other than the user (e.g., family members or the police).

[0954] A "voice recognition engine" refers to software that converts voice data into text data in real time.

[0955] A "generative AI model" refers to a model that uses artificial intelligence technology to perform specific tasks based on provided data.

[0956] "Voice message" refers to a warning that is given to the user in the form of a voice.

[0957] "Visual message" refers to a warning that is displayed on the user's terminal screen.

[0958] MODE FOR CARRYING OUT THE INVENTION

[0959] As an embodiment of the present invention, specific system operations and program processing will be described below.

[0960] System configuration

[0961] The system consists of a user-held device (e.g., a smartphone or dedicated device) and a central server. The device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence (e.g., ChatGPT), alert sending functionality, and communication functionality. The server has a database and handles alert notifications and information analysis.

[0962] Program processing

[0963] Acquiring voice data and converting it to text

[0964] Subject: Terminal

[0965] The device is always on and picks up the user's speech through a microphone. This speech data is sent in real time to a speech recognition engine (such as Google Cloud Speech-to-Text or Nuance's Dragon) and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve highly accurate text conversion.

[0966] Text data analysis

[0967] Subject: Terminal

[0968] Generative AI (such as OpenAI's ChatGPT) analyzes the converted text data in real time, extracting keywords and phrases related to fraud and matching them with an internal fraud pattern database that contains information on past fraud cases and indicators.

[0969] Alert generation and user notification

[0970] Subject: Terminal

[0971] If a fraud pattern is detected, the device immediately generates an alert that includes information about the potential fraud and specific actions the user should take. The alert may also include an audio and / or visual message to the user, such as "This call may be fraudulent. Please hang up immediately."

[0972] Alert transmission to the server and server response

[0973] Subject: Terminal and Server

[0974] The generated alert is sent from the device to a server. The server receives the alert and notifies pre-registered third parties, such as family members or the police. Notifications are sent via email, SMS, push notifications, etc. Furthermore, the notification also includes the user's current location information, allowing for a prompt response.

[0975] Specific examples

[0976] Scenario 1: Refund fraud phone call

[0977] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[0978] The device records the conversation and converts the audio data into text data.

[0979] Generative artificial intelligence analyzes string data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[0980] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[0981] The terminal transmits the alert data to the server.

[0982] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[0983] Scenario 2: Harmless phone call

[0984] A user is having a normal conversation with a friend.

[0985] The device records the conversation and converts the audio data into text data.

[0986] Generative AI analyzes the string data but does not detect any keywords or patterns specifically related to fraud.

[0987] The device will not generate an alert and will notify the user as usual.

[0988] Prompt Sentence Examples

[0989] A user receives a call from someone claiming to be a city hall employee saying, "You have a tax refund, so please go to an ATM." Analyze this conversation and determine whether it is a scam.

[0990] The above is an embodiment of the present invention, and the system allows users to automatically detect fraudulent activity before they become victims and take prompt action.

[0991] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0992] Step 1: Acquire audio data

[0993] Subject: Terminal

[0994] Input: Ambient audio environment

[0995] How it works: The device's microphone constantly monitors the surrounding audio. When the user starts speaking, the device captures the audio data and starts recording.

[0996] Output: Captured audio data

[0997] Step 2: Convert audio data to text

[0998] Subject: Terminal

[0999] Input: Captured audio data

[1000] How it works: The device sends voice data in real time to a speech recognition engine (such as Google Cloud Speech-to-Text or Nuance's Dragon), which performs noise filtering and voice clearing and converts the voice data into text data.

[1001] Output: Converted string data

[1002] Step 3: Analyzing the text data

[1003] Subject: Terminal

[1004] Input: Converted string data

[1005] How it works: Generative AI (e.g., ChatGPT) analyzes text data to detect keywords and phrases associated with fraud, which are then compared against an internal fraud pattern database.

[1006] Output: Analysis results (data on possible fraudulent activity)

[1007] Step 4: Alert Generation

[1008] Subject: Terminal

[1009] Input: Analysis results

[1010] Specific Actions: When a fraudulent pattern is detected, the device immediately generates a warning message. The warning message includes information about the potential fraud and specific actions the user should take. For example, a message could read, "This call may be fraudulent. Please hang up immediately." The message can be in both audio and visual formats.

[1011] Output: Generated warning message

[1012] Step 5: User Notification

[1013] Subject: Terminal

[1014] Input: The generated warning message

[1015] Specific operation: The generated warning message is notified to the user via audio and visual means. The user's device will display a text message stating "Possible fraud" along with a voice message saying "This call may be fraudulent. Please hang up immediately."

[1016] Output: The warning message

[1017] Step 6: Sending alerts to the server

[1018] Subject: Terminal

[1019] Input: The generated warning message

[1020] Specific operation: The device sends the generated warning message to a server via the Internet. Communication uses the TCP / IP protocol, and data is encrypted for security.

[1021] Output: Alert data sent

[1022] Step 7: Server Alert Processing and Notification

[1023] Subject: Server

[1024] Input: Alert data sent

[1025] Specific operation: The server analyzes the received alert and notifies pre-registered family members and police. The notification includes the alert message and the user's current location. Notifications can be sent via email, SMS, or push notification.

[1026] Output: Alert message and location information sent to a third party

[1027] (Application example 1)

[1028] 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."

[1029] There is a need for a means to reduce the risk of elderly people and users who are not familiar with technology being caught up in fraudulent phone calls and fraudulent conversations, and to enable them to live their daily lives safely. In particular, it is essential to provide a system that analyzes the content of calls received by users in real time, promptly issues a warning in the event of a possible fraud, and notifies the appropriate third party.

[1030] 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.

[1031] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data using automatic generation technology to detect fraud patterns, means for generating an alert and notifying the user by voice or visual means when a fraud pattern is detected, means for sending the alert to the server, and means for the server to receive the alert and notify a registered third party by email or SMS. This allows users to deal with fraud risks in real time and prevent damage before it occurs through prompt warnings and notifications.

[1032] "User" refers to a person who uses this system.

[1033] "Voice data" refers to information that is a digital recording of a user's voice.

[1034] "Text data" refers to data that has been analyzed and converted into a character string of characters from audio data.

[1035] "Automatic generation technology" refers to technology that uses generative artificial intelligence to generate or analyze information.

[1036] "Fraud patterns" refer to keywords and phrases that indicate the possibility of fraud, based on information about past fraud cases and indicators of fraud.

[1037] "Alert" refers to a warning message provided to a user when potential fraud is detected.

[1038] "Server" refers to a central computer system that has a database and handles alert notifications and data analysis.

[1039] "Registered third parties" refer to family members, police, or other relevant parties that the user has pre-defined.

[1040] "Email" refers to a means of sending and receiving digital messages over the Internet.

[1041] "SMS" refers to a means of sending short text messages using a mobile phone.

[1042] As a specific embodiment of the present invention, the operation of the system and program processing will be described below.

[1043] System configuration

[1044] This system consists of a device used by the user (a smartphone or dedicated device) and a central server. The device is equipped with a voice input device, voice recognition function, generative artificial intelligence using automatic generation technology, alert sending function, and communication function. The server has a database and is responsible for sending alerts and analyzing data.

[1045] Program processing

[1046] Acquiring voice data and converting it to text

[1047] The device is always on and picks up the user's speech via a microphone. This voice data is sent in real time to a speech recognition function and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve highly accurate text conversion. The Google Cloud Speech-to-Text API is used for speech recognition.

[1048] Text data analysis

[1049] Generative AI (ChatGPT or equivalent technology) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[1050] Alert generation and user notification

[1051] If a fraud pattern is detected, the device will immediately generate an alert, which will include information about the potential fraud and specific actions the user should take. The alert will be communicated to the user via audio and / or visual messages, and will be delivered via the device's on-screen and audio output capabilities.

[1052] Alert transmission to the server and server response

[1053] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, or other means.

[1054] Specific examples

[1055] Scenario 1: Refund fraud phone call

[1056] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[1057] The device records the conversation and converts the audio data into text data.

[1058] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[1059] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[1060] The terminal transmits the alert data to the server.

[1061] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[1062] Scenario 2: Harmless phone call

[1063] A user is having a normal conversation with a friend.

[1064] The device records the conversation and converts the audio into text data.

[1065] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[1066] The device will not generate an alert and will notify the user as usual.

[1067] Example prompt sentences to use

[1068] Analyze the likelihood that the following statements are fraudulent.

[1069] The refund is in your account. Please process it at an ATM.

[1070] If you suspect fraud, please explain why.

