system

Abstract

The system according to this embodiment aims to prevent fraud using voice calls. [Solution] The system according to the embodiment comprises a proxy answering unit, an analysis unit, a transfer unit, a monitoring unit, and a list management unit. The proxy answering unit uses AI to answer calls on behalf of the caller. The analysis unit analyzes the content of the conversation answered by the proxy answering unit and determines the risk of fraud. The transfer unit transfers the call if the analysis unit determines that the risk of fraud is low. The monitoring unit monitors the conversation during the call and issues a warning if signs of fraud are detected. The list management unit manages whitelist, graylist, and blacklist telephone numbers.

Description

Technical Field

[0006] , , ,

[0005] , ,

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the conventional technology, there was a problem that it was difficult to prevent fraud using voice calls.

[0005] The system according to the embodiment aims to prevent fraud using voice calls.

Means for Solving the Problems

[0006] The system according to this embodiment comprises a proxy answering unit, an analysis unit, a transfer unit, a monitoring unit, and a list management unit. The proxy answering unit uses AI to answer calls on behalf of the caller. The analysis unit analyzes the content of the conversation answered by the proxy answering unit and determines the risk of fraud. The transfer unit transfers the call if the analysis unit determines that the risk of fraud is low. The monitoring unit monitors the conversation even while it is in progress and issues a warning if signs of fraud are detected. The list management unit manages telephone numbers in a whitelist, graylist, and blacklist. [Effects of the Invention]

[0007] The system according to this embodiment can prevent fraud using voice calls. [Brief explanation of the drawing]

[0008] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10]This shows an emotion map where multiple emotions are mapped. [Modes for carrying out the invention]

[0009] Hereinafter, an example of an embodiment of the system relating to the technology of this disclosure will be described with reference to the attached drawings.

[0010] First, let's explain the terminology used in the following explanation.

[0011] In the following embodiments, the signed processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Furthermore, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include CPU (Central Processing Unit), GPU (Graphics Processing Unit), GPGPU (General-Purpose computing on Graphics Processing Units), APU (Accelerated Processing Unit), or TPU (Tensor Processing Unit).

[0012] In the following embodiments, signed RAM (Random Access Memory) is a memory that temporarily stores information and is used as work memory by the processor.

[0013] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0014] In the following embodiments, the labeled communication I / F (Interface) is an interface including a communication processor, an antenna, etc. The communication I / F manages communication between a plurality of computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0015] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B". That is, "A and / or B" means that it may be only A, only B, or a combination of A and B. Also, in this specification, when expressing three or more matters connected by "and / or", the same concept as "A and / or B" is applied.

[0016] [First Embodiment] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0017] As shown in FIG. 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0018] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. 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. Also, the database 24 and the communication I / F 26 are 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).

[0019] The smart device 14 comprises a computer 36, a receiving device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The receiving device 38, output device 40, and camera 42 are also connected to the bus 52.

[0020] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, and accepts user input. The touch panel 38A accepts user input via touch by detecting contact with an object (e.g., a pen or finger). The microphone 38B accepts user input via voice by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 (see Figure 2) acquires the data indicating the user input.

[0021] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user by outputting the data in a form perceptible to the user (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0022] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0023] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0024] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0025] Storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotions using the emotion identification model 59 and perform identification processing using the user's emotions. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotions, including but not limited to these examples. Furthermore, emotion estimation and prediction also include, for example, emotion analysis.

[0026] In the smart device 14, specific processing is performed by the processor 46. The storage 50 stores a specific processing program 60. The specific processing program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the specific processing program 60 from the storage 50 and executes the read specific processing program 60 on the RAM 48. The specific processing is realized by the processor 46 operating as a control unit 46A according to the specific processing program 60 executed on the RAM 48. The smart device 14 also has a data generation model 58 and an emotion identification model 59, similar to the data generation model and emotion identification model 59, and can perform processing similar to that of the specific processing unit 290 using these models.

[0027] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device (e.g., a generation server) may have the data generation model 58. In this case, the data processing device 12 obtains processing results (such as prediction results) using the data generation model 58 by communicating with the server device having the data generation model 58. The data processing device 12 may also be a server device or a terminal device owned by a user (e.g., a mobile phone, robot, home appliance, etc.). Next, an example of processing by the data processing system 10 according to the first embodiment will be described.

[0028] (Example of form 1) The AI ​​voice telephone service according to an embodiment of the present invention is a system for preventing fraud using voice telephones. This service first has the AI ​​voice telephone service answer on behalf of the caller upon incoming calls, assess the risk of fraudulent calls, and then transfer the call. Even after the call has started, it monitors the conversation to be vigilant against and eliminate fraud. For example, when a call comes in, the AI ​​voice telephone service answers on behalf of the caller. At this time, the AI ​​conveys a message to the caller such as "Fraud prevention AI is answering on your behalf." Next, the AI ​​analyzes the content of the caller's conversation and assesses the risk of fraud. For example, it checks the caller's trustworthiness through questions such as "Don't you know your mother's birthday?" or "Are you your boss at work? Why aren't you in HR?" If the AI ​​determines that the risk of fraud is low, it transfers the call to the recipient. At this time, the AI ​​conveys a message to the caller such as "Connecting you now♪". On the other hand, if the AI ​​determines that the risk of fraud is high, it rejects the call and informs the caller that "This call may be fraudulent, so we will end the call." Furthermore, even after the call has been transferred, the AI ​​continues to monitor the conversation. For example, if typical scam phrases such as "Are you so-and-so's mother?" or "So-and-so has been in the hospital after an accident!" are detected, the AI ​​will issue a warning and terminate the call. This service manages whitelisted, graylisted, and blacklisted phone numbers and decides how to handle calls based on the list. Incoming calls from whitelisted phone numbers are forwarded to the recipient as usual. Incoming calls from graylisted phone numbers are answered by the AI, which assesses the risk of fraud. Incoming calls from blacklisted phone numbers are automatically rejected. This system prevents voice call scams and ensures user safety. In particular, it is expected to significantly reduce the risk of "kind dads and moms" and "kind grandpas and grandmas" who care about the safety and well-being of their children becoming victims of scams. Thus, the AI ​​voice call service can prevent voice call scams from occurring.

[0029] The AI ​​voice telephone service according to this embodiment comprises a proxy answering unit, an analysis unit, a transfer unit, a monitoring unit, and a list management unit. The proxy answering unit uses AI to answer calls on behalf of the caller. For example, when a call comes in, the proxy answering unit conveys a message to the caller such as "Anti-fraud AI is answering on your behalf." The proxy answering unit can also record the content of the caller's conversation and send it to the analysis unit. Furthermore, the proxy answering unit can refer to telephone numbers on a whitelist, graylist, and blacklist and make appropriate responses. For example, the proxy answering unit answers calls from telephone numbers registered on the whitelist as usual, and asks questions to determine the risk of fraud for calls from telephone numbers registered on the graylist. It automatically rejects calls from telephone numbers registered on the blacklist. The analysis unit analyzes the content of conversations answered by the proxy answering unit and determines the risk of fraud. For example, the analysis unit converts the conversation content into text data using speech recognition technology and evaluates the risk of fraud using natural language processing technology. The analysis unit can also detect specific keywords or phrases and determine the possibility of fraud. For example, the analysis unit verifies the caller's trustworthiness through questions such as, "Is this [Name]? Don't you know your mother's birthday?" or "Are you [Name]'s boss? Why not the HR department?" The call forwarding unit forwards the call if the analysis unit determines that the risk of fraud is low. For example, if the analysis unit determines that the risk of fraud is low, the call forwarding unit sends a message to the caller such as, "Connecting you now♪" and forwards the call to the recipient. The call forwarding unit can also select the call forwarding destination based on a whitelist. The monitoring unit monitors the conversation during the call and issues a warning if signs of fraud are detected. For example, if the monitoring unit detects typical fraud phrases such as, "Are you [Name]'s mother?" or "[Name] has been in an emergency hospitalization after an accident!" during the call, it can issue a warning and terminate the call. The monitoring unit can also analyze the call content in real time and detect signs of fraud. The list management unit manages the whitelist, graylist, and blacklist of phone numbers. For example, the list management unit allows users to manually update the list.Furthermore, the list management unit can automatically update the list based on the results of the proxy response unit and the analysis unit. As a result, the AI ​​voice telephone service according to this embodiment can prevent fraud using voice telephones.

[0030] The proxy answering unit uses AI to answer incoming calls. For example, when a call comes in, the proxy answering unit will send a message to the caller such as, "The fraud prevention AI is answering your call." Specifically, the AI ​​can use speech synthesis technology to deliver the message to the caller in a natural voice. The proxy answering unit can also record the caller's conversation and send it to the analysis unit. The recorded audio data is saved as an audio file and uploaded to a cloud server for the analysis unit to access. Furthermore, the proxy answering unit can refer to whitelisted, graylisted, and blacklisted phone numbers and provide appropriate responses. For example, the proxy answering unit will answer calls from whitelisted phone numbers as usual, and for calls from graylisted phone numbers, it will ask questions to determine the risk of fraud. Specifically, the AI ​​will ask the caller questions such as, "May I have your name and the purpose of your call?" and send the answer to the analysis unit. Calls from blacklisted phone numbers will be automatically rejected. Possible rejection messages include, "We do not accept calls from this phone number." This allows the proxy response unit to quickly assess the reliability of the caller and take appropriate action.

[0031] The analysis unit analyzes the conversation content that has been answered by the proxy response unit to determine the risk of fraud. For example, the analysis unit uses speech recognition technology to convert the conversation content into text data and natural language processing technology to evaluate the risk of fraud. Specifically, speech recognition technology converts the caller's voice into text with high accuracy, and natural language processing technology extracts specific keywords and phrases from the text data. For example, the reliability of the caller is confirmed through questions such as, "Hey, [Name]? Don't you know your mother's birthday?" or "Are you [Name]'s boss? Why aren't you in the HR department?" Furthermore, the analysis unit can also compare the conversation content with a database of past fraud cases to check for similar patterns. This allows the analysis unit to analyze the caller's conversation content from multiple angles and determine the risk of fraud with high accuracy. The analysis unit also provides the analysis results to the transmission unit and monitoring unit in real time to support a rapid response.

[0032] The call forwarding unit forwards calls when the analysis unit determines that the risk of fraud is low. For example, if the analysis unit determines that the risk of fraud is low, the call forwarding unit will send a message such as "Connecting you now♪" to the caller and forward the call to the recipient. Specifically, the call forwarding unit receives a signal from the analysis unit and automatically connects the call to the recipient's phone. The call forwarding unit can also select the call forwarding destination based on a whitelist. For example, it can select the most appropriate recipient from multiple phone numbers registered in the whitelist and forward the call. This allows the call forwarding unit to forward only highly reliable calls to recipients, reducing the risk of fraud. Furthermore, the call forwarding unit can monitor the call forwarding status in real time and respond immediately if an anomaly is detected.

[0033] The monitoring unit monitors the conversation even during a call and issues a warning if signs of fraud are detected. For example, if the monitoring unit detects typical fraud phrases such as "Are you so-and-so's mother?" or "So-and-so has been in the hospital after an accident!" during a call, it can issue a warning and terminate the call. Specifically, the monitoring unit combines speech recognition technology and natural language processing technology to analyze the call content in real time and detect signs of fraud. The monitoring unit can also analyze the call content in real time and detect signs of fraud. For example, if a specific keyword or phrase is detected during a call, the monitoring unit will immediately issue a warning and terminate the call. Furthermore, the monitoring unit can record the call content so that the analysis unit can later analyze it in detail. This allows the monitoring unit to monitor the risk of fraud during a call in real time and respond quickly.

[0034] The list management unit manages whitelisted, graylisted, and blacklisted phone numbers. For example, users can manually update the list. Specifically, users can add, delete, and edit phone numbers through a dedicated application or web interface. The list management unit can also automatically update the list based on the results of the proxy response unit and the analysis unit. For example, if the analysis unit determines that a particular phone number is at high risk of fraud, it can automatically add that phone number to the blacklist. Conversely, highly reliable phone numbers are added to the whitelist. This allows the list management unit to always manage the list based on the latest information, improving the overall reliability of the system. Furthermore, the list management unit can record the update history of the list for later review. This allows the list management unit to track user and system operation history and manage the list more effectively.

[0035] The questioning unit can verify the sender's credibility through specific questions. For example, the questioning unit can ask the sender questions such as, "Do you know your mother's birthday?" or "Are you your boss at work? Why aren't you in HR?" to verify the sender's credibility. The questioning unit can also analyze the sender's answers and evaluate their consistency and reliability. For example, the questioning unit can check whether the sender's answers match past data to assess reliability. Furthermore, the questioning unit can ask additional questions based on the sender's answers. For example, if the questioning unit has doubts about the sender's answers, it can ask additional questions to verify reliability. In this way, the questioning unit can reduce the risk of fraud by verifying the sender's credibility.

[0036] The rejection unit can reject calls if it determines that there is a high risk of fraud. For example, if the analysis unit determines that there is a high risk of fraud, the rejection unit will inform the caller, "This call may be fraudulent, so we will end it," and reject the call. The rejection unit can also reject calls if it determines that the caller is unreliable. For example, the rejection unit will reject a call if the caller's responses are inconsistent or if certain keywords or phrases are detected. Furthermore, the rejection unit can automatically reject incoming calls from phone numbers on a blacklist. For example, if an incoming call comes from a phone number on the blacklist, the rejection unit will automatically reject the call. In this way, the rejection unit can protect users by rejecting calls that have a high risk of fraud.

[0037] The warning unit can issue a warning if it detects signs of fraud during a call. For example, if it detects typical fraudulent phrases such as "Are you so-and-so's mother?" or "So-and-so has been in the hospital after an accident!" during a call, it can issue a warning and terminate the call. The warning unit can also analyze the call content in real time to detect signs of fraud. For example, it can detect specific keywords or phrases to determine the possibility of fraud. Furthermore, the warning unit can analyze changes in the caller's voice tone and speed to detect signs of fraud. For example, it will issue a warning if the caller's voice tone changes unnaturally or if the voice speed is abnormally fast. In this way, the warning unit can prevent fraud by detecting signs of fraud during a call and issuing a warning.

[0038] The list update unit can update the list based on the proxy response results. For example, the list update unit automatically updates the whitelist, graylist, and blacklist based on the results of the proxy response unit and the analysis unit. The list update unit also allows users to manually update the list. For example, the list update unit allows users to add specific phone numbers to the whitelist or move them to the blacklist. Furthermore, the list update unit can adjust the frequency and timing of list updates. For example, the list update unit can set a schedule for periodically updating the list. This allows the list update unit to improve the accuracy of the system by updating the list based on the proxy response results.

