system

Abstract

The system according to the embodiment aims to detect fraudulent calls in real time and encourage victims to take countermeasures. [Solution] The system according to the embodiment comprises a learning unit, an analysis unit, a guidance unit, a disconnection unit, a text conversion unit, and a transmission unit. The learning unit learns the characteristics of fraudulent talk. The analysis unit analyzes the content of the call in real time based on the characteristics learned by the learning unit. The guidance unit plays guidance to the victim when fraud is detected by the analysis unit. The disconnection unit disconnects the call after the guidance unit plays the guidance. The text conversion unit converts the content of the call analyzed by the analysis unit into text. The transmission unit sends the content converted into text by the text conversion unit via SMS.

Description

[Technical Field]

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

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

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] Conventional technology does not adequately detect and counter fraudulent calls, and there is room for improvement to prevent damage.

[0005] The system according to the embodiment aims to detect fraudulent calls in real time and encourage victims to take countermeasures. [Means for solving the problem]

[0006] The system according to the embodiment includes a learning unit, an analysis unit, a guidance unit, a disconnection unit, a text conversion unit, and a transmission unit. The learning unit learns the characteristics of fraudulent talk. The analysis unit analyzes the content of the call in real time based on the characteristics learned by the learning unit. The guidance unit plays guidance to the victim when fraud is detected by the analysis unit. The disconnection unit disconnects the call after the guidance unit plays the guidance. The text conversion unit converts the content of the call analyzed by the analysis unit into text. The transmission unit sends the content converted into text by the text conversion unit via SMS. [Effects of the Invention]

[0007] The system according to the embodiment can detect fraudulent calls in real time and urge victims to take countermeasures. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9]1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. DETAILED DESCRIPTION OF THE INVENTION

[0009] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0010] First, the terms used in the following description will be explained.

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

[0012] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

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

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

[0015] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

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

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

[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, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

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

[0020] The reception device 38 includes a touch panel 38A and a microphone 38B, and receives user input. The touch panel 38A detects contact with a pointer (for example, a pen or a finger) to receive user input by the touch of the pointer. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 (see FIG. 2) acquires the data indicating the user input.

[0021] Output device 40 includes a display 40A and a speaker 40B, and presents data to a user by outputting the data in a form of expression that the user can perceive (e.g., audio and / or text). Display 40A displays visible information such as text and images in accordance with instructions from processor 46. Speaker 40B outputs audio in accordance with instructions from processor 46. Camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0022] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

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

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

[0026] In the smart device 14, the specific processing is performed by the processor 46. The storage 50 stores a specific processing program 60. The specific processing program 60 is used together 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 the control unit 46A in accordance with the specific processing program 60 executed on the RAM 48. Note that the smart device 14 has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can also perform processing similar to that of the specific processing unit 290 using these models.

[0027] Note that a device other than the data processing device 12 may 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 a processing result (prediction result, etc.) using the data generation model 58 by communicating with the server device having the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device owned by a user (e.g., a mobile phone, a robot, a home appliance, etc.). Next, an example of processing by the data processing system 10 according to the first embodiment will be described.

[0028] (Example 1) A fraud prevention system according to an embodiment of the present invention learns the characteristics of fraudulent conversations and compares them during a call. This fraud prevention system uses a generation AI to learn the characteristics of fraudulent conversations, analyzes the content of the call in real time, and, if fraud is detected, plays a message to the victim and disconnects the call. Furthermore, the call content is converted into text and sent via SMS, which can be used as information for consultation with the police or family. This system can deter successful fraud and instill in consumers the idea of ​​preventing "it's my son" scams. For example, a generation AI is used to learn the characteristics of fraudulent conversations. Data on past fraudulent conversations is collected and trained on the generation AI. The call content is then analyzed in real time. The generation AI analyzes the call content and compares it with the characteristics of fraudulent conversations. If fraud is detected, a message is played to the victim and the call is disconnected. For example, a message such as "This may be a scam. Please hang up." is played. The call content is then converted into text and sent via SMS, which can be used as information for consultation with the police or family. For example, if the content of a call is something like "Please transfer money," the content will be converted to text and sent via SMS. This system will deter successful fraud and establish in consumers' minds the image of preventing "it's my son" fraud. This allows the fraud prevention system to learn the characteristics of fraudulent conversations, detect fraud in real time during the call, provide guidance to the victim, disconnect the call, and convert the content of the call into text and send it, thereby deterring successful fraud and protecting victims.