[1071] This system allows users to deal with fraud risks in real time and prevent damage before it occurs through prompt warnings and notifications. Notifications are also sent to third parties such as family members and the police, allowing for a prompt response.

[1072] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1073] Step 1:

[1074] The device acquires the user's voice data. Specifically, it records the user's conversation in real time through the device's microphone and saves the voice data in digital format. The input is the user's voice, and the output is digital voice data.

[1075] Step 2:

[1076] The device sends the acquired voice data to a voice recognition engine and converts it into text data. Specifically, noise filtering and voice clearing are performed on the voice data to achieve highly accurate text conversion. The Google Cloud Speech-to-Text API is used for voice recognition. The input is digital voice data, and the output is text data.

[1077] Step 3:

[1078] The device uses automated generation technology to send text data to a generative artificial intelligence (AI) for analysis. Specifically, the AI ​​(ChatGPT or equivalent technology) analyzes the text data and extracts fraud-related keywords and phrases. The input is the text data, and the output is a list of fraud-related keywords and phrases.

[1079] Step 4:

[1080] The device detects fraud patterns based on the analysis results from the generative AI. Specifically, it compares the results with a fraud pattern database to determine whether keywords or phrases indicate the possibility of fraud. The input is a list of keywords or phrases, and the output is a judgment result regarding the possibility of fraud.

[1081] Step 5:

[1082] If a fraud pattern is detected, the terminal immediately generates an alert. Specifically, it creates an alert message indicating the possibility of fraud and instructing the user on specific actions to take. The input is the result of the fraud possibility judgment, and the output is the alert message.

[1083] Step 6:

[1084] The terminal notifies the user of the generated alert message by voice and visual means. Specifically, the terminal displays the alert message on the screen and plays an audio alert using the audio output function. The input is the alert message, and the output is the notification to the user.

[1085] Step 7:

[1086] The terminal sends alert data to a central server. Specifically, it sends a data packet containing the alert content and the user's current location information to the server. The input is the alert message and location information, and the output is the data transmission to the server.

[1087] Step 8:

[1088] The server notifies pre-registered third parties based on the received alert data. Specifically, it sends the alert content and location information to family members or the police via email or SMS. The input is the alert data, and the output is a notification to the third party.

[1089] 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.

[1090] As an embodiment for carrying out the present invention, the operation and program processing of a system incorporating an emotion engine will be described below.

[1091] System configuration

[1092] The system consists of a user-held device (a smartphone or dedicated device), a central server, and an emotion engine. The device is equipped with a voice input device, voice recognition functionality, generative AI (such as ChatGPT), the emotion engine, an alert function, and communication functionality. The server has a database and handles alert notifications and data analysis.

[1093] Program processing

[1094] Acquiring voice data and converting it to text

[1095] Subject: Terminal

[1096] The device is always on and picks up the user's speech through a microphone. This voice data is sent to a speech recognition engine in real time and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve high-precision text conversion.

[1097] Text data analysis

[1098] Subject: Terminal

[1099] The generative AI (ChatGPT) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[1100] Acquiring and analyzing emotion data

[1101] Subject: Terminal

[1102] The emotion engine extracts emotional information from the user's voice and input data. The device measures the user's level of stress and anxiety, and combines this with fraud patterns to determine the level of risk. The emotion engine also analyzes the emotional data in real time and responds according to the user's emotional state.

[1103] Alert generation and user notification

[1104] Subject: Terminal

[1105] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the device immediately generates an alert. This alert includes information about the likelihood of fraud, specific recommended actions, and a counseling message based on the user's emotional state. For example, "This call may be fraudulent. Please hang up immediately. Also, you appear to be emotionally upset. Please remain calm."

[1106] Alert transmission to the server and server response

[1107] Subject: Terminal, Server

[1108] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, push notification, or other methods.

[1109] Specific examples

[1110] Scenario 1: Refund fraud phone call

[1111] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[1112] The device records the conversation and converts the audio data into text data.

[1113] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[1114] The emotion engine detects high stress levels from the user's tone of voice and language.

[1115] The device determines that the call is likely to be fraudulent and generates an alert that reads, "This call may be fraudulent. Please hang up immediately. Also, you appear to be upset, so please remain calm and take appropriate action," and notifies the user via audio and visual means.

[1116] The terminal transmits the alert data to the server.

[1117] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[1118] Scenario 2: Regular phone call

[1119] A user is having a normal conversation with a friend.

[1120] The device records the conversation and converts the audio into text data.

[1121] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[1122] The emotion engine determines that the user's emotional state is normal.

[1123] The device will not generate an alert and will notify the user as usual.

[1124] System convenience and scalability

[1125] This system provides users with a highly intuitive and easy-to-use interface. It specializes in detecting and preventing fraud targeting the elderly, allowing them to live their daily lives with peace of mind. It also notifies third parties, such as family members and the police, enabling a prompt response. The introduction of an emotion engine provides a comprehensive crime prevention system that also responds to the user's emotional state.

[1126] The above is an embodiment of the present invention. This system can provide an effective means for preventing fraud damage before it occurs.

[1127] The processing flow will be explained below.

[1128] Step 1:

[1129] Subject: Terminal

[1130] The device records the user's conversation in real time through a microphone and temporarily stores the recorded voice data in the device's memory.

[1131] Step 2:

[1132] Subject: Terminal

[1133] The device passes the recorded voice data to a speech recognition engine for conversion to text data, which performs noise filtering and voice clearing to achieve highly accurate text conversion.

[1134] Step 3:

[1135] Subject: Terminal

[1136] The generative AI (ChatGPT) analyzes text data and extracts key phrases and keywords, which are then compared against an internal fraud pattern database.

[1137] Step 4:

[1138] Subject: Terminal

[1139] The emotion engine analyzes the voice data and extracts emotional information from the user's tone of voice and vocabulary, including levels of stress and anxiety.

[1140] Step 5:

[1141] Subject: Terminal

[1142] The emotion engine combines the extracted emotion information with the results of text analysis to determine the likelihood of fraud and the user's emotional state. If a fraud pattern is detected and the user is determined to be in a high-stress state, the device will immediately generate an alert.

[1143] Step 6:

[1144] Subject: Terminal

[1145] The device then notifies the user of the generated alerts, which may include information about potential fraud, specific recommended actions, and counseling messages based on the user's emotional state. Notifications may be audio or visual.

[1146] Step 7:

[1147] Subject: Terminal

[1148] The terminal transmits alert data to the server via the network. The alert data includes the alert message as well as the user's location information.

[1149] Step 8:

[1150] Subject: Server

[1151] The server receives the alert data sent from the device, analyzes the received alert data, and determines which third parties (family members or the police) need to be notified.

[1152] Step 9:

[1153] Subject: Server

[1154] The server notifies pre-registered third parties of the alert data via email, SMS, push notification, etc. The notification content includes the alert message and current user information.

[1155] Step 10:

[1156] Subject: Family / Police

[1157] Family members and police receive notifications from the server and can respond quickly based on the content of the notification, for example, by contacting the user or heading to the scene.

[1158] The above are the specific processing steps of the system that combines the emotion engine. By explaining the specific processing performed at each step in detail, the method of implementing the present invention and the specific content of its operation will become clear.

[1159] Example 2

[1160] 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."

[1161] In recent years, fraudulent phone scams have become more sophisticated, with an increasing number of victims targeting elderly people in particular. However, conventional fraud prevention systems only detect fraudulent patterns and are unable to respond in a way that takes into account the user's emotional state. As a result, even if users suspect a fraud, they are often unable to take appropriate action due to stress and anxiety. This results in the problem of not being able to prevent fraud victims from occurring.

[1162] 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.

[1163] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data using generative artificial intelligence to detect fraud patterns, means for acquiring and analyzing user emotional data, means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable, means for transmitting the alert to the server, and means for the server to receive the alert and notify a registered third party. This improves the accuracy of fraud detection and enables appropriate responses according to the user's emotional state.

[1164] "User" refers to an individual or end user who uses the System.

[1165] "Voice data" refers to data that is a digital recording of a user's voice.

[1166] "Text data" refers to data that has been converted from voice data into a string of characters using a voice recognition function.

[1167] "Generative artificial intelligence" refers to an artificial intelligence model that learns from large amounts of data and analyzes and generates natural language.

[1168] "Fraud patterns" refer to patterns that include specific keywords or phrases defined based on past fraud cases and indicators of fraud.

[1169] "Emotion data" refers to data related to the user's psychological state or emotions extracted from the user's voice or input data.

[1170] "Alert" means a notification generated when potential fraud is detected.

[1171] "Server" refers to a central management system that manages data and has notification functions.