[0039] The caller ID unit can analyze the characteristics of the caller's voice upon receiving an incoming call and evaluate its reliability by comparing it with past data. For example, the caller ID unit can analyze the tone and speed of the caller's voice and compare it with patterns of past fraudulent calls. The caller ID unit can also compare the characteristics of the caller's voice with past call history and evaluate the degree of similarity. Furthermore, the caller ID unit can compare the characteristics of the caller's voice with fraud characteristics learned by the AI ​​and evaluate its reliability. In this way, the caller ID unit can evaluate reliability by analyzing the characteristics of the caller's voice. Some or all of the above processing in the caller ID unit may be performed using AI, for example, or without AI. For example, the caller ID unit can input the caller's voice characteristic data into a generating AI and have the generating AI perform the reliability evaluation.

[0040] The proxy response unit can analyze the caller's background noise when responding on behalf of the caller and determine the possibility of fraud. For example, the proxy response unit evaluates the possibility of fraud if the caller's background noise contains certain noises. The proxy response unit can also determine the possibility of fraud if the caller's background noise contains the voices of other people. Furthermore, the proxy response unit can also evaluate the possibility of fraud if the caller's background noise contains certain environmental sounds (e.g., call center sounds). In this way, the proxy response unit can determine the possibility of fraud by analyzing the caller's background noise. Some or all of the above processing in the proxy response unit may be performed using AI, for example, or without AI. For example, the proxy response unit can input the caller's background noise data into a generating AI and have the generating AI perform the determination of the possibility of fraud.

[0041] The proxy answering unit can adjust the response content by considering the caller's geographical location information when answering on behalf of the caller. For example, if the caller is coming from a specific region, the proxy answering unit will provide information related to that region. The proxy answering unit can also be wary of incoming calls from regions with a high risk of fraud based on the caller's geographical location information. Furthermore, the proxy answering unit can adjust the response content by considering region-specific fraud patterns based on the caller's geographical location information. As a result, the proxy answering unit can provide a more appropriate response by adjusting the response content by considering the caller's geographical location information. Some or all of the above processing in the proxy answering unit may be performed using AI, for example, or without AI. For example, the proxy answering unit can input the caller's geographical location information data into a generating AI and have the generating AI perform the adjustment of the response content.

[0042] The proxy answering unit can optimize the response content by referring to the caller's past call history when answering on behalf of the caller. For example, if the caller has made a call in the past that is suspected of being fraudulent, the proxy answering unit will increase its vigilance. The proxy answering unit can also provide a highly reliable response content based on the caller's past call history. Furthermore, the proxy answering unit can analyze the caller's past call history, assess the risk of fraud, and adjust the response content accordingly. In this way, the proxy answering unit can optimize the response content by referring to the caller's past call history. Some or all of the above processing in the proxy answering unit may be performed using AI, for example, or not using AI. For example, the proxy answering unit can input the caller's past call history data into a generating AI and have the generating AI perform the optimization of the response content.

[0043] The analysis unit can analyze the tone and speed of the caller's voice during analysis and assess the likelihood of fraud. For example, the analysis unit assesses the likelihood of fraud if the tone of the caller's voice changes unnaturally. The analysis unit can also determine the likelihood of fraud if the speed of the caller's voice is abnormally fast. Furthermore, the analysis unit can also assess the likelihood of fraud by comparing the tone and speed of the caller's voice with patterns of past fraudulent calls. In this way, the analysis unit can assess the likelihood of fraud by analyzing the tone and speed of the caller's voice. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can input the tone and speed data of the caller's voice into a generating AI and have the generating AI perform the fraud likelihood assessment.

[0044] The analysis unit can detect specific phrases and expressions used by the caller during analysis and determine the risk of fraud. For example, the analysis unit evaluates the risk of fraud if the caller uses a specific fraudulent phrase. The analysis unit can also determine the risk of fraud if the caller's expressions match those of past fraudulent calls. Furthermore, the analysis unit can have an AI learn specific phrases used by the caller and evaluate the risk of fraud. This allows the analysis unit to determine the risk of fraud by detecting specific phrases and expressions used by the caller. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can input caller phrase and expression data into a generating AI and have the generating AI perform the fraud risk determination.

[0045] The analysis unit can improve the accuracy of the analysis by referring to the caller's past call history during the analysis. For example, the analysis unit improves the accuracy of the analysis based on the caller's past call history. The analysis unit can also analyze the caller's past call history and assess the risk of fraud. Furthermore, the analysis unit can correct the analysis results by referring to the caller's past call history. In this way, the analysis unit can improve the accuracy of the analysis by referring to the caller's past call history. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without using AI. For example, the analysis unit can input the caller's past call history data into a generating AI and have the generating AI perform the analysis accuracy improvement.

[0046] The analysis unit can correct the analysis results by considering the geographical location information of the caller during the analysis. For example, the analysis unit can assess the risk of fraud based on the geographical location information of the caller. The analysis unit can also correct the analysis results by considering the geographical location information of the caller. Furthermore, the analysis unit can adjust the analysis results by considering region-specific fraud patterns based on the geographical location information of the caller. This allows the analysis unit to perform more accurate analysis by correcting the analysis results by considering the geographical location information of the caller. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without using AI. For example, the analysis unit can input the geographical location information data of the caller into a generating AI and have the generating AI perform the correction of the analysis results.

[0047] The forwarding unit can re-evaluate the sender's trustworthiness during forwarding and determine whether or not to forward the message. For example, the forwarding unit can re-evaluate the sender's trustworthiness before forwarding and determine the risk of fraud. The forwarding unit can also re-evaluate the sender's trustworthiness during forwarding and determine whether or not to forward the message. Furthermore, the forwarding unit can re-evaluate the sender's trustworthiness before forwarding and refuse to forward the message if necessary. This allows the forwarding unit to perform more reliable forwarding by re-evaluating the sender's trustworthiness during forwarding. Some or all of the above processing in the forwarding unit may be performed using AI, for example, or without AI. For example, the forwarding unit can input sender trustworthiness data into a generating AI and have the generating AI perform the trustworthiness re-evaluation.

[0048] The transfer unit can optimize the transfer by considering the recipient's current situation at the time of transfer. For example, if the recipient is busy, the transfer unit can delay the transfer or transfer it at a different time. The transfer unit can also consider the recipient's current location and transfer at an appropriate time. Furthermore, the transfer unit can transfer at the optimal time based on the recipient's schedule information. In this way, the transfer unit can optimize the transfer by considering the recipient's current situation, enabling transfers at a more appropriate time. Some or all of the above processing in the transfer unit may be performed using AI, for example, or without AI. For example, the transfer unit can input the recipient's schedule information data into a generating AI and have the generating AI perform the transfer optimization.

[0049] The transfer unit can determine whether or not to transfer a call by referring to the caller's past call history at the time of transfer. For example, the transfer unit can determine whether or not to transfer a call based on the caller's past call history. The transfer unit can also analyze the caller's past call history, assess the risk of fraud, and then decide whether or not to transfer a call. Furthermore, the transfer unit can perform a highly reliable transfer by referring to the caller's past call history. Thus, the transfer unit can determine whether or not to transfer a call by referring to the caller's past call history. Some or all of the above processing in the transfer unit may be performed using AI, for example, or not using AI. For example, the transfer unit can input the caller's past call history data into a generating AI and have the generating AI perform the determination of whether or not to transfer a call.

[0050] The transfer unit can adjust the timing of the transfer by considering the recipient's schedule information during the transfer process. For example, the transfer unit can perform the transfer at the optimal timing based on the recipient's schedule information. The transfer unit can also adjust the timing of the transfer by considering the recipient's schedule information. Furthermore, the transfer unit can refer to the recipient's schedule information and perform the transfer at an appropriate time. As a result, the transfer unit can perform the transfer at a more appropriate timing by adjusting the timing of the transfer by considering the recipient's schedule information. Some or all of the above processing in the transfer unit may be performed using AI, for example, or without AI. For example, the transfer unit can input the recipient's schedule information data into a generating AI and have the generating AI perform the adjustment of the transfer timing.

[0051] The monitoring unit can analyze the caller's voice tone and speed in real time during a call and detect signs of fraud. For example, the monitoring unit can detect signs of fraud if the caller's voice tone changes unnaturally. The monitoring unit can also determine if signs of fraud exist if the caller's voice speed is abnormally fast. Furthermore, the monitoring unit can also detect signs of fraud by comparing the caller's voice tone and speed with patterns from past fraudulent calls. In this way, the monitoring unit can detect signs of fraud by analyzing the caller's voice tone and speed in real time. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input the caller's voice tone and speed data into a generating AI and have the generating AI perform the detection of signs of fraud.

[0052] The monitoring unit can detect specific phrases and expressions in real time during a call and assess the risk of fraud. For example, the monitoring unit assesses the risk of fraud if the caller uses a specific fraudulent phrase. The monitoring unit can also determine the risk of fraud if the caller's expressions match those of past fraudulent calls. Furthermore, the monitoring unit can have an AI learn specific phrases used by the caller and assess the risk of fraud. This allows the monitoring unit to assess the risk of fraud by detecting specific phrases and expressions in real time. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input caller phrase and expression data into a generating AI and have the generating AI perform the fraud risk assessment.

[0053] The monitoring unit can improve the accuracy of monitoring by referring to the caller's past call history during a call. For example, the monitoring unit improves the accuracy of monitoring based on the caller's past call history. The monitoring unit can also analyze the caller's past call history and assess the risk of fraud. Furthermore, the monitoring unit can correct the monitoring results by referring to the caller's past call history. In this way, the monitoring unit can improve the accuracy of monitoring by referring to the caller's past call history. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input the caller's past call history data into a generating AI and have the generating AI perform the improvement of monitoring accuracy.

[0054] The monitoring unit can correct the monitoring results during a call by taking into account the caller's geographical location information. For example, the monitoring unit can assess the risk of fraud based on the caller's geographical location information. The monitoring unit can also correct the monitoring results by taking into account the caller's geographical location information. Furthermore, the monitoring unit can adjust the monitoring results by taking into account region-specific fraud patterns based on the caller's geographical location information. This allows the monitoring unit to perform more accurate monitoring by correcting the monitoring results by taking into account the caller's geographical location information. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without using AI. For example, the monitoring unit can input the caller's geographical location information data into a generating AI and have the generating AI perform the correction of the monitoring results.

[0055] The list management unit can improve the accuracy of the list by referring to the caller's past call history during list management. For example, the list management unit improves the accuracy of the list based on the caller's past call history. The list management unit can also analyze the caller's past call history and assess the risk of fraud. Furthermore, the list management unit can correct the contents of the list by referring to the caller's past call history. In this way, the list management unit can improve the accuracy of the list by referring to the caller's past call history. Some or all of the above processing in the list management unit may be performed using AI, for example, or without using AI. For example, the list management unit can input the caller's past call history data into a generating AI and have the generating AI perform the list accuracy improvement.

[0056] The list management unit can optimize the list content by considering the geographical location information of the sender during list management. For example, the list management unit optimizes the list content based on the geographical location information of the sender. The list management unit can also correct the list content by considering the geographical location information of the sender. Furthermore, the list management unit can adjust the list content by considering region-specific fraud patterns based on the geographical location information of the sender. As a result, the list management unit can perform more accurate list management by optimizing the list content by considering the geographical location information of the sender. Some or all of the above processing in the list management unit may be performed using AI, for example, or without using AI. For example, the list management unit can input geographical location data of the sender into a generating AI and have the generating AI perform the optimization of the list content.

[0057] The list management unit can improve the accuracy of the list by analyzing the characteristics of the caller's voice during list management. For example, the list management unit can improve the accuracy of the list by analyzing the tone and speed of the caller's voice. The list management unit can also correct the list contents by comparing the characteristics of the caller's voice with past call history. Furthermore, the list management unit can improve the accuracy of the list by having an AI learn the characteristics of the caller's voice. In this way, the list management unit can improve the accuracy of the list by analyzing the characteristics of the caller's voice. Some or all of the above processing in the list management unit may be performed using an AI, for example, or without an AI. For example, the list management unit can input caller voice characteristic data into a generating AI and have the generating AI perform list accuracy improvement.

[0058] The list management unit can detect specific phrases and expressions used by callers during list management and update the list content. For example, if a caller uses a specific fraudulent phrase, the list management unit will update the list content. The list management unit can also correct the list content if the caller's expressions match those of past fraudulent calls. Furthermore, the list management unit can update the list content by having an AI learn specific phrases used by callers. This allows the list management unit to update the list content by detecting specific phrases and expressions used by callers. Some or all of the above processing in the list management unit may be performed using an AI, for example, or without an AI. For example, the list management unit can input caller phrase and expression data into a generating AI and have the generating AI perform the update of the list content.

[0059] The questioning unit can analyze the caller's tone and speed of voice when a question is asked and evaluate its reliability. For example, the questioning unit evaluates reliability if the caller's tone of voice changes unnaturally. It can also determine reliability if the caller's speed of voice is abnormally fast. Furthermore, the questioning unit can compare the caller's tone and speed of voice with patterns of past fraudulent calls and evaluate reliability. In this way, the questioning unit can evaluate reliability by analyzing the caller's tone and speed of voice. Some or all of the above processing in the questioning unit may be performed using AI, for example, or without AI. For example, the questioning unit can input the caller's tone and speed data into a generating AI and have the generating AI perform the reliability evaluation.

[0060] The questioning unit can detect specific phrases and expressions used by the caller when asking a question and determine its reliability. For example, the questioning unit evaluates reliability if the caller uses a specific fraudulent phrase. The questioning unit can also determine reliability if the caller's expression matches that of past fraudulent calls. Furthermore, the questioning unit can have an AI learn specific phrases used by the caller and evaluate reliability based on that learning. This allows the questioning unit to determine reliability by detecting specific phrases and expressions used by the caller. Some or all of the above processing in the questioning unit may be performed using AI, for example, or without AI. For example, the questioning unit can input caller phrase and expression data into a generating AI and have the generating AI perform the reliability determination.

[0061] The questioning unit can optimize the question content by referring to the caller's past call history when asking a question. For example, the questioning unit can provide highly reliable question content based on the caller's past call history. The questioning unit can also analyze the caller's past call history, assess the risk of fraud, and adjust the question content accordingly. Furthermore, the questioning unit can provide optimal question content by referring to the caller's past call history. In this way, the questioning unit can optimize the question content by referring to the caller's past call history. Some or all of the above processing in the questioning unit may be performed using AI, for example, or not using AI. For example, the questioning unit can input the caller's past call history data into a generating AI and have the generating AI perform the optimization of the question content.

[0062] The questioning unit can adjust the content of questions by considering the caller's geographical location information. For example, the questioning unit can adjust the content of questions by considering region-specific fraud patterns based on the caller's geographical location information. The questioning unit can also provide highly reliable questions by considering the caller's geographical location information. Furthermore, the questioning unit can provide optimal questions based on the caller's geographical location information. In this way, the questioning unit can ask more appropriate questions by adjusting the content of questions by considering the caller's geographical location information. Some or all of the above processing in the questioning unit may be performed using AI, for example, or without AI. For example, the questioning unit can input the caller's geographical location information data into a generating AI and have the generating AI perform the adjustment of the question content.