[0029] The fraud prevention system according to the embodiment includes a learning unit, an analysis unit, a guidance unit, a disconnection unit, a text conversion unit, and a transmission unit. The learning unit learns the characteristics of fraudulent talk. For example, the learning unit collects data on past fraudulent talk and trains the generation AI. The generation AI learns the characteristics of fraudulent talk using a text generation AI (e.g., LLM) or a multimodal generation AI. The analysis unit analyzes the content of the call in real time based on the characteristics learned by the learning unit. For example, the analysis unit analyzes the content of the call in real time and compares it with the characteristics of the fraudulent talk. The generation AI analyzes the content of the call and compares it with the characteristics of the fraudulent talk. The guidance unit plays guidance to the victim when fraud is detected by the analysis unit. For example, the guidance unit plays guidance such as "There is a possibility of fraud. Please hang up the call." The disconnection unit disconnects the call after the guidance unit has played the guidance. For example, the disconnection unit disconnects the call after the guidance has been played. The text conversion unit converts the content of the call analyzed by the analysis unit into text. The text conversion unit, for example, converts the contents of the call into text. The generation AI converts the contents of the call into text using voice recognition technology. The transmission unit sends the contents converted into text by the text conversion unit via SMS. The transmission unit, for example, sends the contents converted into text via SMS. As a result, the fraud prevention system according to the embodiment can learn the characteristics of fraudulent conversations, detect fraud in real time during a call, provide guidance to the victim, disconnect the call, and convert the contents of the call into text and send it, thereby preventing successful fraud and protecting the victim.

[0030] The learning unit collects data on past fraudulent talk and allows the generation AI to learn from it. The learning unit, for example, collects data on past fraudulent talk and allows the generation AI to learn from it. The generation AI learns the characteristics of fraudulent talk using text generation AI (e.g., LLM) or multimodal generation AI. The generation AI, for example, collects data on past fraudulent talk and learns the characteristics of fraudulent talk based on that data. The generation AI, for example, analyzes data on fraudulent talk and learns specific phrases and speaking patterns. The generation AI, for example, analyzes data on fraudulent talk and learns specific voice tones and rhythms. In this way, by collecting data on past fraudulent talk and allowing the generation AI to learn, the characteristics of fraudulent talk can be learned more accurately.

[0031] The analysis unit can analyze the content of the call in real time and compare it with characteristics of fraudulent talk. The analysis unit, for example, analyzes the content of the call in real time and compares it with characteristics of fraudulent talk. The generation AI analyzes the content of the call and compares it with characteristics of fraudulent talk. The generation AI, for example, analyzes the content of the call in real time and compares it with characteristics of fraudulent talk. The generation AI, for example, analyzes the content of the call and compares it with specific phrases or speaking patterns. The generation AI, for example, analyzes the content of the call and compares it with specific voice tones and rhythms. This improves the accuracy of fraud detection by analyzing the content of the call in real time and comparing it with characteristics of fraudulent talk.

[0032] The guidance unit can play guidance such as "Fraud has been detected. Please hang up." when fraud is detected. For example, the guidance unit plays guidance such as "Fraud has been detected. Please hang up." when fraud is detected. For example, the guidance unit plays a voice message when fraud is detected. For example, the guidance unit plays a text message when fraud is detected. For example, the guidance unit plays both voice and text messages when fraud is detected. This allows victims to respond quickly by playing appropriate guidance when fraud is detected.

[0033] The disconnection unit can disconnect the call after the guidance unit has played the guidance. The disconnection unit, for example, disconnects the call after the guidance has been played. The disconnection unit, for example, automatically disconnects the call after the guidance has been played. The disconnection unit, for example, manually disconnects the call after the guidance has been played. The disconnection unit, for example, disconnects the call after a certain time has elapsed after the guidance has been played. In this way, by disconnecting the call after the guidance has been played, it is possible to prevent fraud.