[1172] "Registered third parties" refers to family members, friends, police, and other related parties who have been pre-registered as recipients of alerts from the server.

[1173] "Location information" refers to data that identifies a user's current geographic location.

[1174] As an embodiment for carrying out the present invention, the operation and program processing of a system incorporating an emotion engine will be described below.

[1175] System configuration

[1176] The system consists of a user-held device (a smartphone or dedicated device), a central server, and an emotion engine. The device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence, the emotion engine, an alert function, and communication functionality. The server has a database and can send alerts and perform data analysis.

[1177] Acquiring voice data and converting it to text

[1178] Subject: Terminal

[1179] The device is always on and picks up the user's speech via a microphone. The speech data is sent to a speech recognition engine such as the Google Speech-to-Text API, where it undergoes noise filtering and voice clearing before being converted into highly accurate text data. For example, a speech such as "Hello, this is City Hall" can be instantly converted into text.

[1180] Text data analysis

[1181] Subject: Terminal

[1182] The converted text data is analyzed by generative artificial intelligence. The device uses this text data to extract keywords and phrases related to fraud. For example, it checks against a database to see if it contains keywords such as "refund," "ATM," or "account information." This database contains information on past fraud cases and signs of fraud, enabling highly accurate fraud detection.

[1183] Acquiring and analyzing emotion data

[1184] Subject: Terminal

[1185] The emotion engine extracts emotional information from the user's voice and input data. The device uses the AWS Comprehend API, for example, to analyze emotions in real time from the user's tone of voice and vocabulary. For example, it can detect the tone of voice when the user is feeling doubtful or stressed and determine that the user is in a high stress state.

[1186] Alert generation and user notification

[1187] Subject: Terminal

[1188] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the device will immediately generate an alert. For example, if keywords such as "refund," "ATM," and "account information" are detected, the device will generate a message to the user via audio and visual notification, such as "This call may be fraudulent. Please hang up immediately. Also, you appear to be emotionally upset, so please remain calm."

[1189] Alert transmission to the server and server response

[1190] Subject: Terminal, Server

[1191] The generated alert is sent from the device to a server. The server receives this alert and sends a notification to a pre-registered third party (family member or police). For example, the server uses the Twilio API to send an emergency message (SMS or email). Push notifications can also be sent using Firebase, and notifications will be sent including the user's current location information.

[1192] Specific examples

[1193] Scenario 1: Refund fraud phone call

[1194] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[1195] The device records the conversation and converts the audio data into text using the Google Speech-to-Text API.

[1196] Generative artificial intelligence analyzes this text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[1197] The emotion engine detects high stress levels from the user's tone of voice and language.

[1198] The device will determine that the call is likely to be fraudulent and will send an alert to the user saying, "This call may be fraudulent. Please hang up immediately. Also, you seem upset, so please remain calm and take action."

[1199] The terminal transmits the alert data to the server.

[1200] The server receives the alert and uses the Twilio API to send notifications to family members and police, including the user's current location.

[1201] Scenario 2: Regular phone call

[1202] A user is having a normal conversation with a friend.

[1203] The device records the conversation and converts the audio into text using the Google Speech-to-Text API.

[1204] Generative AI analyzes text data but does not detect any keywords or patterns related to fraud.

[1205] The emotion engine determines that the user's emotional state is normal.

[1206] The device will not generate an alert and will notify the user as usual.

[1207] Examples of prompt statements

[1208] "Simulate the system's response when a user receives a fraudulent phone call."

[1209] "Please explain how the emotion engine and alert system work when making a regular phone call with a friend."

[1210] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1211] Step 1:

[1212] Subject: Terminal

[1213] The device constantly captures the user's voice data through the microphone. This voice data is input and sent in real time to a voice recognition engine such as the Google Speech-to-Text API. The voice recognition engine analyzes the voice while performing noise filtering and voice clearing, and converts it into a string of characters (text data). The output text data is a specific sentence such as "Hello, this is City Hall."

[1214] Step 2:

[1215] Subject: Terminal

[1216] The device inputs the converted text data into a generative AI (e.g., ChatGPT) and begins analysis. During the analysis process, fraud-related keywords and phrases (e.g., "refund," "ATM," and "account information") are extracted from the text data. The generative AI model compares the data with an internal fraud pattern database to determine the likelihood of fraud. The output is a judgment indicating whether a fraud pattern has been detected.

[1217] Step 3:

[1218] Subject: Terminal

[1219] The device inputs the user's voice and text data into an emotion engine (e.g., AWS Comprehend) and analyzes the emotion data. Specifically, it measures the user's stress and anxiety levels from their voice and vocabulary. The emotion engine performs tone analysis and determines if a high level of stress or unstable emotions is detected. The analysis results are output as emotion data.

[1220] Step 4:

[1221] Subject: Terminal

[1222] The device combines the fraud pattern judgment data detected in step 2 with the emotion data acquired in step 3 to determine the overall risk level. An alert is generated based on this judgment result. The alert contains specific messages such as, "This call may be fraudulent. Please hang up immediately. Also, you seem upset, so please remain calm and take action." The generated alert message is output.

[1223] Step 5:

[1224] Subject: Terminal

[1225] The device sends the generated alert to the server. The sent data also includes the user's location information. The alert message and location information are collected as input data and sent to the server via an appropriate communication protocol. The output data is a notification packet containing the alert message and location information.

[1226] Step 6:

[1227] Subject: Server

[1228] The server processes the alert notification received from the device. It sends an emergency message to pre-registered third parties (e.g., family members or police) using the Twilio API or Firebase push notification. The notification message contains the alert content and the user's current location information. The output data is the emergency notification message sent to the third party.

[1229] (Application example 2)

[1230] 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."

[1231] In modern society, fraud methods have become more sophisticated, making elderly people especially vulnerable to fraud. Fraud methods are so ingenious that simply detecting fraud patterns is not enough. To prevent fraud before it happens, a comprehensive crime prevention system that also takes into account the user's emotional state is necessary. However, current technology has not yet realized a system that can analyze the user's emotional state in real time and notify them of any signs of fraud.

[1232] 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 acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data and detecting fraud patterns, means for analyzing the user's emotional state, means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable, means for transmitting the alert to the server, and means for the server to receive the alert and notify a registered third party. This enables a comprehensive security system that takes into account the user's emotional state as well as the detection of fraud patterns.

[1233] "User voice data" is a digital signal containing the voice uttered by the user.

[1234] The "means for converting acquired voice data into text data" refers to a software and hardware system that utilizes voice recognition technology to convert voice data into a corresponding text format.

[1235] "Means for analyzing text data and detecting fraud patterns" refers to algorithms that analyze text data and identify signs and keywords of fraud based on past fraud cases, etc.

[1236] The "means for analyzing the user's emotional state" is an emotion analysis engine that analyzes the user's emotions such as stress, anxiety, and fear in real time from voice and text.

[1237] "Means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable" refers to a system that generates a warning message and notifies the user when it is determined that there is a high possibility of fraud and the user is in an unstable emotional state.

[1238] The "means for transmitting an alert to a server" is a communication system that securely transmits the generated alert message to a server over the Internet.

[1239] "Means for the server to receive alerts and notify registered third parties" refers to a function that analyzes alerts received on the server and notifies pre-set contacts (for example, family members or the police) via email, SMS, etc.

[1240] System configuration

[1241] The present invention is a system for acquiring and analyzing user voice data, which is composed of the following main components:

[1242] A means of obtaining user voice data

[1243] A means of converting voice data into text data using voice recognition technology

[1244] A method for analyzing text data and detecting fraud patterns using generative artificial intelligence

[1245] A sentiment analysis engine for analyzing users' emotional state in real time

[1246] A means of generating alerts and notifying users

[1247] A means of sending alerts to the server

[1248] A means for the server to receive alerts and notify registered third parties

[1249] Program processing

[1250] The system uses voice recognition technology to capture the user's voice data, then analyzes the text data using generative artificial intelligence models (such as ChatGPT and BERT), detects fraud patterns from the analyzed text data, and evaluates the user's emotional state using a sentiment analysis engine.

[1251] Hardware and Software Details

[1252] Hardware: Smartphone, microphone

[1253] Software: Python, speech_recognition library, transformers library, requests library

[1254] The smartphone's microphone is used to continuously record the user's voice, which is then converted into text in real time by a speech recognition engine. The converted text is then analyzed by generative artificial intelligence to detect fraud patterns, keywords, and phrases. At the same time, a sentiment analysis engine evaluates the user's emotional state from the text and voice to determine whether they are experiencing increased stress or anxiety.