[0063] The rejection unit can detect specific phrases or expressions used by the caller when rejecting a call, thereby strengthening the basis for rejection. For example, the rejection unit strengthens the basis for rejection if the caller uses specific fraudulent phrases. The rejection unit can also clarify the basis for rejection if the caller's expressions match those of past fraudulent calls. Furthermore, the rejection unit can have an AI learn specific phrases used by the caller, thereby strengthening the basis for rejection. This allows the rejection unit to strengthen the basis for rejection by detecting specific phrases or expressions used by the caller. Some or all of the above processing in the rejection unit may be performed using AI, for example, or without AI. For example, the rejection unit can input caller phrase and expression data into a generating AI and have the generating AI perform the strengthening of the basis for rejection.

[0064] The rejection unit can determine whether to reject a call by referring to the caller's past call history. For example, the rejection unit can determine whether to reject a call based on the caller's past call history. The rejection unit can also analyze the caller's past call history, assess the risk of fraud, and then decide whether to reject the call. Furthermore, the rejection unit can make a highly reliable rejection by referring to the caller's past call history. Thus, the rejection unit can determine whether to reject a call by referring to the caller's past call history. Some or all of the above processing in the rejection unit may be performed using AI, for example, or without AI. For example, the rejection unit can input the caller's past call history data into a generating AI and have the generating AI perform the decision on whether to reject the call.

[0065] The rejection unit can adjust the reason for rejection by considering the caller's geographical location information at the time of rejection. For example, the rejection unit can assess the risk of fraud based on the caller's geographical location information. The rejection unit can also adjust the reason for rejection by considering the caller's geographical location information. Furthermore, the rejection unit can adjust the reason for rejection by considering region-specific fraud patterns based on the caller's geographical location information. This allows the rejection unit to perform more accurate rejections by adjusting the reason for rejection by considering the caller's geographical location information. Some or all of the above processing in the rejection unit may be performed using AI, for example, or without using AI. For example, the rejection unit can input the caller's geographical location information data into a generating AI and have the generating AI perform the adjustment of the reason for rejection.

[0066] The warning unit can analyze the caller's voice tone and speed at the time of a warning and optimize the timing of the warning. For example, if the caller's voice tone changes unnaturally, the warning unit will optimize the timing of the warning. The warning unit can also determine the timing of the warning if the caller's voice speed is abnormally fast. Furthermore, the warning unit can compare the caller's voice tone and speed with patterns of past fraudulent calls and optimize the timing of the warning. In this way, the warning unit can optimize the timing of the warning by analyzing the caller's voice tone and speed. Some or all of the above processing in the warning unit may be performed using AI, for example, or without AI. For example, the warning unit can input the caller's voice tone and speed data into a generating AI and have the generating AI perform the optimization of the warning timing.

[0067] The warning unit can detect specific phrases and expressions used by the caller when issuing a warning, thereby strengthening the basis for the warning. For example, the warning unit strengthens the basis for the warning if the caller uses a specific fraudulent phrase. The warning unit can also clarify the basis for the warning if the caller's expression matches that of past fraudulent calls. Furthermore, the warning unit can have an AI learn specific phrases used by the caller, thereby strengthening the basis for the warning. This allows the warning unit to strengthen the basis for the warning by detecting specific phrases and expressions used by the caller. Some or all of the above processing in the warning unit may be performed using AI, for example, or without AI. For example, the warning unit can input caller phrase and expression data into a generating AI and have the generating AI perform the strengthening of the basis for the warning.

[0068] The warning unit can optimize the content of a warning by referring to the caller's past call history at the time of the warning. For example, the warning unit optimizes the content of a warning based on the caller's past call history. The warning unit can also analyze the caller's past call history, assess the risk of fraud, and adjust the content of the warning. Furthermore, the warning unit can provide the most appropriate warning content by referring to the caller's past call history. In this way, the warning unit can optimize the content of a warning by referring to the caller's past call history. Some or all of the above processing in the warning unit may be performed using AI, for example, or without AI. For example, the warning unit can input the caller's past call history data into a generating AI and have the generating AI perform the optimization of the warning content.

[0069] The warning unit can adjust the content of a warning by considering the caller's geographical location information at the time of the warning. For example, the warning unit can adjust the warning content by considering region-specific fraud patterns based on the caller's geographical location information. The warning unit can also provide highly reliable warning content by considering the caller's geographical location information. Furthermore, the warning unit can provide optimal warning content based on the caller's geographical location information. As a result, the warning unit can provide more accurate warnings by adjusting the content of the warning by considering the caller's geographical location information. Some or all of the above processing in the warning unit may be performed using AI, for example, or without AI. For example, the warning unit can input the caller's geographical location information data into a generating AI and have the generating AI perform the correction of the warning content.

[0070] The list update unit can improve the accuracy of the list by analyzing the tone and speed of the caller's voice when updating the list. For example, the list update unit can analyze the tone and speed of the caller's voice to improve the accuracy of the list. The list update unit can also correct the list content by comparing the characteristics of the caller's voice with past call history. Furthermore, the list update unit can improve the accuracy of the list by having the AI ​​learn the characteristics of the caller's voice. In this way, the list update unit can improve the accuracy of the list by analyzing the tone and speed of the caller's voice. Some or all of the above processing in the list update unit may be performed using AI, for example, or without using AI. For example, the list update unit can input caller voice characteristic data into a generating AI and have the generating AI perform the list accuracy improvement.

[0071] The list update unit can detect specific phrases or expressions used by callers when updating the list and update the list content. For example, if a caller uses a specific fraudulent phrase, the list update unit will update the list content. The list update unit can also correct the list content if the caller's expressions match those of past fraudulent calls. Furthermore, the list update unit can have an AI learn specific phrases used by callers and update the list content accordingly. This allows the list update unit to update the list content by detecting specific phrases or expressions used by callers. Some or all of the above processing in the list update unit may be performed using an AI, for example, or without an AI. For example, the list update unit can input caller phrase and expression data into a generating AI and have the generating AI perform the list content update.

[0072] The list update unit can update the list content by referring to the caller's past call history when updating the list. For example, the list update unit updates the list content based on the caller's past call history. The list update unit can also analyze the caller's past call history, assess the risk of fraud, and adjust the list content accordingly. Furthermore, the list update unit can refer to the caller's past call history to provide the most optimal list content. Thus, the list update unit can update the list content by referring to the caller's past call history. Some or all of the above processing in the list update unit may be performed using AI, for example, or without AI. For example, the list update unit can input the caller's past call history data into a generating AI and have the generating AI perform the list content update.

[0073] The list update unit can optimize the list content by considering the geographical location information of the sender when updating the list. For example, the list update unit adjusts the list content by considering region-specific fraud patterns based on the geographical location information of the sender. The list update unit can also provide highly reliable list content by considering the geographical location information of the sender. Furthermore, the list update unit can provide optimal list content based on the geographical location information of the sender. As a result, the list update unit can achieve more accurate list management by optimizing the list content by considering the geographical location information of the sender. Some or all of the above processing in the list update unit may be performed using AI, for example, or without using AI. For example, the list update unit can input the geographical location information data of the sender into a generating AI and have the generating AI perform the optimization of the list content.

[0074] The system according to the embodiment is not limited to the example described above, and various modifications are possible, for example, as follows.

[0075] The caller ID unit can analyze the characteristics of the caller's voice upon receiving an incoming call and evaluate its reliability by comparing it with past data. For example, the caller ID unit can analyze the tone and speed of the caller's voice and compare it with patterns of past fraudulent calls. The caller ID unit can also compare the characteristics of the caller's voice with past call history and evaluate the degree of similarity. Furthermore, the caller ID unit can compare the characteristics of the caller's voice with fraud characteristics learned by the AI ​​and evaluate its reliability. In this way, the caller ID unit can evaluate reliability by analyzing the characteristics of the caller's voice. Some or all of the above processing in the caller ID unit may be performed using AI, for example, or without AI. For example, the caller ID unit can input the caller's voice characteristic data into a generating AI and have the generating AI perform the reliability evaluation.

[0076] The proxy response unit can analyze the caller's background noise when responding on behalf of the caller and determine the possibility of fraud. For example, the proxy response unit evaluates the possibility of fraud if the caller's background noise contains certain noises. The proxy response unit can also determine the possibility of fraud if the caller's background noise contains the voices of other people. Furthermore, the proxy response unit can also evaluate the possibility of fraud if the caller's background noise contains certain environmental sounds (e.g., call center sounds). In this way, the proxy response unit can determine the possibility of fraud by analyzing the caller's background noise. Some or all of the above processing in the proxy response unit may be performed using AI, for example, or without AI. For example, the proxy response unit can input the caller's background noise data into a generating AI and have the generating AI perform the determination of the possibility of fraud.

[0077] The forwarding unit can re-evaluate the sender's trustworthiness during forwarding and determine whether or not to forward the message. For example, the forwarding unit can re-evaluate the sender's trustworthiness before forwarding and determine the risk of fraud. The forwarding unit can also re-evaluate the sender's trustworthiness during forwarding and determine whether or not to forward the message. Furthermore, the forwarding unit can re-evaluate the sender's trustworthiness before forwarding and refuse to forward the message if necessary. This allows the forwarding unit to perform more reliable forwarding by re-evaluating the sender's trustworthiness during forwarding. Some or all of the above processing in the forwarding unit may be performed using AI, for example, or without AI. For example, the forwarding unit can input sender trustworthiness data into a generating AI and have the generating AI perform the trustworthiness re-evaluation.

[0078] The monitoring unit can analyze the caller's voice tone and speed in real time during a call and detect signs of fraud. For example, the monitoring unit can detect signs of fraud if the caller's voice tone changes unnaturally. The monitoring unit can also determine if signs of fraud exist if the caller's voice speed is abnormally fast. Furthermore, the monitoring unit can also detect signs of fraud by comparing the caller's voice tone and speed with patterns from past fraudulent calls. In this way, the monitoring unit can detect signs of fraud by analyzing the caller's voice tone and speed in real time. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input the caller's voice tone and speed data into a generating AI and have the generating AI perform the detection of signs of fraud.

[0079] The list management unit can improve the accuracy of the list by referring to the caller's past call history during list management. For example, the list management unit improves the accuracy of the list based on the caller's past call history. The list management unit can also analyze the caller's past call history and assess the risk of fraud. Furthermore, the list management unit can correct the contents of the list by referring to the caller's past call history. In this way, the list management unit can improve the accuracy of the list by referring to the caller's past call history. Some or all of the above processing in the list management unit may be performed using AI, for example, or without using AI. For example, the list management unit can input the caller's past call history data into a generating AI and have the generating AI perform the list accuracy improvement.

[0080] The following briefly describes the processing flow for example form 1.

[0081] Step 1: The proxy answering unit uses AI to answer incoming calls. For example, when a call comes in, it sends a message to the caller such as, "The fraud prevention AI is answering your call." The proxy answering unit can also record the caller's conversation and send it to the analysis unit. Furthermore, the proxy answering unit can refer to whitelisted, graylisted, and blacklisted phone numbers and respond appropriately. For example, it will answer calls from whitelisted phone numbers as usual, ask questions to determine the risk of fraud for calls from graylisted phone numbers, and automatically reject calls from blacklisted phone numbers. Step 2: The analysis unit analyzes the conversation content that was answered by the proxy response unit and determines the risk of fraud. For example, it uses speech recognition technology to convert the conversation content into text data and natural language processing technology to evaluate the risk of fraud. It can also detect specific keywords or phrases to determine the possibility of fraud. For example, it can verify the trustworthiness of the caller through questions such as, "Hey, [Name]? Don't you know your mother's birthday?" or "Are you [Name]'s boss? Why aren't you in the HR department?" Step 3: The forwarding unit forwards the call if the analysis unit determines that the risk of fraud is low. For example, if the analysis unit determines that the risk of fraud is low, it will send a message such as "Connecting you now♪" to the caller and forward the call to the recipient. It is also possible to select the call forwarding destination based on a whitelist. Step 4: The monitoring unit monitors the conversation even while a call is in progress and issues a warning if signs of fraud are detected. For example, if typical fraud phrases such as "Are you so-and-so's mother?" or "So-and-so has been in the hospital after an accident!" are detected during a call, a warning will be issued and the call can be terminated. It can also analyze the call content in real time to detect signs of fraud. Step 5: The list management unit manages the whitelist, graylist, and blacklist of phone numbers. For example, users can manually update the list. Alternatively, the list can be automatically updated based on the results of the proxy answering unit and the analysis unit.

[0082] (Example of form 2) The AI ​​voice telephone service according to an embodiment of the present invention is a system for preventing fraud using voice telephones. This service first has the AI ​​voice telephone service answer on behalf of the caller upon incoming calls, assess the risk of fraudulent calls, and then transfer the call. Even after the call has started, it monitors the conversation to be vigilant against and eliminate fraud. For example, when a call comes in, the AI ​​voice telephone service answers on behalf of the caller. At this time, the AI ​​conveys a message to the caller such as "Fraud prevention AI is answering on your behalf." Next, the AI ​​analyzes the content of the caller's conversation and assesses the risk of fraud. For example, it checks the caller's trustworthiness through questions such as "Don't you know your mother's birthday?" or "Are you your boss at work? Why aren't you in HR?" If the AI ​​determines that the risk of fraud is low, it transfers the call to the recipient. At this time, the AI ​​conveys a message to the caller such as "Connecting you now♪". On the other hand, if the AI ​​determines that the risk of fraud is high, it rejects the call and informs the caller that "This call may be fraudulent, so we will end the call." Furthermore, even after the call has been transferred, the AI ​​continues to monitor the conversation. For example, if typical scam phrases such as "Are you so-and-so's mother?" or "So-and-so has been in the hospital after an accident!" are detected, the AI ​​will issue a warning and terminate the call. This service manages whitelisted, graylisted, and blacklisted phone numbers and decides how to handle calls based on the list. Incoming calls from whitelisted phone numbers are forwarded to the recipient as usual. Incoming calls from graylisted phone numbers are answered by the AI, which assesses the risk of fraud. Incoming calls from blacklisted phone numbers are automatically rejected. This system prevents voice call scams and ensures user safety. In particular, it is expected to significantly reduce the risk of "kind dads and moms" and "kind grandpas and grandmas" who care about the safety and well-being of their children becoming victims of scams. Thus, the AI ​​voice call service can prevent voice call scams from occurring.