[0034] The text conversion unit can convert the contents of a call into text. The text conversion unit, for example, converts the contents of a call into text. The text conversion unit, for example, converts the contents of a call into text using voice recognition technology. The text conversion unit, for example, converts the contents of a call into text in real time. The text conversion unit, for example, records the contents of a call and converts them into text later. By converting the contents of a call into text, the contents can be more easily checked later.

[0035] The transmitting unit can transmit the content converted into text by the text conversion unit via SMS. The transmitting unit, for example, transmits the text content via SMS. The transmitting unit, for example, transmits the text content to the police. The transmitting unit, for example, transmits the text content to a family member. The transmitting unit, for example, transmits the text content to multiple destinations. In this way, by transmitting the text content via SMS, the content can be used as material for consultation with the police or family members.

[0036] During learning, the learning unit can analyze the voice patterns of fraudulent talk and learn voice features. For example, the generation AI analyzes the voice data of fraudulent talk and learns specific voice patterns (e.g., voice tone and rhythm). For example, the generation AI analyzes the voice data of fraudulent talk and learns the emphasis of specific words and pauses. For example, the learning unit analyzes the voice data of fraudulent talk and learns specific emotional expressions (e.g., impatience and relief). In this way, by learning the voice patterns of fraudulent talk, fraud detection based on voice features becomes possible.

[0037] During learning, the learning unit can extract highly accurate features based on contextual information about the fraudulent conversation. For example, the generation AI in the learning unit analyzes the context of the fraudulent conversation and extracts features related to a specific scenario (e.g., a family emergency). For example, the generation AI in the learning unit analyzes the context of the fraudulent conversation and extracts specific phrases and phrasing patterns. For example, the generation AI in the learning unit analyzes the context of the fraudulent conversation and extracts specific patterns of conversation flow and development. This enables more accurate fraud detection by taking into account the contextual information about the fraudulent conversation.

[0038] During learning, the learning unit can take into account the regional characteristics of fraudulent speech and learn the characteristics of each region. For example, the generation AI collects data on fraudulent speech that occurs frequently in a specific region and learns the characteristics unique to that region. For example, the generation AI learns the characteristics of fraudulent speech by taking into account differences in language and dialect in each region. For example, the generation AI learns the characteristics of fraudulent speech by taking into account the culture and customs of each region. In this way, by taking into account regional characteristics, the system can learn the characteristics of fraudulent speech in each region and detect frauds that are unique to that region.

[0039] During learning, the learning unit can analyze the characteristics of fraudulent talk by time of day and learn the patterns for each time of day. For example, the generation AI classifies fraudulent talk data by time of day and learns the characteristics that are common in specific time periods. For example, the generation AI learns the patterns of fraudulent talk by time of day (for example, the difference between daytime and nighttime). For example, the generation AI analyzes the success rate of fraudulent talk during specific time periods and learns the characteristics unique to that time period. In this way, by learning the characteristics for each time period, the accuracy of detecting fraudulent talk during specific time periods is improved.

[0040] During the analysis, the analysis unit can analyze background sounds of the call and identify elements that increase the likelihood of fraud. For example, the generation AI analyzes the background sounds of the call and identifies specific sounds (e.g., noise or silence) as elements that increase the likelihood of fraud. For example, the generation AI analyzes the background sounds of the call and identifies specific environmental sounds (e.g., office sounds or home sounds) as elements that increase the likelihood of fraud. For example, the analysis unit analyzes the background sounds of the call and identifies specific sound patterns (e.g., intermittent sounds or continuous sounds) as elements that increase the likelihood of fraud. In this way, by analyzing the background sounds of the call, elements that increase the likelihood of fraud are identified, improving the accuracy of fraud detection.