[1255] Alerting and Notifications

[1256] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the system immediately generates an alert and notifies the user. The alert includes the possibility of fraud, specific recommended actions, and a counseling message based on the user's emotional state. The generated alert is sent to the server, which then notifies registered third parties.

[1257] Specific examples

[1258] Scenario 1: Refund fraud phone call

[1259] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[1260] The device records the conversation and converts the audio data into text data.

[1261] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[1262] The emotion engine detects high stress levels from the user's tone of voice and language.

[1263] The device determines that the call is likely to be fraudulent and generates an alert that reads, "This call may be fraudulent. Please hang up immediately. Also, you appear to be upset, so please remain calm and take appropriate action," and notifies the user via audio and visual means.

[1264] The terminal transmits the alert data to the server.

[1265] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[1266] Prompt Sentence Examples

[1267] "This call is from City Hall to inform you about your refund. Please go to an ATM and follow the instructions provided."

[1268] This invention allows for a comprehensive security system that takes into account the emotional state of the user as well as detecting fraud patterns.

[1269] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1270] Step 1:

[1271] The terminal acquires the user's voice data. The user's voice is constantly recorded through a microphone and the voice data is acquired. The input is voice data, and the output is also voice data.

[1272] Step 2:

[1273] The device converts the acquired voice data into text data. It uses a voice recognition engine (e.g., Google Recognition API) to convert the voice data into text format. The input is voice data, and the output is text data.

[1274] Step 3:

[1275] The device analyzes text data using generative artificial intelligence (e.g., ChatGPT or BERT) to detect fraud-related keywords and phrases. The input is text data, and the output is the detection results of fraud patterns.

[1276] Step 4:

[1277] The device uses an emotion analysis engine to analyze the user's emotional state. It evaluates the user's emotions from voice and text data and determines the level of stress and anxiety. The input is voice and text data, and the output is the evaluation result of the emotional state.

[1278] Step 5:

[1279] If a fraud pattern is detected and the user's emotional state is unstable, the device generates an alert and notifies the user. Specifically, it generates a message such as "There is a possibility of fraud. Please hang up the phone immediately" and notifies the user visually or audibly. The input is the fraud pattern detection result and the emotional state evaluation result, and the output is the alert message.

[1280] Step 6:

[1281] The terminal sends the generated alert to the server. The alert message is sent to the server via the Internet and recorded in the database. The input is the alert message, and the output is the alert sending completion status.

[1282] Step 7:

[1283] The server analyzes the received alert and notifies pre-registered contacts (family, police, etc.) via email, SMS, push notification, etc. The input is the alert message, and the output is the notification message.

[1284] Specific working example:

[1285] For example, suppose the user speaks the following prompt:

[1286] "This call is from City Hall to inform you about your refund. Please go to an ATM and follow the instructions provided."

[1287] In this case, the system detected a fraud pattern and the user's emotional state indicated high stress, so it generated an alert and notified the user, "This may be a scam. Please hang up immediately." Similar alert messages were also sent to family members and the police.

[1288] 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.

[1289] 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.

[1290] 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.

[1291] [Fourth embodiment]

[1292] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1293] 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.

[1294] 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).

[1295] 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.

[1296] 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.

[1297] 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).

[1298] 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.

[1299] 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.

[1300] 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.

[1301] 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.

[1302] 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.

[1303] 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.

[1304] 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."

[1305] As an embodiment for carrying out the present invention, specific system operations and program processing will be described below.

[1306] System configuration

[1307] The system consists of a user-held device (a smartphone or dedicated device) and a central server. This device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence (such as ChatGPT), alert sending functionality, and communication functionality. The server has a database and handles alert notifications and data analysis.

[1308] Program processing

[1309] Acquiring voice data and converting it to text

[1310] Subject: Terminal

[1311] The device is always on and picks up the user's speech through a microphone. This voice data is sent to a speech recognition engine in real time and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve high-precision text conversion.

[1312] Text data analysis

[1313] Subject: Terminal

[1314] Generative AI (ChatGPT or equivalent technology) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[1315] Alert generation and user notification

[1316] Subject: Terminal

[1317] If a fraud pattern is detected, the device immediately generates an alert, which includes information about the potential fraud and specific actions the user should take, and is notified to the user through an audio message, a visual message, or both.

[1318] Alert transmission to the server and server response

[1319] Subject: Terminal, Server

[1320] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, push notification, or other methods.

[1321] Specific examples

[1322] Scenario 1: Refund fraud phone call

[1323] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[1324] The device records the conversation and converts the audio data into text data.

[1325] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[1326] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[1327] The terminal transmits the alert data to the server.

[1328] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[1329] Scenario 2: Harmless phone call

[1330] A user is having a normal conversation with a friend.

[1331] The device records the conversation and converts the audio into text data.

[1332] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[1333] The device will not generate an alert and will notify the user as usual.

[1334] System convenience

[1335] The system provides users with a highly intuitive and easy-to-use interface. It specializes in detecting and preventing fraud targeting the elderly, allowing them to live their daily lives with peace of mind. It also notifies third parties, such as family members and the police, so they can respond quickly.

[1336] The above is an embodiment of the present invention. This system can provide an effective means for preventing fraud damage before it occurs.

[1337] The processing flow will be explained below.

[1338] Step 1:

[1339] Subject: Terminal

[1340] The device records the user's conversation in real time through a microphone and temporarily stores the recorded voice data in the device's memory.

[1341] Step 2:

[1342] Subject: Terminal

[1343] The device passes the recorded voice data to a speech recognition engine, which converts it into text data. The speech recognition engine performs noise filtering and voice clearing to convert the conversation into text with high accuracy.

[1344] Step 3:

[1345] Subject: Terminal

[1346] Generative AI (ChatGPT) analyzes text data and extracts key phrases and keywords, which are then compared with past fraud patterns.

[1347] Step 4:

[1348] Subject: Terminal

[1349] If a fraud pattern is detected, the device will immediately generate an alert that includes the likelihood of fraud and a recommended action, such as "This call may be fraudulent. Please hang up immediately."

[1350] Step 5:

[1351] Subject: Terminal

[1352] The device notifies the user of the generated alert either audibly or visually. In the case of an audio alert, the user is warned through the device's speaker, and in the case of a visual alert, a warning message is displayed on the screen.

[1353] Step 6:

[1354] Subject: Terminal

[1355] The terminal transmits alert data to the server via the network. The alert data includes the alert message as well as information about the user's current situation (e.g., location information).

[1356] Step 7:

[1357] Subject: Server

[1358] The server receives the alert data sent from the device, analyzes the received alert data, and determines which third parties (family members or the police) need to be notified.

[1359] Step 8:

[1360] Subject: Server

[1361] The server notifies pre-registered third parties of the alert data via email, SMS, push notification, etc. The notification content includes the alert message and current user information.

[1362] Step 9:

[1363] Subject: Family / Police

[1364] Family members and police receive notifications from the server and can respond quickly based on the content of the notification, for example, by contacting the user or heading to the scene.

[1365] The above is a description of the processing steps of the system and their specific operations. By explaining the specific processing performed at each step in detail, the method of implementing the present invention and the specific content of its operations will become clear.

[1366] Example 1

[1367] 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."

[1368] In recent years, telephone fraud has been on the rise, with elderly people being particularly targeted. Early detection and warnings are essential to prevent these types of fraud. However, current manual countermeasures tend to miss the timing, and it is difficult for elderly people to make the right decision. Therefore, there is a need to develop a system that automatically detects potential fraud using user voice data and issues an immediate warning.

[1369] 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.

[1370] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into character string data, means for analyzing the character string data and detecting fraudulent activity patterns, means for generating a warning and notifying the user when a fraudulent activity pattern is detected, means for transmitting the warning to the server, means for the server to receive the warning and notify a registered third party, means for converting voice data into character string data using a voice recognition engine, means for analyzing the character string data using a generative AI model and detecting fraudulent activity patterns, and means for notifying the user of the warning by voice message or visual message. This makes it possible to automatically detect fraud before a user becomes a victim of fraud and respond quickly.

[1371] "User" refers to an individual who uses the system.

[1372] "Voice data" refers to data that is a digital recording of a user's spoken words.

[1373] "Character string data" refers to data that has been converted from audio data into text format.

[1374] "Fraud patterns" refer to keywords or phrases associated with fraud or other illegal activity.

[1375] "Warning" refers to a message that notifies the user when the system detects possible fraud.

[1376] "Server" refers to a centralized device that manages databases and provides alert notifications.

[1377] "Third party" refers to an individual or organization other than the user (e.g., family members or the police).

[1378] A "voice recognition engine" refers to software that converts voice data into text data in real time.