[0083] The AI ​​voice telephone service according to this embodiment comprises a proxy answering unit, an analysis unit, a transfer unit, a monitoring unit, and a list management unit. The proxy answering unit uses AI to answer calls on behalf of the caller. For example, when a call comes in, the proxy answering unit conveys a message to the caller such as "Anti-fraud AI is answering on your behalf." The proxy answering unit can also record the content of the caller's conversation and send it to the analysis unit. Furthermore, the proxy answering unit can refer to telephone numbers on a whitelist, graylist, and blacklist and make appropriate responses. For example, the proxy answering unit answers calls from telephone numbers registered on the whitelist as usual, and asks questions to determine the risk of fraud for calls from telephone numbers registered on the graylist. It automatically rejects calls from telephone numbers registered on the blacklist. The analysis unit analyzes the content of conversations answered by the proxy answering unit and determines the risk of fraud. For example, the analysis unit converts the conversation content into text data using speech recognition technology and evaluates the risk of fraud using natural language processing technology. The analysis unit can also detect specific keywords or phrases and determine the possibility of fraud. For example, the analysis unit verifies the caller's trustworthiness through questions such as, "Is this [Name]? Don't you know your mother's birthday?" or "Are you [Name]'s boss? Why not the HR department?" The call forwarding unit forwards the call if the analysis unit determines that the risk of fraud is low. For example, if the analysis unit determines that the risk of fraud is low, the call forwarding unit sends a message to the caller such as, "Connecting you now♪" and forwards the call to the recipient. The call forwarding unit can also select the call forwarding destination based on a whitelist. The monitoring unit monitors the conversation during the call and issues a warning if signs of fraud are detected. For example, if the monitoring unit detects typical fraud phrases such as, "Are you [Name]'s mother?" or "[Name] has been in an emergency hospitalization after an accident!" during the call, it can issue a warning and terminate the call. The monitoring unit can also analyze the call content in real time and detect signs of fraud. The list management unit manages the whitelist, graylist, and blacklist of phone numbers. For example, the list management unit allows users to manually update the list.Furthermore, the list management unit can automatically update the list based on the results of the proxy response unit and the analysis unit. As a result, the AI ​​voice telephone service according to this embodiment can prevent fraud using voice telephones.

[0084] The proxy answering unit uses AI to answer incoming calls. For example, when a call comes in, the proxy answering unit will send a message to the caller such as, "The fraud prevention AI is answering your call." Specifically, the AI ​​can use speech synthesis technology to deliver the message to the caller in a natural voice. The proxy answering unit can also record the caller's conversation and send it to the analysis unit. The recorded audio data is saved as an audio file and uploaded to a cloud server for the analysis unit to access. Furthermore, the proxy answering unit can refer to whitelisted, graylisted, and blacklisted phone numbers and provide appropriate responses. For example, the proxy answering unit will answer calls from whitelisted phone numbers as usual, and for calls from graylisted phone numbers, it will ask questions to determine the risk of fraud. Specifically, the AI ​​will ask the caller questions such as, "May I have your name and the purpose of your call?" and send the answer to the analysis unit. Calls from blacklisted phone numbers will be automatically rejected. Possible rejection messages include, "We do not accept calls from this phone number." This allows the proxy response unit to quickly assess the reliability of the caller and take appropriate action.

[0085] The analysis unit analyzes the conversation content that has been answered by the proxy response unit to determine the risk of fraud. For example, the analysis unit uses speech recognition technology to convert the conversation content into text data and natural language processing technology to evaluate the risk of fraud. Specifically, speech recognition technology converts the caller's voice into text with high accuracy, and natural language processing technology extracts specific keywords and phrases from the text data. For example, the reliability of the caller is confirmed through questions such as, "Hey, [Name]? Don't you know your mother's birthday?" or "Are you [Name]'s boss? Why aren't you in the HR department?" Furthermore, the analysis unit can also compare the conversation content with a database of past fraud cases to check for similar patterns. This allows the analysis unit to analyze the caller's conversation content from multiple angles and determine the risk of fraud with high accuracy. The analysis unit also provides the analysis results to the transmission unit and monitoring unit in real time to support a rapid response.

[0086] The call forwarding unit forwards calls when the analysis unit determines that the risk of fraud is low. For example, if the analysis unit determines that the risk of fraud is low, the call forwarding unit will send a message such as "Connecting you now♪" to the caller and forward the call to the recipient. Specifically, the call forwarding unit receives a signal from the analysis unit and automatically connects the call to the recipient's phone. The call forwarding unit can also select the call forwarding destination based on a whitelist. For example, it can select the most appropriate recipient from multiple phone numbers registered in the whitelist and forward the call. This allows the call forwarding unit to forward only highly reliable calls to recipients, reducing the risk of fraud. Furthermore, the call forwarding unit can monitor the call forwarding status in real time and respond immediately if an anomaly is detected.

[0087] The monitoring unit monitors the conversation even during a call and issues a warning if signs of fraud are detected. For example, if the monitoring unit detects typical fraud phrases such as "Are you so-and-so's mother?" or "So-and-so has been in the hospital after an accident!" during a call, it can issue a warning and terminate the call. Specifically, the monitoring unit combines speech recognition technology and natural language processing technology to analyze the call content in real time and detect signs of fraud. The monitoring unit can also analyze the call content in real time and detect signs of fraud. For example, if a specific keyword or phrase is detected during a call, the monitoring unit will immediately issue a warning and terminate the call. Furthermore, the monitoring unit can record the call content so that the analysis unit can later analyze it in detail. This allows the monitoring unit to monitor the risk of fraud during a call in real time and respond quickly.

[0088] The list management unit manages whitelisted, graylisted, and blacklisted phone numbers. For example, users can manually update the list. Specifically, users can add, delete, and edit phone numbers through a dedicated application or web interface. The list management unit can also automatically update the list based on the results of the proxy response unit and the analysis unit. For example, if the analysis unit determines that a particular phone number is at high risk of fraud, it can automatically add that phone number to the blacklist. Conversely, highly reliable phone numbers are added to the whitelist. This allows the list management unit to always manage the list based on the latest information, improving the overall reliability of the system. Furthermore, the list management unit can record the update history of the list for later review. This allows the list management unit to track user and system operation history and manage the list more effectively.

[0089] The questioning unit can verify the sender's credibility through specific questions. For example, the questioning unit can ask the sender questions such as, "Do you know your mother's birthday?" or "Are you your boss at work? Why aren't you in HR?" to verify the sender's credibility. The questioning unit can also analyze the sender's answers and evaluate their consistency and reliability. For example, the questioning unit can check whether the sender's answers match past data to assess reliability. Furthermore, the questioning unit can ask additional questions based on the sender's answers. For example, if the questioning unit has doubts about the sender's answers, it can ask additional questions to verify reliability. In this way, the questioning unit can reduce the risk of fraud by verifying the sender's credibility.

[0090] The rejection unit can reject calls if it determines that there is a high risk of fraud. For example, if the analysis unit determines that there is a high risk of fraud, the rejection unit will inform the caller, "This call may be fraudulent, so we will end it," and reject the call. The rejection unit can also reject calls if it determines that the caller is unreliable. For example, the rejection unit will reject a call if the caller's responses are inconsistent or if certain keywords or phrases are detected. Furthermore, the rejection unit can automatically reject incoming calls from phone numbers on a blacklist. For example, if an incoming call comes from a phone number on the blacklist, the rejection unit will automatically reject the call. In this way, the rejection unit can protect users by rejecting calls that have a high risk of fraud.

[0091] The warning unit can issue a warning if it detects signs of fraud during a call. For example, if it detects typical fraudulent phrases such as "Are you so-and-so's mother?" or "So-and-so has been in the hospital after an accident!" during a call, it can issue a warning and terminate the call. The warning unit can also analyze the call content in real time to detect signs of fraud. For example, it can detect specific keywords or phrases to determine the possibility of fraud. Furthermore, the warning unit can analyze changes in the caller's voice tone and speed to detect signs of fraud. For example, it will issue a warning if the caller's voice tone changes unnaturally or if the voice speed is abnormally fast. In this way, the warning unit can prevent fraud by detecting signs of fraud during a call and issuing a warning.

[0092] The list update unit can update the list based on the proxy response results. For example, the list update unit automatically updates the whitelist, graylist, and blacklist based on the results of the proxy response unit and the analysis unit. The list update unit also allows users to manually update the list. For example, the list update unit allows users to add specific phone numbers to the whitelist or move them to the blacklist. Furthermore, the list update unit can adjust the frequency and timing of list updates. For example, the list update unit can set a schedule for periodically updating the list. This allows the list update unit to improve the accuracy of the system by updating the list based on the proxy response results.

[0093] The proxy response unit can estimate the user's emotions and adjust the tone and wording of its proxy response based on the estimated emotions. For example, if the user is nervous, the proxy response unit can respond in a calm tone to provide reassurance. If the user is relaxed, the proxy response unit can respond in a friendly tone to create a sense of familiarity. Furthermore, if the user is in a hurry, the proxy response unit can respond in a quick and concise tone to save time. In this way, the proxy response unit can provide a proxy response with an appropriate tone and wording according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the proxy response unit may be performed using AI or not using AI. For example, the proxy response unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0094] The caller ID unit can analyze the characteristics of the caller's voice upon receiving an incoming call and evaluate its reliability by comparing it with past data. For example, the caller ID unit can analyze the tone and speed of the caller's voice and compare it with patterns of past fraudulent calls. The caller ID unit can also compare the characteristics of the caller's voice with past call history and evaluate the degree of similarity. Furthermore, the caller ID unit can compare the characteristics of the caller's voice with fraud characteristics learned by the AI ​​and evaluate its reliability. In this way, the caller ID unit can evaluate reliability by analyzing the characteristics of the caller's voice. Some or all of the above processing in the caller ID unit may be performed using AI, for example, or without AI. For example, the caller ID unit can input the caller's voice characteristic data into a generating AI and have the generating AI perform the reliability evaluation.

[0095] The proxy response unit can analyze the caller's background noise when responding on behalf of the caller and determine the possibility of fraud. For example, the proxy response unit evaluates the possibility of fraud if the caller's background noise contains certain noises. The proxy response unit can also determine the possibility of fraud if the caller's background noise contains the voices of other people. Furthermore, the proxy response unit can also evaluate the possibility of fraud if the caller's background noise contains certain environmental sounds (e.g., call center sounds). In this way, the proxy response unit can determine the possibility of fraud by analyzing the caller's background noise. Some or all of the above processing in the proxy response unit may be performed using AI, for example, or without AI. For example, the proxy response unit can input the caller's background noise data into a generating AI and have the generating AI perform the determination of the possibility of fraud.

[0096] The proxy response unit can estimate the user's emotions and customize the content of the proxy response based on the estimated emotions. For example, if the user is stressed, the proxy response unit can provide simple and reassuring content. If the user is relaxed, the proxy response unit can also provide content that includes detailed information. Furthermore, if the user is in a hurry, the proxy response unit can provide quick and to-the-point content. This allows the proxy response unit to provide appropriate proxy responses according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. The generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the proxy response unit may be performed using AI, for example, or not using AI. For example, the proxy response unit can input the user's voice data into the generative AI and have the generative AI perform emotion estimation.

[0097] The proxy answering unit can adjust the response content by considering the caller's geographical location information when answering on behalf of the caller. For example, if the caller is coming from a specific region, the proxy answering unit will provide information related to that region. The proxy answering unit can also be wary of incoming calls from regions with a high risk of fraud based on the caller's geographical location information. Furthermore, the proxy answering unit can adjust the response content by considering region-specific fraud patterns based on the caller's geographical location information. As a result, the proxy answering unit can provide a more appropriate response by adjusting the response content by considering the caller's geographical location information. Some or all of the above processing in the proxy answering unit may be performed using AI, for example, or without AI. For example, the proxy answering unit can input the caller's geographical location information data into a generating AI and have the generating AI perform the adjustment of the response content.

[0098] The proxy answering unit can optimize the response content by referring to the caller's past call history when answering on behalf of the caller. For example, if the caller has made a call in the past that is suspected of being fraudulent, the proxy answering unit will increase its vigilance. The proxy answering unit can also provide a highly reliable response content based on the caller's past call history. Furthermore, the proxy answering unit can analyze the caller's past call history, assess the risk of fraud, and adjust the response content accordingly. In this way, the proxy answering unit can optimize the response content by referring to the caller's past call history. Some or all of the above processing in the proxy answering unit may be performed using AI, for example, or not using AI. For example, the proxy answering unit can input the caller's past call history data into a generating AI and have the generating AI perform the optimization of the response content.

[0099] The analysis unit can estimate the user's emotions and adjust the accuracy of the analysis based on the estimated emotions. For example, if the user is nervous, the analysis unit can increase the accuracy of the analysis to ensure reliability. Conversely, if the user is relaxed, the analysis unit can adjust the accuracy of the analysis to prioritize efficiency. Furthermore, if the user is in a hurry, the analysis unit can perform the analysis quickly and provide the results. In this way, the analysis unit can perform more reliable analysis by adjusting the accuracy of the analysis according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can input the user's voice data into the generative AI and have the generative AI perform emotion estimation.

[0100] The analysis unit can analyze the tone and speed of the caller's voice during analysis and assess the likelihood of fraud. For example, the analysis unit assesses the likelihood of fraud if the tone of the caller's voice changes unnaturally. The analysis unit can also determine the likelihood of fraud if the speed of the caller's voice is abnormally fast. Furthermore, the analysis unit can also assess the likelihood of fraud by comparing the tone and speed of the caller's voice with patterns of past fraudulent calls. In this way, the analysis unit can assess the likelihood of fraud by analyzing the tone and speed of the caller's voice. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can input the tone and speed data of the caller's voice into a generating AI and have the generating AI perform the fraud likelihood assessment.

[0101] The analysis unit can detect specific phrases and expressions used by the caller during analysis and determine the risk of fraud. For example, the analysis unit evaluates the risk of fraud if the caller uses a specific fraudulent phrase. The analysis unit can also determine the risk of fraud if the caller's expressions match those of past fraudulent calls. Furthermore, the analysis unit can have an AI learn specific phrases used by the caller and evaluate the risk of fraud. This allows the analysis unit to determine the risk of fraud by detecting specific phrases and expressions used by the caller. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can input caller phrase and expression data into a generating AI and have the generating AI perform the fraud risk determination.

[0102] The analysis unit can estimate the user's emotions and adjust the display method of the analysis results based on the estimated user emotions. For example, if the user is nervous, the analysis unit can provide a simple and highly visible display method. If the user is relaxed, the analysis unit can also provide a display method that includes detailed information. Furthermore, if the user is in a hurry, the analysis unit can provide a display method that gets straight to the point. In this way, the analysis unit can provide a more appropriate display by adjusting the display method of the analysis results according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or a generative AI. The generative AI is a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can input the user's voice data into the generative AI and have the generative AI perform emotion estimation.

[0103] The analysis unit can improve the accuracy of the analysis by referring to the caller's past call history during the analysis. For example, the analysis unit improves the accuracy of the analysis based on the caller's past call history. The analysis unit can also analyze the caller's past call history and assess the risk of fraud. Furthermore, the analysis unit can correct the analysis results by referring to the caller's past call history. In this way, the analysis unit can improve the accuracy of the analysis by referring to the caller's past call history. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without using AI. For example, the analysis unit can input the caller's past call history data into a generating AI and have the generating AI perform the analysis accuracy improvement.

[0104] The analysis unit can correct the analysis results by considering the geographical location information of the caller during the analysis. For example, the analysis unit can assess the risk of fraud based on the geographical location information of the caller. The analysis unit can also correct the analysis results by considering the geographical location information of the caller. Furthermore, the analysis unit can adjust the analysis results by considering region-specific fraud patterns based on the geographical location information of the caller. This allows the analysis unit to perform more accurate analysis by correcting the analysis results by considering the geographical location information of the caller. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without using AI. For example, the analysis unit can input the geographical location information data of the caller into a generating AI and have the generating AI perform the correction of the analysis results.

[0105] The transfer unit can estimate the user's emotions and adjust the timing of the transfer based on the estimated emotions. For example, if the user is nervous, the transfer unit can transfer quickly to provide a sense of security. The transfer unit can also transfer at an appropriate time if the user is relaxed. Furthermore, if the user is in a hurry, the transfer unit can transfer quickly to save time. This allows the transfer unit to transfer at a more appropriate time by adjusting the timing according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the transfer unit may be performed using AI, or not. For example, the transfer unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0106] The forwarding unit can re-evaluate the sender's trustworthiness during forwarding and determine whether or not to forward the message. For example, the forwarding unit can re-evaluate the sender's trustworthiness before forwarding and determine the risk of fraud. The forwarding unit can also re-evaluate the sender's trustworthiness during forwarding and determine whether or not to forward the message. Furthermore, the forwarding unit can re-evaluate the sender's trustworthiness before forwarding and refuse to forward the message if necessary. This allows the forwarding unit to perform more reliable forwarding by re-evaluating the sender's trustworthiness during forwarding. Some or all of the above processing in the forwarding unit may be performed using AI, for example, or without AI. For example, the forwarding unit can input sender trustworthiness data into a generating AI and have the generating AI perform the trustworthiness re-evaluation.

[0107] The transfer unit can optimize the transfer by considering the recipient's current situation at the time of transfer. For example, if the recipient is busy, the transfer unit can delay the transfer or transfer it at a different time. The transfer unit can also consider the recipient's current location and transfer at an appropriate time. Furthermore, the transfer unit can transfer at the optimal time based on the recipient's schedule information. In this way, the transfer unit can optimize the transfer by considering the recipient's current situation, enabling transfers at a more appropriate time. Some or all of the above processing in the transfer unit may be performed using AI, for example, or without AI. For example, the transfer unit can input the recipient's schedule information data into a generating AI and have the generating AI perform the transfer optimization.

[0108] The transfer unit can estimate the user's emotions and determine the priority of transfers based on the estimated emotions. For example, if the user is tense, the transfer unit can transfer quickly to provide reassurance. It can also transfer at an appropriate time if the user is relaxed. Furthermore, if the user is in a hurry, the transfer unit can transfer quickly to save time. This allows the transfer unit to perform more appropriate transfers by determining the priority of transfers according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the transfer unit may be performed using AI, or not. For example, the transfer unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0109] The transfer unit can determine whether or not to transfer a call by referring to the caller's past call history at the time of transfer. For example, the transfer unit can determine whether or not to transfer a call based on the caller's past call history. The transfer unit can also analyze the caller's past call history, assess the risk of fraud, and then decide whether or not to transfer a call. Furthermore, the transfer unit can perform a highly reliable transfer by referring to the caller's past call history. Thus, the transfer unit can determine whether or not to transfer a call by referring to the caller's past call history. Some or all of the above processing in the transfer unit may be performed using AI, for example, or not using AI. For example, the transfer unit can input the caller's past call history data into a generating AI and have the generating AI perform the determination of whether or not to transfer a call.

[0110] The transfer unit can adjust the timing of the transfer by considering the recipient's schedule information during the transfer process. For example, the transfer unit can perform the transfer at the optimal timing based on the recipient's schedule information. The transfer unit can also adjust the timing of the transfer by considering the recipient's schedule information. Furthermore, the transfer unit can refer to the recipient's schedule information and perform the transfer at an appropriate time. As a result, the transfer unit can perform the transfer at a more appropriate timing by adjusting the timing of the transfer by considering the recipient's schedule information. Some or all of the above processing in the transfer unit may be performed using AI, for example, or without AI. For example, the transfer unit can input the recipient's schedule information data into a generating AI and have the generating AI perform the adjustment of the transfer timing.

[0111] The monitoring unit can estimate the user's emotions and adjust the accuracy of monitoring based on the estimated emotions. For example, if the user is tense, the monitoring unit can increase the accuracy of monitoring to ensure reliability. Conversely, if the user is relaxed, the monitoring unit can adjust the accuracy of monitoring to prioritize efficiency. Furthermore, if the user is in a hurry, the monitoring unit can perform monitoring quickly and provide results. In this way, the monitoring unit can enable more reliable monitoring by adjusting the accuracy of monitoring according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the monitoring unit may be performed using AI, or not using AI. For example, the monitoring unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0112] The monitoring unit can analyze the caller's voice tone and speed in real time during a call and detect signs of fraud. For example, the monitoring unit can detect signs of fraud if the caller's voice tone changes unnaturally. The monitoring unit can also determine if signs of fraud exist if the caller's voice speed is abnormally fast. Furthermore, the monitoring unit can also detect signs of fraud by comparing the caller's voice tone and speed with patterns from past fraudulent calls. In this way, the monitoring unit can detect signs of fraud by analyzing the caller's voice tone and speed in real time. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input the caller's voice tone and speed data into a generating AI and have the generating AI perform the detection of signs of fraud.

[0113] The monitoring unit can detect specific phrases and expressions in real time during a call and assess the risk of fraud. For example, the monitoring unit assesses the risk of fraud if the caller uses a specific fraudulent phrase. The monitoring unit can also determine the risk of fraud if the caller's expressions match those of past fraudulent calls. Furthermore, the monitoring unit can have an AI learn specific phrases used by the caller and assess the risk of fraud. This allows the monitoring unit to assess the risk of fraud by detecting specific phrases and expressions in real time. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input caller phrase and expression data into a generating AI and have the generating AI perform the fraud risk assessment.

[0114] The monitoring unit can estimate the user's emotions and adjust the monitoring frequency based on the estimated emotions. For example, if the user is nervous, the monitoring unit can increase the monitoring frequency to ensure reliability. Conversely, if the user is relaxed, the monitoring unit can adjust the monitoring frequency to prioritize efficiency. Furthermore, if the user is in a hurry, the monitoring unit can perform monitoring quickly and provide results. This allows the monitoring unit to provide more appropriate monitoring by adjusting the monitoring frequency according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above-described processing in the monitoring unit may be performed using AI, or not. For example, the monitoring unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0115] The monitoring unit can improve the accuracy of monitoring by referring to the caller's past call history during a call. For example, the monitoring unit improves the accuracy of monitoring based on the caller's past call history. The monitoring unit can also analyze the caller's past call history and assess the risk of fraud. Furthermore, the monitoring unit can correct the monitoring results by referring to the caller's past call history. In this way, the monitoring unit can improve the accuracy of monitoring by referring to the caller's past call history. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input the caller's past call history data into a generating AI and have the generating AI perform the improvement of monitoring accuracy.

[0116] The monitoring unit can correct the monitoring results during a call by taking into account the caller's geographical location information. For example, the monitoring unit can assess the risk of fraud based on the caller's geographical location information. The monitoring unit can also correct the monitoring results by taking into account the caller's geographical location information. Furthermore, the monitoring unit can adjust the monitoring results by taking into account region-specific fraud patterns based on the caller's geographical location information. This allows the monitoring unit to perform more accurate monitoring by correcting the monitoring results by taking into account the caller's geographical location information. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without using AI. For example, the monitoring unit can input the caller's geographical location information data into a generating AI and have the generating AI perform the correction of the monitoring results.

[0117] The list management unit can estimate the user's emotions and adjust the list update frequency based on the estimated emotions. For example, if the user is stressed, the list management unit can increase the list update frequency to ensure reliability. Conversely, if the user is relaxed, the list management unit can adjust the update frequency to prioritize efficiency. Furthermore, if the user is in a hurry, the list management unit can quickly update the list and provide results. In this way, the list management unit can achieve more reliable list management by adjusting the list update frequency according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the list management unit may be performed using AI, or not using AI. For example, the list management unit can input the user's voice data into the generative AI and have the generative AI perform emotion estimation.

[0118] The list management unit can improve the accuracy of the list by referring to the caller's past call history during list management. For example, the list management unit improves the accuracy of the list based on the caller's past call history. The list management unit can also analyze the caller's past call history and assess the risk of fraud. Furthermore, the list management unit can correct the contents of the list by referring to the caller's past call history. In this way, the list management unit can improve the accuracy of the list by referring to the caller's past call history. Some or all of the above processing in the list management unit may be performed using AI, for example, or without using AI. For example, the list management unit can input the caller's past call history data into a generating AI and have the generating AI perform the list accuracy improvement.

[0119] The list management unit can optimize the list content by considering the geographical location information of the sender during list management. For example, the list management unit optimizes the list content based on the geographical location information of the sender. The list management unit can also correct the list content by considering the geographical location information of the sender. Furthermore, the list management unit can adjust the list content by considering region-specific fraud patterns based on the geographical location information of the sender. As a result, the list management unit can perform more accurate list management by optimizing the list content by considering the geographical location information of the sender. Some or all of the above processing in the list management unit may be performed using AI, for example, or without using AI. For example, the list management unit can input geographical location data of the sender into a generating AI and have the generating AI perform the optimization of the list content.

[0120] The list management unit can estimate the user's emotions and determine the priority of lists based on the estimated emotions. For example, if the user is stressed, the list management unit will prioritize reliable lists. It can also prioritize efficiency-focused lists if the user is relaxed. Furthermore, if the user is in a hurry, the list management unit can quickly update lists and determine priorities. This allows the list management unit to perform more appropriate list management by prioritizing lists according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the list management unit may be performed using AI, or not. For example, the list management unit can input user voice data into a generative AI and have the generative AI perform emotion estimation.

[0121] The list management unit can improve the accuracy of the list by analyzing the characteristics of the caller's voice during list management. For example, the list management unit can improve the accuracy of the list by analyzing the tone and speed of the caller's voice. The list management unit can also correct the list contents by comparing the characteristics of the caller's voice with past call history. Furthermore, the list management unit can improve the accuracy of the list by having an AI learn the characteristics of the caller's voice. In this way, the list management unit can improve the accuracy of the list by analyzing the characteristics of the caller's voice. Some or all of the above processing in the list management unit may be performed using an AI, for example, or without an AI. For example, the list management unit can input caller voice characteristic data into a generating AI and have the generating AI perform list accuracy improvement.

[0122] The list management unit can detect specific phrases and expressions used by callers during list management and update the list content. For example, if a caller uses a specific fraudulent phrase, the list management unit will update the list content. The list management unit can also correct the list content if the caller's expressions match those of past fraudulent calls. Furthermore, the list management unit can update the list content by having an AI learn specific phrases used by callers. This allows the list management unit to update the list content by detecting specific phrases and expressions used by callers. Some or all of the above processing in the list management unit may be performed using an AI, for example, or without an AI. For example, the list management unit can input caller phrase and expression data into a generating AI and have the generating AI perform the update of the list content.

[0123] The questioning unit can estimate the user's emotions and adjust the content and tone of the questions based on the estimated emotions. For example, if the user is nervous, the questioning unit can ask questions in a calm tone to provide reassurance. If the user is relaxed, the questioning unit can ask questions in a friendly tone to create a sense of familiarity. Furthermore, if the user is in a hurry, the questioning unit can ask questions in a quick and concise tone to save time. In this way, the questioning unit can ask more appropriate questions by adjusting the content and tone of the questions according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the questioning unit may be performed using AI, or not using AI. For example, the questioning unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0124] The questioning unit can analyze the caller's tone and speed of voice when a question is asked and evaluate its reliability. For example, the questioning unit evaluates reliability if the caller's tone of voice changes unnaturally. It can also determine reliability if the caller's speed of voice is abnormally fast. Furthermore, the questioning unit can compare the caller's tone and speed of voice with patterns of past fraudulent calls and evaluate reliability. In this way, the questioning unit can evaluate reliability by analyzing the caller's tone and speed of voice. Some or all of the above processing in the questioning unit may be performed using AI, for example, or without AI. For example, the questioning unit can input the caller's tone and speed data into a generating AI and have the generating AI perform the reliability evaluation.

[0125] The questioning unit can detect specific phrases and expressions used by the caller when asking a question and determine its reliability. For example, the questioning unit evaluates reliability if the caller uses a specific fraudulent phrase. The questioning unit can also determine reliability if the caller's expression matches that of past fraudulent calls. Furthermore, the questioning unit can have an AI learn specific phrases used by the caller and evaluate reliability based on that learning. This allows the questioning unit to determine reliability by detecting specific phrases and expressions used by the caller. Some or all of the above processing in the questioning unit may be performed using AI, for example, or without AI. For example, the questioning unit can input caller phrase and expression data into a generating AI and have the generating AI perform the reliability determination.

[0126] The questioning unit can estimate the user's emotions and adjust the order of questions based on the estimated emotions. For example, if the user is nervous, the questioning unit can start with simple questions to create a sense of reassurance. If the user is relaxed, the questioning unit can ask more detailed questions to confirm their trustworthiness. Furthermore, if the user is in a hurry, the questioning unit can prioritize important questions to save time. In this way, the questioning unit can ask more appropriate questions by adjusting the order of questions according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the questioning unit may be performed using AI or not using AI. For example, the questioning unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0127] The questioning unit can optimize the question content by referring to the caller's past call history when asking a question. For example, the questioning unit can provide highly reliable question content based on the caller's past call history. The questioning unit can also analyze the caller's past call history, assess the risk of fraud, and adjust the question content accordingly. Furthermore, the questioning unit can provide optimal question content by referring to the caller's past call history. In this way, the questioning unit can optimize the question content by referring to the caller's past call history. Some or all of the above processing in the questioning unit may be performed using AI, for example, or not using AI. For example, the questioning unit can input the caller's past call history data into a generating AI and have the generating AI perform the optimization of the question content.