[1379] A "generative AI model" refers to a model that uses artificial intelligence technology to perform specific tasks based on provided data.

[1380] "Voice message" refers to a warning that is given to the user in the form of a voice.

[1381] "Visual message" refers to a warning that is displayed on the user's terminal screen.

[1382] MODE FOR CARRYING OUT THE INVENTION

[1383] As an embodiment of the present invention, specific system operations and program processing will be described below.

[1384] System configuration

[1385] The system consists of a user-held device (e.g., a smartphone or dedicated device) and a central server. The device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence (e.g., ChatGPT), alert sending functionality, and communication functionality. The server has a database and handles alert notifications and information analysis.

[1386] Program processing

[1387] Acquiring voice data and converting it to text

[1388] Subject: Terminal

[1389] The device is always on and picks up the user's speech through a microphone. This speech data is sent in real time to a speech recognition engine (such as Google Cloud Speech-to-Text or Nuance's Dragon) and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve highly accurate text conversion.

[1390] Text data analysis

[1391] Subject: Terminal

[1392] Generative AI (such as OpenAI's ChatGPT) analyzes the converted text data in real time, extracting keywords and phrases related to fraud and matching them with an internal fraud pattern database that contains information on past fraud cases and indicators.

[1393] Alert generation and user notification

[1394] Subject: Terminal

[1395] If a fraud pattern is detected, the device immediately generates an alert that includes information about the potential fraud and specific actions the user should take. The alert may also include an audio and / or visual message to the user, such as "This call may be fraudulent. Please hang up immediately."

[1396] Alert transmission to the server and server response

[1397] Subject: Terminal and Server

[1398] The generated alert is sent from the device to a server. The server receives the alert and notifies pre-registered third parties, such as family members or the police. Notifications are sent via email, SMS, push notifications, etc. Furthermore, the notification also includes the user's current location information, allowing for a prompt response.

[1399] Specific examples

[1400] Scenario 1: Refund fraud phone call

[1401] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[1402] The device records the conversation and converts the audio data into text data.

[1403] Generative artificial intelligence analyzes string data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[1404] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[1405] The terminal transmits the alert data to the server.

[1406] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[1407] Scenario 2: Harmless phone call

[1408] A user is having a normal conversation with a friend.

[1409] The device records the conversation and converts the audio data into text data.

[1410] Generative AI analyzes the string data but does not detect any keywords or patterns specifically related to fraud.

[1411] The device will not generate an alert and will notify the user as usual.

[1412] Prompt Sentence Examples

[1413] A user receives a call from someone claiming to be a city hall employee saying, "You have a tax refund, so please go to an ATM." Analyze this conversation and determine whether it is a scam.

[1414] The above is an embodiment of the present invention, and the system allows users to automatically detect fraudulent activity before they become victims and take prompt action.

[1415] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1416] Step 1: Acquire audio data

[1417] Subject: Terminal

[1418] Input: Ambient audio environment

[1419] How it works: The device's microphone constantly monitors the surrounding audio. When the user starts speaking, the device captures the audio data and starts recording.

[1420] Output: Captured audio data

[1421] Step 2: Convert audio data to text

[1422] Subject: Terminal

[1423] Input: Captured audio data

[1424] How it works: The device sends voice data in real time to a speech recognition engine (such as Google Cloud Speech-to-Text or Nuance's Dragon), which performs noise filtering and voice clearing and converts the voice data into text data.

[1425] Output: Converted string data

[1426] Step 3: Analyzing the text data

[1427] Subject: Terminal

[1428] Input: Converted string data

[1429] How it works: Generative AI (e.g., ChatGPT) analyzes text data to detect keywords and phrases associated with fraud, which are then compared against an internal fraud pattern database.

[1430] Output: Analysis results (data on possible fraudulent activity)

[1431] Step 4: Alert Generation

[1432] Subject: Terminal

[1433] Input: Analysis results

[1434] Specific Actions: When a fraudulent pattern is detected, the device immediately generates a warning message. The warning message includes information about the potential fraud and specific actions the user should take. For example, a message could read, "This call may be fraudulent. Please hang up immediately." The message can be in both audio and visual formats.

[1435] Output: Generated warning message

[1436] Step 5: User Notification

[1437] Subject: Terminal

[1438] Input: The generated warning message

[1439] Specific operation: The generated warning message is notified to the user via audio and visual means. The user's device will display a text message stating "Possible fraud" along with a voice message saying "This call may be fraudulent. Please hang up immediately."

[1440] Output: The warning message

[1441] Step 6: Sending alerts to the server

[1442] Subject: Terminal

[1443] Input: The generated warning message

[1444] Specific operation: The device sends the generated warning message to a server via the Internet. Communication uses the TCP / IP protocol, and data is encrypted for security.

[1445] Output: Alert data sent

[1446] Step 7: Server Alert Processing and Notification

[1447] Subject: Server

[1448] Input: Alert data sent

[1449] Specific operation: The server analyzes the received alert and notifies pre-registered family members and police. The notification includes the alert message and the user's current location. Notifications can be sent via email, SMS, or push notification.

[1450] Output: Alert message and location information sent to a third party

[1451] (Application example 1)

[1452] 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."

[1453] There is a need for a means to reduce the risk of elderly people and users who are not familiar with technology being caught up in fraudulent phone calls and fraudulent conversations, and to enable them to live their daily lives safely. In particular, it is essential to provide a system that analyzes the content of calls received by users in real time, promptly issues a warning in the event of a possible fraud, and notifies the appropriate third party.

[1454] 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.

[1455] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data using automatic generation technology to detect fraud patterns, means for generating an alert and notifying the user by voice or visual means when a fraud pattern is detected, means for sending the alert to the server, and means for the server to receive the alert and notify a registered third party by email or SMS. This allows users to deal with fraud risks in real time and prevent damage before it occurs through prompt warnings and notifications.

[1456] "User" refers to a person who uses this system.

[1457] "Voice data" refers to information that is a digital recording of a user's voice.

[1458] "Text data" refers to data that has been analyzed and converted into a character string of characters from audio data.

[1459] "Automatic generation technology" refers to technology that uses generative artificial intelligence to generate or analyze information.

[1460] "Fraud patterns" refer to keywords and phrases that indicate the possibility of fraud, based on information about past fraud cases and indicators of fraud.

[1461] "Alert" refers to a warning message provided to a user when potential fraud is detected.

[1462] "Server" refers to a central computer system that has a database and handles alert notifications and data analysis.

[1463] "Registered third parties" refer to family members, police, or other relevant parties that the user has pre-defined.

[1464] "Email" refers to a means of sending and receiving digital messages over the Internet.

[1465] "SMS" refers to a means of sending short text messages using a mobile phone.

[1466] As a specific embodiment of the present invention, the operation of the system and program processing will be described below.

[1467] System configuration

[1468] This system consists of a device used by the user (a smartphone or dedicated device) and a central server. The device is equipped with a voice input device, voice recognition function, generative artificial intelligence using automatic generation technology, alert sending function, and communication function. The server has a database and is responsible for sending alerts and analyzing data.

[1469] Program processing

[1470] Acquiring voice data and converting it to text

[1471] The device is always on and picks up the user's speech via a microphone. This voice data is sent in real time to a speech recognition function and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve highly accurate text conversion. The Google Cloud Speech-to-Text API is used for speech recognition.

[1472] Text data analysis

[1473] Generative AI (ChatGPT or equivalent technology) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[1474] Alert generation and user notification

[1475] If a fraud pattern is detected, the device will immediately generate an alert, which will include information about the potential fraud and specific actions the user should take. The alert will be communicated to the user via audio and / or visual messages, and will be delivered via the device's on-screen and audio output capabilities.

[1476] Alert transmission to the server and server response

[1477] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, or other means.

[1478] Specific examples

[1479] Scenario 1: Refund fraud phone call

[1480] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[1481] The device records the conversation and converts the audio data into text data.

[1482] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[1483] The device determines that the call is likely to be fraudulent and generates an alert saying, "This call may be fraudulent. Please hang up immediately," and notifies the user with audio and visual alerts.

[1484] The terminal transmits the alert data to the server.

[1485] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[1486] Scenario 2: Harmless phone call

[1487] A user is having a normal conversation with a friend.

[1488] The device records the conversation and converts the audio into text data.

[1489] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[1490] The device will not generate an alert and will notify the user as usual.

[1491] Example prompt sentences to use

[1492] Analyze the likelihood that the following statements are fraudulent.

[1493] The refund is in your account. Please process it at an ATM.

[1494] If you suspect fraud, please explain why.