[0128] The questioning unit can adjust the content of questions by considering the caller's geographical location information. For example, the questioning unit can adjust the content of questions by considering region-specific fraud patterns based on the caller's geographical location information. The questioning unit can also provide highly reliable questions by considering the caller's geographical location information. Furthermore, the questioning unit can provide optimal questions based on the caller's geographical location information. In this way, the questioning unit can ask more appropriate questions by adjusting the content of questions by considering the caller's geographical location information. Some or all of the above processing in the questioning unit may be performed using AI, for example, or without AI. For example, the questioning unit can input the caller's geographical location information data into a generating AI and have the generating AI perform the adjustment of the question content.

[0129] The rejection unit can estimate the user's emotions and adjust the timing of the rejection based on the estimated emotions. For example, if the user is tense, the rejection unit can quickly reject the request to provide reassurance. It can also reject the request at an appropriate time if the user is relaxed. Furthermore, if the user is in a hurry, the rejection unit can quickly reject the request to save time. This allows the rejection unit to reject requests at a more appropriate time by adjusting the timing according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above-described processes in the rejection unit may be performed using AI, or not. For example, the rejection unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0130] The rejection unit can detect specific phrases or expressions used by the caller when rejecting a call, thereby strengthening the basis for rejection. For example, the rejection unit strengthens the basis for rejection if the caller uses specific fraudulent phrases. The rejection unit can also clarify the basis for rejection if the caller's expressions match those of past fraudulent calls. Furthermore, the rejection unit can have an AI learn specific phrases used by the caller, thereby strengthening the basis for rejection. This allows the rejection unit to strengthen the basis for rejection by detecting specific phrases or expressions used by the caller. Some or all of the above processing in the rejection unit may be performed using AI, for example, or without AI. For example, the rejection unit can input caller phrase and expression data into a generating AI and have the generating AI perform the strengthening of the basis for rejection.

[0131] The rejection unit can estimate the user's emotions and determine the priority of rejections based on the estimated emotions. For example, if the user is tense, the rejection unit can quickly reject the request to provide reassurance. It can also reject the request at an appropriate time if the user is relaxed. Furthermore, if the user is in a hurry, the rejection unit can quickly reject the request to save time. This allows the rejection unit to perform more appropriate rejections by determining the priority of rejections according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above-described processes in the rejection unit may be performed using AI, or not. For example, the rejection unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0132] The rejection unit can determine whether to reject a call by referring to the caller's past call history. For example, the rejection unit can determine whether to reject a call based on the caller's past call history. The rejection unit can also analyze the caller's past call history, assess the risk of fraud, and then decide whether to reject the call. Furthermore, the rejection unit can make a highly reliable rejection by referring to the caller's past call history. Thus, the rejection unit can determine whether to reject a call by referring to the caller's past call history. Some or all of the above processing in the rejection unit may be performed using AI, for example, or without AI. For example, the rejection unit can input the caller's past call history data into a generating AI and have the generating AI perform the decision on whether to reject the call.

[0133] The rejection unit can adjust the reason for rejection by considering the caller's geographical location information at the time of rejection. For example, the rejection unit can assess the risk of fraud based on the caller's geographical location information. The rejection unit can also adjust the reason for rejection by considering the caller's geographical location information. Furthermore, the rejection unit can adjust the reason for rejection by considering region-specific fraud patterns based on the caller's geographical location information. This allows the rejection unit to perform more accurate rejections by adjusting the reason for rejection by considering the caller's geographical location information. Some or all of the above processing in the rejection unit may be performed using AI, for example, or without using AI. For example, the rejection unit can input the caller's geographical location information data into a generating AI and have the generating AI perform the adjustment of the reason for rejection.

[0134] The warning unit can estimate the user's emotions and adjust the content and tone of the warning based on the estimated emotions. For example, if the user is tense, the warning unit can issue a warning in a calm tone to provide reassurance. If the user is relaxed, the warning unit can issue a warning in a friendly tone to create a sense of familiarity. Furthermore, if the user is in a hurry, the warning unit can issue a warning in a quick and concise tone to save time. In this way, the warning unit can provide more appropriate warnings by adjusting the content and tone of the warning according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the warning unit may be performed using AI, or not using AI. For example, the warning unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0135] The warning unit can analyze the caller's voice tone and speed at the time of a warning and optimize the timing of the warning. For example, if the caller's voice tone changes unnaturally, the warning unit will optimize the timing of the warning. The warning unit can also determine the timing of the warning if the caller's voice speed is abnormally fast. Furthermore, the warning unit can compare the caller's voice tone and speed with patterns of past fraudulent calls and optimize the timing of the warning. In this way, the warning unit can optimize the timing of the warning by analyzing the caller's voice tone and speed. Some or all of the above processing in the warning unit may be performed using AI, for example, or without AI. For example, the warning unit can input the caller's voice tone and speed data into a generating AI and have the generating AI perform the optimization of the warning timing.

[0136] The warning unit can detect specific phrases and expressions used by the caller when issuing a warning, thereby strengthening the basis for the warning. For example, the warning unit strengthens the basis for the warning if the caller uses a specific fraudulent phrase. The warning unit can also clarify the basis for the warning if the caller's expression matches that of past fraudulent calls. Furthermore, the warning unit can have an AI learn specific phrases used by the caller, thereby strengthening the basis for the warning. This allows the warning unit to strengthen the basis for the warning by detecting specific phrases and expressions used by the caller. Some or all of the above processing in the warning unit may be performed using AI, for example, or without AI. For example, the warning unit can input caller phrase and expression data into a generating AI and have the generating AI perform the strengthening of the basis for the warning.

[0137] The warning unit can estimate the user's emotions and determine the priority of warnings based on the estimated emotions. For example, if the user is tense, the warning unit can issue a warning quickly to provide reassurance. It can also issue a warning at an appropriate time if the user is relaxed. Furthermore, if the user is in a hurry, the warning unit can issue a warning quickly to save time. This allows the warning unit to provide more appropriate warnings by prioritizing them according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the warning unit may be performed using AI, or not. For example, the warning unit can input user voice data into a generative AI and have the generative AI perform emotion estimation.

[0138] The warning unit can optimize the content of a warning by referring to the caller's past call history at the time of the warning. For example, the warning unit optimizes the content of a warning based on the caller's past call history. The warning unit can also analyze the caller's past call history, assess the risk of fraud, and adjust the content of the warning. Furthermore, the warning unit can provide the most appropriate warning content by referring to the caller's past call history. In this way, the warning unit can optimize the content of a warning by referring to the caller's past call history. Some or all of the above processing in the warning unit may be performed using AI, for example, or without AI. For example, the warning unit can input the caller's past call history data into a generating AI and have the generating AI perform the optimization of the warning content.

[0139] The warning unit can adjust the content of a warning by considering the caller's geographical location information at the time of the warning. For example, the warning unit can adjust the warning content by considering region-specific fraud patterns based on the caller's geographical location information. The warning unit can also provide highly reliable warning content by considering the caller's geographical location information. Furthermore, the warning unit can provide optimal warning content based on the caller's geographical location information. As a result, the warning unit can provide more accurate warnings by adjusting the content of the warning by considering the caller's geographical location information. Some or all of the above processing in the warning unit may be performed using AI, for example, or without AI. For example, the warning unit can input the caller's geographical location information data into a generating AI and have the generating AI perform the correction of the warning content.

[0140] The list update unit can estimate the user's emotions and adjust the list update frequency based on the estimated emotions. For example, if the user is stressed, the list update unit can increase the update frequency to ensure reliability. Conversely, if the user is relaxed, the list update unit can adjust the update frequency to prioritize efficiency. Furthermore, if the user is in a hurry, the list update unit can quickly update the list and provide results. This allows the list update unit to manage lists more reliably by adjusting the update frequency according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or a generative AI. The generative AI is, but is not limited to, a text generation AI (e.g., LLM) or a multimodal generation AI. Some or all of the above processing in the list update unit may be performed using AI or not. For example, the list update unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0141] The list update unit can improve the accuracy of the list by analyzing the tone and speed of the caller's voice when updating the list. For example, the list update unit can analyze the tone and speed of the caller's voice to improve the accuracy of the list. The list update unit can also correct the list content by comparing the characteristics of the caller's voice with past call history. Furthermore, the list update unit can improve the accuracy of the list by having the AI ​​learn the characteristics of the caller's voice. In this way, the list update unit can improve the accuracy of the list by analyzing the tone and speed of the caller's voice. Some or all of the above processing in the list update unit may be performed using AI, for example, or without using AI. For example, the list update unit can input caller voice characteristic data into a generating AI and have the generating AI perform the list accuracy improvement.

[0142] The list update unit can detect specific phrases or expressions used by callers when updating the list and update the list content. For example, if a caller uses a specific fraudulent phrase, the list update unit will update the list content. The list update unit can also correct the list content if the caller's expressions match those of past fraudulent calls. Furthermore, the list update unit can have an AI learn specific phrases used by callers and update the list content accordingly. This allows the list update unit to update the list content by detecting specific phrases or expressions used by callers. Some or all of the above processing in the list update unit may be performed using an AI, for example, or without an AI. For example, the list update unit can input caller phrase and expression data into a generating AI and have the generating AI perform the list content update.

[0143] The list update unit can estimate the user's emotions and determine the priority of lists based on the estimated emotions. For example, if the user is stressed, the list update unit will prioritize reliable lists. If the user is relaxed, the list update unit can also prioritize efficiency-focused lists. Furthermore, if the user is in a hurry, the list update unit can quickly update the lists and determine priorities. This allows the list update unit to manage lists more effectively by prioritizing them according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the list update unit may be performed using AI, or not using AI. For example, the list update unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0144] The list update unit can update the list content by referring to the caller's past call history when updating the list. For example, the list update unit updates the list content based on the caller's past call history. The list update unit can also analyze the caller's past call history, assess the risk of fraud, and adjust the list content accordingly. Furthermore, the list update unit can refer to the caller's past call history to provide the most optimal list content. Thus, the list update unit can update the list content by referring to the caller's past call history. Some or all of the above processing in the list update unit may be performed using AI, for example, or without AI. For example, the list update unit can input the caller's past call history data into a generating AI and have the generating AI perform the list content update.

[0145] The list update unit can optimize the list content by considering the geographical location information of the sender when updating the list. For example, the list update unit adjusts the list content by considering region-specific fraud patterns based on the geographical location information of the sender. The list update unit can also provide highly reliable list content by considering the geographical location information of the sender. Furthermore, the list update unit can provide optimal list content based on the geographical location information of the sender. As a result, the list update unit can achieve more accurate list management by optimizing the list content by considering the geographical location information of the sender. Some or all of the above processing in the list update unit may be performed using AI, for example, or without using AI. For example, the list update unit can input the geographical location information data of the sender into a generating AI and have the generating AI perform the optimization of the list content.

[0146] The system according to the embodiment is not limited to the example described above, and various modifications are possible, for example, as follows.

[0147] The analysis unit can estimate the user's emotions and adjust the accuracy of the analysis based on the estimated emotions. For example, if the user is nervous, the analysis unit can increase the accuracy of the analysis to ensure reliability. Conversely, if the user is relaxed, the analysis unit can adjust the accuracy of the analysis to prioritize efficiency. Furthermore, if the user is in a hurry, the analysis unit can perform the analysis quickly and provide the results. In this way, the analysis unit can perform more reliable analysis by adjusting the accuracy of the analysis according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the analysis unit may be performed using AI, for example, or without AI. For example, the analysis unit can input the user's voice data into the generative AI and have the generative AI perform emotion estimation.

[0148] The caller ID unit can analyze the characteristics of the caller's voice upon receiving an incoming call and evaluate its reliability by comparing it with past data. For example, the caller ID unit can analyze the tone and speed of the caller's voice and compare it with patterns of past fraudulent calls. The caller ID unit can also compare the characteristics of the caller's voice with past call history and evaluate the degree of similarity. Furthermore, the caller ID unit can compare the characteristics of the caller's voice with fraud characteristics learned by the AI ​​and evaluate its reliability. In this way, the caller ID unit can evaluate reliability by analyzing the characteristics of the caller's voice. Some or all of the above processing in the caller ID unit may be performed using AI, for example, or without AI. For example, the caller ID unit can input the caller's voice characteristic data into a generating AI and have the generating AI perform the reliability evaluation.

[0149] The proxy response unit can analyze the caller's background noise when responding on behalf of the caller and determine the possibility of fraud. For example, the proxy response unit evaluates the possibility of fraud if the caller's background noise contains certain noises. The proxy response unit can also determine the possibility of fraud if the caller's background noise contains the voices of other people. Furthermore, the proxy response unit can also evaluate the possibility of fraud if the caller's background noise contains certain environmental sounds (e.g., call center sounds). In this way, the proxy response unit can determine the possibility of fraud by analyzing the caller's background noise. Some or all of the above processing in the proxy response unit may be performed using AI, for example, or without AI. For example, the proxy response unit can input the caller's background noise data into a generating AI and have the generating AI perform the determination of the possibility of fraud.

[0150] The proxy response unit can estimate the user's emotions and adjust the tone and wording of its proxy response based on the estimated emotions. For example, if the user is nervous, the proxy response unit can respond in a calm tone to provide reassurance. If the user is relaxed, the proxy response unit can respond in a friendly tone to create a sense of familiarity. Furthermore, if the user is in a hurry, the proxy response unit can respond in a quick and concise tone to save time. In this way, the proxy response unit can provide a proxy response with an appropriate tone and wording according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the proxy response unit may be performed using AI or not using AI. For example, the proxy response unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0151] The forwarding unit can re-evaluate the sender's trustworthiness during forwarding and determine whether or not to forward the message. For example, the forwarding unit can re-evaluate the sender's trustworthiness before forwarding and determine the risk of fraud. The forwarding unit can also re-evaluate the sender's trustworthiness during forwarding and determine whether or not to forward the message. Furthermore, the forwarding unit can re-evaluate the sender's trustworthiness before forwarding and refuse to forward the message if necessary. This allows the forwarding unit to perform more reliable forwarding by re-evaluating the sender's trustworthiness during forwarding. Some or all of the above processing in the forwarding unit may be performed using AI, for example, or without AI. For example, the forwarding unit can input sender trustworthiness data into a generating AI and have the generating AI perform the trustworthiness re-evaluation.

[0152] The transfer unit can estimate the user's emotions and adjust the timing of the transfer based on the estimated emotions. For example, if the user is nervous, the transfer unit can transfer quickly to provide a sense of security. The transfer unit can also transfer at an appropriate time if the user is relaxed. Furthermore, if the user is in a hurry, the transfer unit can transfer quickly to save time. This allows the transfer unit to transfer at a more appropriate time by adjusting the timing according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, such as an emotion engine or generative AI. Generative AI may be, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the transfer unit may be performed using AI, or not. For example, the transfer unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0153] The monitoring unit can analyze the caller's voice tone and speed in real time during a call and detect signs of fraud. For example, the monitoring unit can detect signs of fraud if the caller's voice tone changes unnaturally. The monitoring unit can also determine if signs of fraud exist if the caller's voice speed is abnormally fast. Furthermore, the monitoring unit can also detect signs of fraud by comparing the caller's voice tone and speed with patterns from past fraudulent calls. In this way, the monitoring unit can detect signs of fraud by analyzing the caller's voice tone and speed in real time. Some or all of the above processing in the monitoring unit may be performed using AI, for example, or without AI. For example, the monitoring unit can input the caller's voice tone and speed data into a generating AI and have the generating AI perform the detection of signs of fraud.

[0154] The monitoring unit can estimate the user's emotions and adjust the accuracy of monitoring based on the estimated emotions. For example, if the user is tense, the monitoring unit can increase the accuracy of monitoring to ensure reliability. Conversely, if the user is relaxed, the monitoring unit can adjust the accuracy of monitoring to prioritize efficiency. Furthermore, if the user is in a hurry, the monitoring unit can perform monitoring quickly and provide results. In this way, the monitoring unit can enable more reliable monitoring by adjusting the accuracy of monitoring according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the monitoring unit may be performed using AI, or not using AI. For example, the monitoring unit can input the user's voice data into a generative AI and have the generative AI perform emotion estimation.

[0155] The list management unit can improve the accuracy of the list by referring to the caller's past call history during list management. For example, the list management unit improves the accuracy of the list based on the caller's past call history. The list management unit can also analyze the caller's past call history and assess the risk of fraud. Furthermore, the list management unit can correct the contents of the list by referring to the caller's past call history. In this way, the list management unit can improve the accuracy of the list by referring to the caller's past call history. Some or all of the above processing in the list management unit may be performed using AI, for example, or without using AI. For example, the list management unit can input the caller's past call history data into a generating AI and have the generating AI perform the list accuracy improvement.

[0156] The list management unit can estimate the user's emotions and adjust the list update frequency based on the estimated emotions. For example, if the user is stressed, the list management unit can increase the list update frequency to ensure reliability. Conversely, if the user is relaxed, the list management unit can adjust the update frequency to prioritize efficiency. Furthermore, if the user is in a hurry, the list management unit can quickly update the list and provide results. In this way, the list management unit can achieve more reliable list management by adjusting the list update frequency according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generative AI. Generative AI is, but is not limited to, text generation AI (e.g., LLM) or multimodal generation AI. Some or all of the above processing in the list management unit may be performed using AI, or not using AI. For example, the list management unit can input the user's voice data into the generative AI and have the generative AI perform emotion estimation.

[0157] The following briefly describes the processing flow for example form 2.

[0158] Step 1: The proxy answering unit uses AI to answer incoming calls. For example, when a call comes in, it sends a message to the caller such as, "The fraud prevention AI is answering your call." The proxy answering unit can also record the caller's conversation and send it to the analysis unit. Furthermore, the proxy answering unit can refer to whitelisted, graylisted, and blacklisted phone numbers and respond appropriately. For example, it will answer calls from whitelisted phone numbers as usual, ask questions to determine the risk of fraud for calls from graylisted phone numbers, and automatically reject calls from blacklisted phone numbers. Step 2: The analysis unit analyzes the conversation content that was answered by the proxy response unit and determines the risk of fraud. For example, it uses speech recognition technology to convert the conversation content into text data and natural language processing technology to evaluate the risk of fraud. It can also detect specific keywords or phrases to determine the possibility of fraud. For example, it can verify the trustworthiness of the caller through questions such as, "Hey, [Name]? Don't you know your mother's birthday?" or "Are you [Name]'s boss? Why aren't you in the HR department?" Step 3: The forwarding unit forwards the call if the analysis unit determines that the risk of fraud is low. For example, if the analysis unit determines that the risk of fraud is low, it will send a message such as "Connecting you now♪" to the caller and forward the call to the recipient. It is also possible to select the call forwarding destination based on a whitelist. Step 4: The monitoring unit monitors the conversation even while a call is in progress and issues a warning if signs of fraud are detected. For example, if typical fraud phrases such as "Are you so-and-so's mother?" or "So-and-so has been in the hospital after an accident!" are detected during a call, a warning will be issued and the call can be terminated. It can also analyze the call content in real time to detect signs of fraud. Step 5: The list management unit manages the whitelist, graylist, and blacklist of phone numbers. For example, users can manually update the list. Alternatively, the list can be automatically updated based on the results of the proxy answering unit and the analysis unit.

[0159] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0160] Data generation model 58 is a form of so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> Examples of generative AI include text generation AI, image generation AI, and multimodal generation AI. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images (e.g., still image data or video data). The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference result in one or more data formats from audio data, text data, and image data. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the specific processing described above using the data generation model 58. The data generation model 58 may be a fine-tuned model that outputs inference results from prompts that do not contain instructions, in which case the data generation model 58 can output inference results from prompts that do not contain instructions. In the data processing device 12, etc., there are multiple types of data generation models 58, and the data generation model 58 includes AI other than generative AI. AI other than generative AI includes, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVMs), k-means clustering, convolutional neural networks (CNNs), recurrent neural networks (RNNs), generative adversarial networks (GANs), or naive Bayes, and can perform various processes, but is not limited to these examples. Also, the AI ​​may be an AI agent. Furthermore, when the processing of each of the above parts is performed by the AI, the processing may be performed by the AI ​​in part or in whole, but is not limited to this example.Furthermore, processing performed by AI, including generative AI, may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by AI, including generative AI.

[0161] Furthermore, the processing performed by the data processing system 10 described above is carried out by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart device 14, but it may also be carried out by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart device 14. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information necessary for processing from the smart device 14 or an external device, and the smart device 14 acquires or collects information necessary for processing from the data processing device 12 or an external device.

[0162] Each of the multiple elements described above, including the proxy answering unit, analysis unit, transfer unit, monitoring unit, and list management unit, is implemented in at least one of the smart device 14 and the data processing unit 12. For example, the proxy answering unit is implemented by the control unit 46A of the smart device 14, and the AI ​​answers the call on behalf of the caller when an incoming call is received. The analysis unit is implemented by the identification processing unit 290 of the data processing unit 12, and analyzes the content of the conversation to determine the risk of fraud. The transfer unit is implemented by the control unit 46A of the smart device 14, and transfers the call if it is determined that the risk of fraud is low. The monitoring unit is implemented by the identification processing unit 290 of the data processing unit 12, and monitors the conversation even while a call is in progress, and issues a warning if signs of fraud are detected. The list management unit is implemented by the identification processing unit 290 of the data processing unit 12, and manages telephone numbers in the whitelist, graylist, and blacklist. The correspondence between each unit and the device or control unit is not limited to the example described above, and various changes are possible.

[0163] [Second Embodiment] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0164] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0165] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN and / or LAN.

[0166] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0167] The microphone 238 receives voice signals from the user and accepts instructions from the user. The microphone 238 captures the voice signals from the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0168] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, which captures images of the area around the user (for example, an imaging range defined by a field of view equivalent to the field of vision of a typical healthy person).

[0169] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0170] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing by the processor 28. The storage 32 stores the specific processing program 56.

[0171] The processor 28 reads a 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 acting as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0172] Storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotions using the emotion identification model 59 and perform identification processing using the user's emotions. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotions, including but not limited to these examples. Furthermore, emotion estimation and prediction also include, for example, emotion analysis.

[0173] In the smart glasses 214, specific processing is performed by the processor 46. The storage 50 stores a specific processing program 60. The processor 46 reads the specific processing program 60 from the storage 50 and executes the read specific processing program 60 on the RAM 48. The specific processing is realized by the processor 46 acting as a control unit 46A according to the specific processing program 60 executed on the RAM 48. The smart glasses 214 also have a data generation model 58 and an emotion identification model 59, similar to the data generation model and emotion identification model 59, and can perform processing similar to that of the specific processing unit 290 using these models.

[0174] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains processing results (such as prediction results) using the data generation model 58 by communicating with the server device that has the data generation model 58. Also, the data processing device 12 may be a server device or a terminal device owned by the user (for example, a mobile phone, robot, home appliance, etc.).

[0175] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0176] The data generation model 58 is a so-called generative AI. An example of a data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and inference data such as audio data representing speech, text data representing text, and image data representing images (e.g., still image data or video data). The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference result in one or more data formats such as audio data, text data, and image data. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the specific processing described above using the data generation model 58. The data generation model 58 may be a fine-tuned model that outputs inference results from prompts that do not contain instructions, in which case the data generation model 58 can output inference results from prompts that do not contain instructions. In the data processing device 12, etc., there are multiple types of data generation models 58, and the data generation model 58 includes AI other than generative AI. AI other than generative AI includes, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVM), k-means clustering, convolutional neural networks (CNN), recurrent neural networks (RNN), generative adversarial networks (GAN), or naive Bayes, and can perform various processes, but is not limited to these examples. Also, the AI ​​may be an AI agent. Furthermore, when the processing of each part described above is performed by the AI, the processing may be performed by the AI ​​in part or in whole, but is not limited to this example. Also, processing performed by an AI including a generative AI may be replaced by rule-based processing, and rule-based processing may be replaced by processing performed by an AI including a generative AI.

[0177] The data processing system 210 according to the second embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 210 is performed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart glasses 214, but it may also be performed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart glasses 214. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information necessary for processing from the smart glasses 214 or an external device, and the smart glasses 214 acquires or collects information necessary for processing from the data processing device 12 or an external device.

[0178] Each of the multiple elements described above, including the proxy answering unit, analysis unit, transfer unit, monitoring unit, and list management unit, is implemented in at least one of the smart glasses 214 and the data processing unit 12. For example, the proxy answering unit is implemented by the control unit 46A of the smart glasses 214, and the AI ​​answers the call on behalf of the caller when an incoming call is received. The analysis unit is implemented by the identification processing unit 290 of the data processing unit 12, and analyzes the content of the conversation to determine the risk of fraud. The transfer unit is implemented by the control unit 46A of the smart glasses 214, and transfers the call if it is determined that the risk of fraud is low. The monitoring unit is implemented by the identification processing unit 290 of the data processing unit 12, and monitors the conversation even while a call is in progress, and issues a warning if signs of fraud are detected. The list management unit is implemented by the identification processing unit 290 of the data processing unit 12, and manages telephone numbers in the whitelist, graylist, and blacklist. The correspondence between each unit and the device or control unit is not limited to the example described above, and various changes are possible.

[0179] [Third Embodiment] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0180] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0181] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN and / or LAN.

[0182] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0183] The microphone 238 receives voice signals from the user and accepts instructions from the user. The microphone 238 captures the voice signals from the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0184] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, which captures images of the area around the user (for example, an imaging range defined by a field of view equivalent to the field of vision of a typical healthy person).

[0185] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0186] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0187] The processor 28 reads a 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 acting as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0188] Storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotions using the emotion identification model 59 and perform identification processing using the user's emotions. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotions, including but not limited to these examples. Furthermore, emotion estimation and prediction also include, for example, emotion analysis.

[0189] In the headset terminal 314, specific processing is performed by the processor 46. The storage 50 stores a specific program 60. The processor 46 reads the specific program 60 from the storage 50 and executes the read specific program 60 on the RAM 48. The specific processing is realized by the processor 46 acting as a control unit 46A according to the specific program 60 executed on the RAM 48. The headset terminal 314 also has a data generation model 58 and an emotion identification model 59, similar to the data generation model and emotion identification model 59, and can perform processing similar to that of the specific processing unit 290 using these models.

[0190] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains processing results (such as prediction results) using the data generation model 58 by communicating with the server device that has the data generation model 58. Also, the data processing device 12 may be a server device or a terminal device owned by the user (for example, a mobile phone, robot, home appliance, etc.).

[0191] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0192] The data generation model 58 is a so-called generative AI. An example of a data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and inference data such as audio data representing speech, text data representing text, and image data representing images (e.g., still image data or video data). The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference result in one or more data formats such as audio data, text data, and image data. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the specific processing described above using the data generation model 58. The data generation model 58 may be a fine-tuned model that outputs inference results from prompts that do not contain instructions, in which case the data generation model 58 can output inference results from prompts that do not contain instructions. In the data processing device 12, etc., there are multiple types of data generation models 58, and the data generation model 58 includes AI other than generative AI. AI other than generative AI includes, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVM), k-means clustering, convolutional neural networks (CNN), recurrent neural networks (RNN), generative adversarial networks (GAN), or naive Bayes, and can perform various processes, but is not limited to these examples. Also, the AI ​​may be an AI agent. Furthermore, when the processing of each part described above is performed by the AI, the processing may be performed by the AI ​​in part or in whole, but is not limited to this example. Also, processing performed by an AI including a generative AI may be replaced by rule-based processing, and rule-based processing may be replaced by processing performed by an AI including a generative AI.

[0193] The data processing system 310 according to the third embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 310 is performed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the headset terminal 314, but may also be performed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the headset terminal 314. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information necessary for processing from the headset terminal 314 or an external device, and the headset terminal 314 acquires or collects information necessary for processing from the data processing device 12 or an external device.

[0194] Each of the multiple elements described above, including the proxy answering unit, analysis unit, transfer unit, monitoring unit, and list management unit, is implemented in at least one of the headset terminal 314 and the data processing unit 12. For example, the proxy answering unit is implemented by the control unit 46A of the headset terminal 314, and the AI ​​answers on behalf of the caller when an incoming call is received. The analysis unit is implemented by the identification processing unit 290 of the data processing unit 12, and analyzes the content of the conversation to determine the risk of fraud. The transfer unit is implemented by the control unit 46A of the headset terminal 314, and transfers the call if it is determined that the risk of fraud is low. The monitoring unit is implemented by the identification processing unit 290 of the data processing unit 12, and monitors the conversation even during a call, and issues a warning if signs of fraud are detected. The list management unit is implemented by the identification processing unit 290 of the data processing unit 12, and manages telephone numbers in the whitelist, graylist, and blacklist. The correspondence between each unit and the device or control unit is not limited to the example described above, and various changes are possible.

[0195] [Fourth Embodiment] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0196] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0197] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN and / or LAN.

[0198] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0199] The microphone 238 receives voice signals from the user and accepts instructions from the user. The microphone 238 captures the voice signals from the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0200] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS image sensor or CCD image sensor, which captures images of the area around the user (for example, an imaging range defined by a field of view equivalent to the field of vision of a typical healthy person).

[0201] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0202] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. The robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0203] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0204] The processor 28 reads a 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 acting as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0205] Storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate the user's emotions using the emotion identification model 59 and perform identification processing using the user's emotions. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotions, including but not limited to these examples. Furthermore, emotion estimation and prediction also include, for example, emotion analysis.

[0206] In robot 414, specific processing is performed by processor 46. A specific program 60 is stored in storage 50. Processor 46 reads the specific program 60 from storage 50 and executes it on RAM 48. The specific processing is achieved by processor 46 acting as a control unit 46A according to the specific program 60 executed on RAM 48. Robot 414 also has data generation model 58 and emotion identification model 59, similar to those of the robot, and can perform processing similar to that of the specific processing unit 290 using these models.