[1495] This system allows users to deal with fraud risks in real time and prevent damage before it occurs through prompt warnings and notifications. Notifications are also sent to third parties such as family members and the police, allowing for a prompt response.

[1496] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1497] Step 1:

[1498] The device acquires the user's voice data. Specifically, it records the user's conversation in real time through the device's microphone and saves the voice data in digital format. The input is the user's voice, and the output is digital voice data.

[1499] Step 2:

[1500] The device sends the acquired voice data to a voice recognition engine and converts it into text data. Specifically, noise filtering and voice clearing are performed on the voice data to achieve highly accurate text conversion. The Google Cloud Speech-to-Text API is used for voice recognition. The input is digital voice data, and the output is text data.

[1501] Step 3:

[1502] The device uses automated generation technology to send text data to a generative artificial intelligence (AI) for analysis. Specifically, the AI ​​(ChatGPT or equivalent technology) analyzes the text data and extracts fraud-related keywords and phrases. The input is the text data, and the output is a list of fraud-related keywords and phrases.

[1503] Step 4:

[1504] The device detects fraud patterns based on the analysis results from the generative AI. Specifically, it compares the results with a fraud pattern database to determine whether keywords or phrases indicate the possibility of fraud. The input is a list of keywords or phrases, and the output is a judgment result regarding the possibility of fraud.

[1505] Step 5:

[1506] If a fraud pattern is detected, the terminal immediately generates an alert. Specifically, it creates an alert message indicating the possibility of fraud and instructing the user on specific actions to take. The input is the result of the fraud possibility judgment, and the output is the alert message.

[1507] Step 6:

[1508] The terminal notifies the user of the generated alert message by voice and visual means. Specifically, the terminal displays the alert message on the screen and plays an audio alert using the audio output function. The input is the alert message, and the output is the notification to the user.

[1509] Step 7:

[1510] The terminal sends alert data to a central server. Specifically, it sends a data packet containing the alert content and the user's current location information to the server. The input is the alert message and location information, and the output is the data transmission to the server.

[1511] Step 8:

[1512] The server notifies pre-registered third parties based on the received alert data. Specifically, it sends the alert content and location information to family members or the police via email or SMS. The input is the alert data, and the output is a notification to the third party.

[1513] 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.

[1514] As an embodiment for carrying out the present invention, the operation and program processing of a system incorporating an emotion engine will be described below.

[1515] System configuration

[1516] The system consists of a user-held device (a smartphone or dedicated device), a central server, and an emotion engine. The device is equipped with a voice input device, voice recognition functionality, generative AI (such as ChatGPT), the emotion engine, an alert function, and communication functionality. The server has a database and handles alert notifications and data analysis.

[1517] Program processing

[1518] Acquiring voice data and converting it to text

[1519] Subject: Terminal

[1520] The device is always on and picks up the user's speech through a microphone. This voice data is sent to a speech recognition engine in real time and converted into text data. The speech recognition engine is equipped with noise filtering and voice clearing functions to achieve high-precision text conversion.

[1521] Text data analysis

[1522] Subject: Terminal

[1523] The generative AI (ChatGPT) analyzes text data in real time, extracting fraud-related keywords and phrases and matching them with an internal fraud pattern database, which contains information on past fraud cases and indicators of fraud.

[1524] Acquiring and analyzing emotion data

[1525] Subject: Terminal

[1526] The emotion engine extracts emotional information from the user's voice and input data. The device measures the user's level of stress and anxiety, and combines this with fraud patterns to determine the level of risk. The emotion engine also analyzes the emotional data in real time and responds according to the user's emotional state.

[1527] Alert generation and user notification

[1528] Subject: Terminal

[1529] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the device immediately generates an alert. This alert includes information about the likelihood of fraud, specific recommended actions, and a counseling message based on the user's emotional state. For example, "This call may be fraudulent. Please hang up immediately. Also, you appear to be emotionally upset. Please remain calm."

[1530] Alert transmission to the server and server response

[1531] Subject: Terminal, Server

[1532] The generated alert is sent from the device to a server, which then notifies pre-registered third parties, such as family members or the police, via email, SMS, push notification, or other methods.

[1533] Specific examples

[1534] Scenario 1: Refund fraud phone call

[1535] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[1536] The device records the conversation and converts the audio data into text data.

[1537] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[1538] The emotion engine detects high stress levels from the user's tone of voice and language.

[1539] The device determines that the call is likely to be fraudulent and generates an alert that reads, "This call may be fraudulent. Please hang up immediately. Also, you appear to be upset, so please remain calm and take appropriate action," and notifies the user via audio and visual means.

[1540] The terminal transmits the alert data to the server.

[1541] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[1542] Scenario 2: Regular phone call

[1543] A user is having a normal conversation with a friend.

[1544] The device records the conversation and converts the audio into text data.

[1545] Generative AI analyzes text data but does not detect any keywords or patterns specifically related to fraud.

[1546] The emotion engine determines that the user's emotional state is normal.

[1547] The device will not generate an alert and will notify the user as usual.

[1548] System convenience and scalability

[1549] This system provides users with a highly intuitive and easy-to-use interface. It specializes in detecting and preventing fraud targeting the elderly, allowing them to live their daily lives with peace of mind. It also notifies third parties, such as family members and the police, enabling a prompt response. The introduction of an emotion engine provides a comprehensive crime prevention system that also responds to the user's emotional state.

[1550] The above is an embodiment of the present invention. This system can provide an effective means for preventing fraud damage before it occurs.

[1551] The processing flow will be explained below.

[1552] Step 1:

[1553] Subject: Terminal

[1554] The device records the user's conversation in real time through a microphone and temporarily stores the recorded voice data in the device's memory.

[1555] Step 2:

[1556] Subject: Terminal

[1557] The device passes the recorded voice data to a speech recognition engine for conversion to text data, which performs noise filtering and voice clearing to achieve highly accurate text conversion.

[1558] Step 3:

[1559] Subject: Terminal

[1560] The generative AI (ChatGPT) analyzes text data and extracts key phrases and keywords, which are then compared against an internal fraud pattern database.

[1561] Step 4:

[1562] Subject: Terminal

[1563] The emotion engine analyzes the voice data and extracts emotional information from the user's tone of voice and vocabulary, including levels of stress and anxiety.

[1564] Step 5:

[1565] Subject: Terminal

[1566] The emotion engine combines the extracted emotion information with the results of text analysis to determine the likelihood of fraud and the user's emotional state. If a fraud pattern is detected and the user is determined to be in a high-stress state, the device will immediately generate an alert.

[1567] Step 6:

[1568] Subject: Terminal

[1569] The device then notifies the user of the generated alerts, which may include information about potential fraud, specific recommended actions, and counseling messages based on the user's emotional state. Notifications may be audio or visual.

[1570] Step 7:

[1571] Subject: Terminal

[1572] The terminal transmits alert data to the server via the network. The alert data includes the alert message as well as the user's location information.

[1573] Step 8:

[1574] Subject: Server

[1575] The server receives the alert data sent from the device, analyzes the received alert data, and determines which third parties (family members or the police) need to be notified.

[1576] Step 9:

[1577] Subject: Server

[1578] The server notifies pre-registered third parties of the alert data via email, SMS, push notification, etc. The notification content includes the alert message and current user information.

[1579] Step 10:

[1580] Subject: Family / Police

[1581] Family members and police receive notifications from the server and can respond quickly based on the content of the notification, for example, by contacting the user or heading to the scene.

[1582] The above are the specific processing steps of the system that combines the emotion engine. By explaining the specific processing performed at each step in detail, the method of implementing the present invention and the specific content of its operation will become clear.

[1583] Example 2

[1584] 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."

[1585] In recent years, fraudulent phone scams have become more sophisticated, with an increasing number of victims targeting elderly people in particular. However, conventional fraud prevention systems only detect fraudulent patterns and are unable to respond in a way that takes into account the user's emotional state. As a result, even if users suspect a fraud, they are often unable to take appropriate action due to stress and anxiety. This results in the problem of not being able to prevent fraud victims from occurring.

[1586] 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.

[1587] In this invention, the server includes means for acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data using generative artificial intelligence to detect fraud patterns, means for acquiring and analyzing user emotional data, means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable, means for transmitting the alert to the server, and means for the server to receive the alert and notify a registered third party. This improves the accuracy of fraud detection and enables appropriate responses according to the user's emotional state.

[1588] "User" refers to an individual or end user who uses the System.

[1589] "Voice data" refers to data that is a digital recording of a user's voice.

[1590] "Text data" refers to data that has been converted from voice data into a string of characters using a voice recognition function.