[0207] Furthermore, other devices besides the data processing device 12 may also have the data generation model 58. For example, a server device may have the data generation model 58. In this case, the data processing device 12 obtains processing results (such as prediction results) using the data generation model 58 by communicating with the server device that has the data generation model 58. Also, the data processing device 12 may be a server device or a terminal device owned by the user (for example, a mobile phone, robot, home appliance, etc.).

[0208] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0209] The data generation model 58 is a so-called generative AI. An example of a data generation model 58 is a generative AI such as ChatGPT. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and inference data such as audio data representing speech, text data representing text, and image data representing images (e.g., still image data or video data). The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference result in one or more data formats such as audio data, text data, and image data. The data generation model 58 includes, for example, text generation AI, image generation AI, and multimodal generation AI. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specific processing unit 290 performs the specific processing described above using the data generation model 58. The data generation model 58 may be a fine-tuned model that outputs inference results from prompts that do not contain instructions, in which case the data generation model 58 can output inference results from prompts that do not contain instructions. In the data processing device 12, etc., there are multiple types of data generation models 58, and the data generation model 58 includes AI other than generative AI. AI other than generative AI includes, for example, linear regression, logistic regression, decision trees, random forests, support vector machines (SVM), k-means clustering, convolutional neural networks (CNN), recurrent neural networks (RNN), generative adversarial networks (GAN), or naive Bayes, and can perform various processes, but is not limited to these examples. Also, the AI ​​may be an AI agent. Furthermore, when the processing of each part described above is performed by the AI, the processing may be performed by the AI ​​in part or in whole, but is not limited to this example. Also, processing performed by an AI including a generative AI may be replaced by rule-based processing, and rule-based processing may be replaced by processing performed by an AI including a generative AI.

[0210] The data processing system 410 according to the fourth embodiment performs the same processing as the data processing system 10 according to the first embodiment. The processing by the data processing system 410 is performed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the robot 414, but it may also be performed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the robot 414. In addition, the specific processing unit 290 of the data processing device 12 acquires or collects information necessary for processing from the robot 414 or an external device, and the robot 414 acquires or collects information necessary for processing from the data processing device 12 or an external device.

[0211] Each of the multiple elements described above, including the proxy answering unit, analysis unit, transfer unit, monitoring unit, and list management unit, is implemented by at least one of the robot 414 and the data processing unit 12. For example, the proxy answering unit is implemented by the control unit 46A of the robot 414, and the AI ​​answers on behalf of the caller when an incoming call is received. The analysis unit is implemented by the identification processing unit 290 of the data processing unit 12, and analyzes the content of the conversation to determine the risk of fraud. The transfer unit is implemented by the control unit 46A of the robot 414, and transfers the call if it is determined that the risk of fraud is low. The monitoring unit is implemented by the identification processing unit 290 of the data processing unit 12, and monitors the conversation even while a call is in progress, and issues a warning if signs of fraud are detected. The list management unit is implemented by the identification processing unit 290 of the data processing unit 12, and manages telephone numbers on the whitelist, graylist, and blacklist. The correspondence between each unit and the device or control unit is not limited to the example described above, and various changes are possible.

[0212] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0213] Figure 9 shows the emotion map 400, in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0214] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0215] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0216] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, and motorcycles, emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated based, for example, on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0217] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0218] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0219] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing method for the specific process may be used, which includes computer 22 and multiple other computers.

[0220] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0221] 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.

[0222] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0223] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0224] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0225] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0226] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0227] Furthermore, although the above-described examples were divided into four embodiments, some or all of these embodiments may be combined. Also, the smart device 14, smart glasses 214, headset terminal 314, and robot 414 are just examples, and they may be combined, or other devices may be used. Also, although the above-described examples were divided into two embodiments, Embodiment 1 and Embodiment 2, these may be combined.

[0228] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and other things that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0229] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[0230] (Note 1) The AI-powered answering unit handles incoming calls, An analysis unit analyzes the content of the conversation that was answered by the proxy response unit and determines the risk of fraud, A call forwarding unit that forwards a call if the analysis unit determines that the risk of fraud is low, The monitoring unit monitors the conversation during the call and issues a warning if signs of fraud are detected. It includes a list management unit that manages phone numbers on a whitelist, graylist, and blacklist. A system characterized by the following features. (Note 2) It includes a questioning section that verifies the reliability of the sender through specific questions. The system described in Appendix 1, characterized by the features described herein. (Note 3) It is equipped with a rejection function that rejects calls if it determines that there is a high risk of fraud. The system described in Appendix 1, characterized by the features described herein. (Note 4) It features a warning unit that issues a warning if signs of fraud are detected during a call. The system described in Appendix 1, characterized by the features described herein. (Note 5) It includes a list update unit that updates the list based on the proxy response result. The system described in Appendix 1, characterized by the features described herein. (Note 6) The aforementioned proxy response unit is It estimates the user's emotions and adjusts the tone and wording of the proxy response based on the estimated emotions. The system described in Appendix 1, characterized by the features described herein. (Note 7) The aforementioned proxy response unit is The system analyzes the characteristics of the caller's voice upon incoming calls and compares it with past data to evaluate its reliability. The system described in Appendix 1, characterized by the features described herein. (Note 8) The aforementioned proxy response unit is When an agent answers the call, the caller's background noise is analyzed to determine the possibility of fraud. The system described in Appendix 1, characterized by the features described herein. (Note 9) The aforementioned proxy response unit is It estimates the user's emotions and customizes the content of the proxy response based on the estimated user emotions. The system described in Appendix 1, characterized by the features described herein. (Note 10) The aforementioned proxy response unit is When answering on behalf of someone else, the response content is adjusted to take into account the caller's geographical location. The system described in Appendix 1, characterized by the features described herein. (Note 11) The aforementioned proxy response unit is When answering a call on behalf of another party, the response is optimized by referring to the caller's past call history. The system described in Appendix 1, characterized by the features described herein. (Note 12) The aforementioned analysis unit, It estimates the user's emotions and adjusts the accuracy of the analysis based on the estimated user emotions. The system described in Appendix 1, characterized by the features described herein. (Note 13) The aforementioned analysis unit, During the analysis, the tone and speed of the caller's voice are analyzed to assess the likelihood of fraud. The system described in Appendix 1, characterized by the features described herein. (Note 14) The aforementioned analysis unit, During analysis, specific phrases and expressions used by the caller are detected to determine the risk of fraud. The system described in Appendix 1, characterized by the features described herein. (Note 15) The aforementioned analysis unit, It estimates the user's emotions and adjusts how the analysis results are displayed based on the estimated emotions. The system described in Appendix 1, characterized by the features described herein. (Note 16) The aforementioned analysis unit, The accuracy of the analysis is improved by referencing the caller's past call history during the analysis process. The system described in Appendix 1, characterized by the features described herein. (Note 17) The aforementioned analysis unit, The analysis results are corrected by considering the geographical location information of the caller during the analysis process. The system described in Appendix 1, characterized by the features described herein. (Note 18) The transfer unit is, It estimates the user's emotions and adjusts the timing of transfers based on the estimated emotions. The system described in Appendix 1, characterized by the features described herein. (Note 19) The transfer unit is, The caller's reliability is re-evaluated during forwarding to determine whether or not to forward the message. The system described in Appendix 1, characterized by the features described herein. (Note 20) The transfer unit is, The system optimizes the transfer by considering the recipient's current status during the transfer process. The system described in Appendix 1, characterized by the features described herein. (Note 21) The transfer unit is, It estimates the user's emotions and determines the priority of transfers based on the estimated user emotions. The system described in Appendix 1, characterized by the features described herein. (Note 22) The transfer unit is, When transferring a call, the system checks the caller's past call history to determine whether or not to transfer the call. The system described in Appendix 1, characterized by the features described herein. (Note 23) The transfer unit is, The timing of the transfer is adjusted to take into account the recipient's schedule information during the transfer process. The system described in Appendix 1, characterized by the features described herein. (Note 24) The aforementioned monitor unit is It estimates the user's emotions and adjusts the accuracy of monitoring based on the estimated user emotions. The system described in Appendix 1, characterized by the features described herein. (Note 25) The aforementioned monitor unit is The system analyzes the caller's tone and speed of voice in real time during a call to detect signs of fraud. The system described in Appendix 1, characterized by the features described herein. (Note 26) The aforementioned monitor unit is The system detects specific phrases and expressions in real time during a call and assesses the risk of fraud. The system described in Appendix 1, characterized by the features described herein. (Note 27) The aforementioned monitor unit is It estimates the user's emotions and adjusts the monitoring frequency based on the estimated emotions. The system described in Appendix 1, characterized by the features described herein. (Note 28) The aforementioned monitor unit is By referencing the caller's past call history during a call, the accuracy of monitoring can be improved. The system described in Appendix 1, characterized by the features described herein. (Note 29) The aforementioned monitor unit is During a call, the monitoring results are corrected by considering the caller's geographical location information. The system described in Appendix 1, characterized by the features described herein. (Note 30) The aforementioned list management unit, It estimates the user's sentiment and adjusts the list update frequency based on the estimated user sentiment. The system described in Appendix 1, characterized by the features described herein. (Note 31) The aforementioned list management unit, When managing lists, refer to the caller's past call history to improve the accuracy of the lists. The system described in Appendix 1, characterized by the features described herein. (Note 32) The aforementioned list management unit, When managing lists, the content of the list is optimized by considering the geographical location information of the sender. The system described in Appendix 1, characterized by the features described herein. (Note 33) The aforementioned list management unit, It estimates the user's sentiment and determines the priority of the list based on the estimated user sentiment. The system according to appended note 1, characterized in that (Appended note 34) The list management unit Analyzes the characteristics of the caller's voice during list management to improve the accuracy of the list The system according to appended note 1, characterized in that (Appended note 35) The list management unit Detects specific phrases or turns of phrase used by the caller during list management and updates the content of the list The system according to appended note 1, characterized in that (Appended note 36) The question unit Estimates the user's emotion and adjusts the content and tone of the question based on the estimated user emotion The system according to appended note 2, characterized in that (Appended note 37) The question unit Analyzes the tone and speed of the caller's voice during question asking to evaluate the reliability The system according to appended note 2, characterized in that (Appended note 38) The question unit Detects specific phrases or turns of phrase used by the caller during question asking and determines the reliability The system according to appended note 2, characterized in that (Appended note 39) The question unit Estimates the user's emotion and adjusts the order of the questions based on the estimated user emotion The system according to appended note 2, characterized in that (Appended note 40) The question unit Refers to the caller's past call history during question asking to optimize the question content The system according to appended note 2, characterized in that (Appended note 41) The question unit Considers the geographical location information of the caller during question asking to adjust the question content The system according to appended note 2, characterized in that (Appended note 42) The rejecting part is, It estimates the user's emotions and adjusts the timing of rejection based on those estimated emotions. The system described in Appendix 3, characterized by the features described herein. (Note 43) The rejecting part is, The system detects specific phrases and expressions used by callers when rejecting a request, thereby strengthening the basis for the rejection. The system described in Appendix 3, characterized by the features described herein. (Note 44) The rejecting part is, It estimates the user's emotions and determines the priority of rejections based on the estimated user emotions. The system described in Appendix 3, characterized by the features described herein. (Note 45) The rejecting part is, When rejecting a call, the system checks the caller's past call history to determine whether or not to reject it. The system described in Appendix 3, characterized by the features described herein. (Note 46) The rejecting part is, When rejecting a request, the reason for rejection should be adjusted to take into account the caller's geographical location. The system described in Appendix 3, characterized by the features described herein. (Note 47) The aforementioned warning unit is It estimates the user's emotions and adjusts the content and tone of warnings based on those emotions. The system described in Appendix 4, characterized by the features described herein. (Note 48) The aforementioned warning unit is The system analyzes the tone and speed of the caller's voice during a warning to optimize the timing of the warning. The system described in Appendix 4, characterized by the features described herein. (Note 49) The aforementioned warning unit is The system detects specific phrases and expressions used by callers when issuing warnings, thereby strengthening the basis for those warnings. The system described in Appendix 4, characterized by the features described herein. (Note 50) The aforementioned warning unit is Estimate the user's emotion and determine the warning priority based on the estimated user emotion The system according to Appendix 4, characterized by the above. (Appendix 51) The warning unit Optimizes the warning content by referring to the caller's past call history at the time of warning The system according to Appendix 4, characterized by the above. (Appendix 52) The warning unit Corrects the warning content by considering the caller's geographical location information at the time of warning The system according to Appendix 4, characterized by the above. (Appendix 53) The list update unit Estimates the user's emotion and adjusts the update frequency of the list based on the estimated user emotion The system according to Appendix 5, characterized by the above. (Appendix 54) The list update unit Analyzes the tone and speed of the caller's voice at the time of list update to improve the accuracy of the list The system according to Appendix 5, characterized by the above. <000096​​​​​​​​​​​​​​​​​​​​​​​​​​​​When updating the list, the list content is optimized by considering the geographical location information of the sender. The system described in Appendix 5, characterized by the features described herein. [Explanation of symbols]

[0231] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots

Claims

1. A proxy answering unit where AI answers calls on behalf of the caller, An analysis unit analyzes the content of the conversation that was answered by the proxy response unit and determines the risk of fraud, A call forwarding unit that forwards a call if the analysis unit determines that the risk of fraud is low, The monitoring unit monitors the conversation during the call and issues a warning if signs of fraud are detected. It includes a list management unit that manages phone numbers on a whitelist, graylist, and blacklist. A system characterized by the following features.

2. It includes a questioning section that verifies the reliability of the sender through specific questions. The system according to feature 1.

3. It is equipped with a rejection function that rejects calls if it determines that there is a high risk of fraud. The system according to feature 1.

4. It features a warning unit that issues a warning if signs of fraud are detected during a call. The system according to feature 1.

5. It includes a list update unit that updates the list based on the proxy response result. The system according to feature 1.

6. The aforementioned proxy response unit is It estimates the user's emotions and adjusts the tone and wording of the proxy response based on the estimated emotions. The system according to feature 1.

7. The aforementioned proxy response unit is The system analyzes the characteristics of the caller's voice upon incoming calls and compares it with past data to evaluate its reliability. The system according to feature 1.

8. The aforementioned proxy response unit is When an agent answers the call, the caller's background noise is analyzed to determine the possibility of fraud. The system according to feature 1.

9. The aforementioned proxy response unit is It estimates the user's emotions and customizes the content of the proxy response based on the estimated user emotions. The system according to feature 1.

10. The aforementioned proxy response unit is When answering on behalf of someone else, the response content is adjusted to take into account the caller's geographical location. The system according to feature 1.

Images