[1591] "Generative artificial intelligence" refers to an artificial intelligence model that learns from large amounts of data and analyzes and generates natural language.

[1592] "Fraud patterns" refer to patterns that include specific keywords or phrases defined based on past fraud cases and indicators of fraud.

[1593] "Emotion data" refers to data related to the user's psychological state or emotions extracted from the user's voice or input data.

[1594] "Alert" means a notification generated when potential fraud is detected.

[1595] "Server" refers to a central management system that manages data and has notification functions.

[1596] "Registered third parties" refers to family members, friends, police, and other related parties who have been pre-registered as recipients of alerts from the server.

[1597] "Location information" refers to data that identifies a user's current geographic location.

[1598] As an embodiment for carrying out the present invention, the operation and program processing of a system incorporating an emotion engine will be described below.

[1599] System configuration

[1600] The system consists of a user-held device (a smartphone or dedicated device), a central server, and an emotion engine. The device is equipped with a voice input device, voice recognition functionality, generative artificial intelligence, the emotion engine, an alert function, and communication functionality. The server has a database and can send alerts and perform data analysis.

[1601] Acquiring voice data and converting it to text

[1602] Subject: Terminal

[1603] The device is always on and picks up the user's speech via a microphone. The speech data is sent to a speech recognition engine such as the Google Speech-to-Text API, where it undergoes noise filtering and voice clearing before being converted into highly accurate text data. For example, a speech such as "Hello, this is City Hall" can be instantly converted into text.

[1604] Text data analysis

[1605] Subject: Terminal

[1606] The converted text data is analyzed by generative artificial intelligence. The device uses this text data to extract keywords and phrases related to fraud. For example, it checks against a database to see if it contains keywords such as "refund," "ATM," or "account information." This database contains information on past fraud cases and signs of fraud, enabling highly accurate fraud detection.

[1607] Acquiring and analyzing emotion data

[1608] Subject: Terminal

[1609] The emotion engine extracts emotional information from the user's voice and input data. The device uses the AWS Comprehend API, for example, to analyze emotions in real time from the user's tone of voice and vocabulary. For example, it can detect the tone of voice when the user is feeling doubtful or stressed and determine that the user is in a high stress state.

[1610] Alert generation and user notification

[1611] Subject: Terminal

[1612] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the device will immediately generate an alert. For example, if keywords such as "refund," "ATM," and "account information" are detected, the device will generate a message to the user via audio and visual notification, such as "This call may be fraudulent. Please hang up immediately. Also, you appear to be emotionally upset, so please remain calm."

[1613] Alert transmission to the server and server response

[1614] Subject: Terminal, Server

[1615] The generated alert is sent from the device to a server. The server receives this alert and sends a notification to a pre-registered third party (family member or police). For example, the server uses the Twilio API to send an emergency message (SMS or email). Push notifications can also be sent using Firebase, and notifications will be sent including the user's current location information.

[1616] Specific examples

[1617] Scenario 1: Refund fraud phone call

[1618] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[1619] The device records the conversation and converts the audio data into text using the Google Speech-to-Text API.

[1620] Generative artificial intelligence analyzes this text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[1621] The emotion engine detects high stress levels from the user's tone of voice and language.

[1622] The device will determine that the call is likely to be fraudulent and will send an alert to the user saying, "This call may be fraudulent. Please hang up immediately. Also, you seem upset, so please remain calm and take action."

[1623] The terminal transmits the alert data to the server.

[1624] The server receives the alert and uses the Twilio API to send notifications to family members and police, including the user's current location.

[1625] Scenario 2: Regular phone call

[1626] A user is having a normal conversation with a friend.

[1627] The device records the conversation and converts the audio into text using the Google Speech-to-Text API.

[1628] Generative AI analyzes text data but does not detect any keywords or patterns related to fraud.

[1629] The emotion engine determines that the user's emotional state is normal.

[1630] The device will not generate an alert and will notify the user as usual.

[1631] Examples of prompt statements

[1632] "Simulate the system's response when a user receives a fraudulent phone call."

[1633] "Please explain how the emotion engine and alert system work when making a regular phone call with a friend."

[1634] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1635] Step 1:

[1636] Subject: Terminal

[1637] The device constantly captures the user's voice data through the microphone. This voice data is input and sent in real time to a voice recognition engine such as the Google Speech-to-Text API. The voice recognition engine analyzes the voice while performing noise filtering and voice clearing, and converts it into a string of characters (text data). The output text data is a specific sentence such as "Hello, this is City Hall."

[1638] Step 2:

[1639] Subject: Terminal

[1640] The device inputs the converted text data into a generative AI (e.g., ChatGPT) and begins analysis. During the analysis process, fraud-related keywords and phrases (e.g., "refund," "ATM," and "account information") are extracted from the text data. The generative AI model compares the data with an internal fraud pattern database to determine the likelihood of fraud. The output is a judgment indicating whether a fraud pattern has been detected.

[1641] Step 3:

[1642] Subject: Terminal

[1643] The device inputs the user's voice and text data into an emotion engine (e.g., AWS Comprehend) and analyzes the emotion data. Specifically, it measures the user's stress and anxiety levels from their voice and vocabulary. The emotion engine performs tone analysis and determines if a high level of stress or unstable emotions is detected. The analysis results are output as emotion data.

[1644] Step 4:

[1645] Subject: Terminal

[1646] The device combines the fraud pattern judgment data detected in step 2 with the emotion data acquired in step 3 to determine the overall risk level. An alert is generated based on this judgment result. The alert contains specific messages such as, "This call may be fraudulent. Please hang up immediately. Also, you seem upset, so please remain calm and take action." The generated alert message is output.

[1647] Step 5:

[1648] Subject: Terminal

[1649] The device sends the generated alert to the server. The sent data also includes the user's location information. The alert message and location information are collected as input data and sent to the server via an appropriate communication protocol. The output data is a notification packet containing the alert message and location information.

[1650] Step 6:

[1651] Subject: Server

[1652] The server processes the alert notification received from the device. It sends an emergency message to pre-registered third parties (e.g., family members or police) using the Twilio API or Firebase push notification. The notification message contains the alert content and the user's current location information. The output data is the emergency notification message sent to the third party.

[1653] (Application example 2)

[1654] 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."

[1655] In modern society, fraud methods have become more sophisticated, making elderly people especially vulnerable to fraud. Fraud methods are so ingenious that simply detecting fraud patterns is not enough. To prevent fraud before it happens, a comprehensive crime prevention system that also takes into account the user's emotional state is necessary. However, current technology has not yet realized a system that can analyze the user's emotional state in real time and notify them of any signs of fraud.

[1656] 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 acquiring user voice data, means for converting the acquired voice data into text data, means for analyzing the text data and detecting fraud patterns, means for analyzing the user's emotional state, means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable, means for transmitting the alert to the server, and means for the server to receive the alert and notify a registered third party. This enables a comprehensive security system that takes into account the user's emotional state as well as the detection of fraud patterns.

[1657] "User voice data" is a digital signal containing the voice uttered by the user.

[1658] The "means for converting acquired voice data into text data" refers to a software and hardware system that utilizes voice recognition technology to convert voice data into a corresponding text format.

[1659] "Means for analyzing text data and detecting fraud patterns" refers to algorithms that analyze text data and identify signs and keywords of fraud based on past fraud cases, etc.

[1660] The "means for analyzing the user's emotional state" is an emotion analysis engine that analyzes the user's emotions such as stress, anxiety, and fear in real time from voice and text.

[1661] "Means for generating an alert and notifying the user when a fraud pattern is detected and the user's emotional state is unstable" refers to a system that generates a warning message and notifies the user when it is determined that there is a high possibility of fraud and the user is in an unstable emotional state.

[1662] The "means for transmitting an alert to a server" is a communication system that securely transmits the generated alert message to a server over the Internet.

[1663] "Means for the server to receive alerts and notify registered third parties" refers to a function that analyzes alerts received on the server and notifies pre-set contacts (for example, family members or the police) via email, SMS, etc.

[1664] System configuration

[1665] The present invention is a system for acquiring and analyzing user voice data, which is composed of the following main components:

[1666] A means of obtaining user voice data

[1667] A means of converting voice data into text data using voice recognition technology

[1668] A method for analyzing text data and detecting fraud patterns using generative artificial intelligence

[1669] A sentiment analysis engine for analyzing users' emotional state in real time

[1670] A means of generating alerts and notifying users

[1671] A means of sending alerts to the server

[1672] A means for the server to receive alerts and notify registered third parties

[1673] Program processing

[1674] The system uses voice recognition technology to capture the user's voice data, then analyzes the text data using generative artificial intelligence models (such as ChatGPT and BERT), detects fraud patterns from the analyzed text data, and evaluates the user's emotional state using a sentiment analysis engine.

[1675] Hardware and Software Details

[1676] Hardware: Smartphone, microphone

[1677] Software: Python, speech_recognition library, transformers library, requests library

[1678] The smartphone's microphone is used to continuously record the user's voice, which is then converted into text in real time by a speech recognition engine. The converted text is then analyzed by generative artificial intelligence to detect fraud patterns, keywords, and phrases. At the same time, a sentiment analysis engine evaluates the user's emotional state from the text and voice to determine whether they are experiencing increased stress or anxiety.

[1679] Alerting and Notifications

[1680] If a fraud pattern is detected and the user's emotional state is determined to be unstable, the system immediately generates an alert and notifies the user. The alert includes the possibility of fraud, specific recommended actions, and a counseling message based on the user's emotional state. The generated alert is sent to the server, which then notifies registered third parties.

[1681] Specific examples

[1682] Scenario 1: Refund fraud phone call

[1683] The user receives a call from someone claiming to be a "city hall employee" who explains about the refund.

[1684] The device records the conversation and converts the audio data into text data.

[1685] Generative artificial intelligence analyzes text data and detects fraud-related keywords such as "refund," "ATM," and "account information."

[1686] The emotion engine detects high stress levels from the user's tone of voice and language.

[1687] The device determines that the call is likely to be fraudulent and generates an alert that reads, "This call may be fraudulent. Please hang up immediately. Also, you appear to be upset, so please remain calm and take appropriate action," and notifies the user via audio and visual means.

[1688] The terminal transmits the alert data to the server.

[1689] The server receives the alert and notifies family members or the police, which includes the alert message along with the user's current location.

[1690] Prompt Sentence Examples

[1691] "This call is from City Hall to inform you about your refund. Please go to an ATM and follow the instructions provided."

[1692] This invention allows for a comprehensive security system that takes into account the emotional state of the user as well as detecting fraud patterns.

[1693] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1694] Step 1:

[1695] The terminal acquires the user's voice data. The user's voice is constantly recorded through a microphone and the voice data is acquired. The input is voice data, and the output is also voice data.

[1696] Step 2:

[1697] The device converts the acquired voice data into text data. It uses a voice recognition engine (e.g., Google Recognition API) to convert the voice data into text format. The input is voice data, and the output is text data.

[1698] Step 3:

[1699] The device analyzes text data using generative artificial intelligence (e.g., ChatGPT or BERT) to detect fraud-related keywords and phrases. The input is text data, and the output is the detection results of fraud patterns.

[1700] Step 4:

[1701] The device uses an emotion analysis engine to analyze the user's emotional state. It evaluates the user's emotions from voice and text data and determines the level of stress and anxiety. The input is voice and text data, and the output is the evaluation result of the emotional state.

[1702] Step 5:

[1703] If a fraud pattern is detected and the user's emotional state is unstable, the device generates an alert and notifies the user. Specifically, it generates a message such as "There is a possibility of fraud. Please hang up the phone immediately" and notifies the user visually or audibly. The input is the fraud pattern detection result and the emotional state evaluation result, and the output is the alert message.

[1704] Step 6:

[1705] The terminal sends the generated alert to the server. The alert message is sent to the server via the Internet and recorded in the database. The input is the alert message, and the output is the alert sending completion status.

[1706] Step 7:

[1707] The server analyzes the received alert and notifies pre-registered contacts (family, police, etc.) via email, SMS, push notification, etc. The input is the alert message, and the output is the notification message.

[1708] Specific working example:

[1709] For example, suppose the user speaks the following prompt:

[1710] "This call is from City Hall to inform you about your refund. Please go to an ATM and follow the instructions provided."

[1711] In this case, the system detected a fraud pattern and the user's emotional state indicated high stress, so it generated an alert and notified the user, "This may be a scam. Please hang up immediately." Similar alert messages were also sent to family members and the police.

[1712] 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.

[1713] 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.

[1714] 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.

[1715] 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.

[1716] 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.

[1717] 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.

[1718] 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).

[1719] 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.

[1720] 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."

[1721] 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.

[1722] 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).

[1723] 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.

[1724] 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.

[1725] 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.

[1726] 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.

[1727] 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.

[1728] 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.

[1729] 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.

[1730] 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.

[1731] 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.

[1732] 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.

[1733] The following is further disclosed regarding the above embodiment.

[1734] (Claim 1)

[1735] means for acquiring user voice data;

[1736] A means for converting the acquired voice data into text data;

[1737] means for analyzing the text data to detect fraud patterns;

[1738] means for generating an alert and notifying a user when a fraud pattern is detected;

[1739] a means for sending the alert to a server;

[1740] a means for the server to receive the alert and notify a registered third party;

[1741] A system including:

[1742] (Claim 2)

[1743] The system of claim 1, wherein the alert notification includes user location information.

[1744] (Claim 3)

[1745] 10. The system of claim 1, wherein the system uses generative artificial intelligence to analyze text data and detect fraudulent patterns.

[1746] "Example 1"

[1747] (Claim 1)

[1748] means for acquiring user voice data;

[1749] A means for converting the acquired voice data into character string data;

[1750] means for analyzing the string data and detecting fraud patterns;

[1751] means for generating an alert and notifying a user when a fraudulent pattern is detected;

[1752] means for sending the alert to a server;

[1753] a means for the server to receive the alert and notify a registered third party;

[1754] A means for converting voice data into character string data using a voice recognition engine;

[1755] a means for analyzing string data using a generative AI model to detect fraudulent patterns;

[1756] means for notifying the user of the warning by audio or visual message;

[1757] A system including:

[1758] (Claim 2)

[1759] The system of claim 1, wherein the alert notification includes user location information.

[1760] (Claim 3)

[1761] 10. The system of claim 1, wherein the system uses a generative AI model to analyze string data and detect fraudulent patterns.

[1762] "Application Example 1"

[1763] (Claim 1)

[1764] means for acquiring user voice data;

[1765] A means for converting the acquired voice data into text data;

[1766] means for analyzing text data using automated generation techniques to detect fraud patterns;

[1767] means for generating an alert and providing audio or visual notification to the user when a fraud pattern is detected;

[1768] a means for sending the alert to a server;

[1769] A means for the server to receive the alert and notify registered third parties via email or SMS;

[1770] A system including:

[1771] (Claim 2)

[1772] The system of claim 1, wherein the alert notification includes user location information.

[1773] (Claim 3)

[1774] 10. The system of claim 1, wherein the system uses automated generation techniques to analyze text data and detect fraud patterns.

[1775] "Example 2: Combining Emotion Engines"

[1776] (Claim 1)

[1777] means for acquiring user voice data;

[1778] A means for converting the acquired voice data into text data;

[1779] a means for analyzing text data using generative artificial intelligence to detect fraud patterns;

[1780] A means for acquiring and analyzing user emotion data;

[1781] means for generating an alert and notifying the user if a fraud pattern is detected and the user's emotional state is unstable;

[1782] a means for sending the alert to a server;

[1783] a means for the server to receive the alert and notify a registered third party;

[1784] A system including:

[1785] (Claim 2)

[1786] The system of claim 1, wherein the alert notification includes user location information.

[1787] (Claim 3)

[1788] 10. The system of claim 1, wherein the emotion engine is used to analyze the user's emotion data.

[1789] "Application example 2 when combining emotion engines"

[1790] (Claim 1)

[1791] means for acquiring user voice data;

[1792] A means for converting the acquired voice data into text data;

[1793] means for analyzing the text data to detect fraud patterns;

[1794] means for analyzing the emotional state of a user;

[1795] means for generating an alert and notifying the user if a fraud pattern is detected and the user's emotional state is unstable;

[1796] a means for sending the alert to a server;

[1797] a means for the server to receive the alert and notify a registered third party;

[1798] A system including:

[1799] (Claim 2)

[1800] The system of claim 1, wherein the alert notification includes user location information.

[1801] (Claim 3)

[1802] 10. The system of claim 1, wherein the system uses generative artificial intelligence to analyze text data and detect fraudulent patterns. [Explanation of symbols]

[1803] 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. means for acquiring user voice data; A means for converting the acquired voice data into text data; means for analyzing the text data to detect fraud patterns; means for generating an alert and notifying a user when a fraud pattern is detected; a means for sending the alert to a server; a means for the server to receive the alert and notify a registered third party; A system including:

2. The system according to claim 1 , wherein the alert notification includes user location information.

3. 10. The system of claim 1, wherein generative artificial intelligence is used to analyze text data and detect fraudulent patterns.

Images