[0041] During analysis, the analysis unit can analyze the speed and tone of the call and compare them with characteristics of fraudulent talk. For example, the generation AI analyzes the speed of the call and compares a specific speed (e.g., fast speaking or slow speaking) with characteristics of fraudulent talk. For example, the generation AI analyzes the tone of the call and compares a specific tone (e.g., emphasized tone or flat tone) with characteristics of fraudulent talk. For example, the generation AI analyzes the combination of speed and tone of the call and compares a specific pattern with characteristics of fraudulent talk. In this way, by analyzing the speed and tone of the call, it can be compared with characteristics of fraudulent talk, improving the accuracy of fraud detection.

[0042] During analysis, the analysis unit can identify region-specific fraudulent talk by taking into account the geographical information of the call. For example, the analysis unit uses a generation AI to analyze the geographical information of the call and identify fraudulent talk that occurs frequently in a specific region. For example, the analysis unit uses a generation AI to analyze the geographical information of the call and identify fraudulent talk by taking into account differences in language and dialect between regions. For example, the analysis unit uses a generation AI to analyze the geographical information of the call and identify fraudulent talk by taking into account the culture and customs of each region. In this way, by taking into account the geographical information of the call, region-specific fraudulent talk can be identified, improving the accuracy of fraud detection.

[0043] During analysis, the analysis unit can evaluate the likelihood of fraud based on contextual information of the call. For example, the generation AI analyzes the context of the call and evaluates the likelihood of fraud related to a specific scenario (e.g., a family emergency). For example, the generation AI analyzes the context of the call and evaluates the likelihood of fraud by referring to specific phrases and phrasing patterns. For example, the generation AI analyzes the context of the call and evaluates the likelihood of fraud by referring to specific patterns of conversation flow and development. This allows for a more accurate evaluation of the likelihood of fraud by referring to the contextual information of the call.

[0044] When providing guidance, the guidance unit can refer to the user's past call history to provide optimal guidance. For example, if the user has a history of falling for fraudulent talk in the past, the guidance unit provides strong warning guidance. For example, if the user has a history of ignoring fraudulent talk in the past, the guidance unit provides normal warning guidance. For example, if the user has a history of suspecting fraudulent talk in the past, the guidance unit provides mild warning guidance. In this way, by referring to the user's past call history, more appropriate guidance can be provided.

[0045] The guidance unit can generate different guidance messages depending on the content of the call when providing guidance. For example, if the content of the call is a request for money, the guidance unit plays the guidance, "This may be a scam. Please hang up." For example, if the content of the call is a request for personal information, the guidance unit plays the guidance, "This may be a scam. Please hang up." For example, if the content of the call is a report of an emergency, the guidance unit plays the guidance, "This may be a scam. Please hang up." In this way, by generating guidance messages depending on the content of the call, more effective guidance can be provided.

[0046] The guidance unit can change the way the guidance is presented depending on the age and gender of the user when providing guidance. For example, the guidance unit provides guidance in a slow tone of voice for an elderly person. For example, the guidance unit provides guidance in a normal tone of voice for a young person. For example, the guidance unit provides guidance in a gentle tone of voice for a female person. This allows for more effective guidance to be provided by providing guidance according to the age and gender of the user.

[0047] The guidance unit can adjust the frequency of the guidance based on the length of the call when providing the guidance. For example, if the call continues for a long time, the guidance unit plays the guidance at regular intervals to warn of the possibility of fraud. For example, if the call is short, the guidance unit plays the guidance only once. For example, if the call is medium in length, the guidance unit plays the guidance at appropriate intervals. This allows for more effective guidance to be provided by adjusting the frequency of the guidance based on the length of the call.

[0048] The disconnection unit evaluates the importance of the call at the time of disconnection, and allows important calls to be continued with priority. For example, the disconnection unit allows the call to continue if the content of the call is about an emergency. For example, the disconnection unit allows the call to continue if the content of the call is about important business. For example, the disconnection unit allows the call to continue if the content of the call is about an important family topic. In this way, by evaluating the importance of the call, important calls can be continued with priority.

[0049] The disconnection unit can analyze the content of the call when disconnecting the call and disconnect the call only if there is a high possibility of fraud. For example, the disconnection unit disconnects the call if the content of the call is a request for money. For example, the disconnection unit disconnects the call if the content of the call is a request for personal information. For example, the disconnection unit disconnects the call if the content of the call is a report of an emergency. In this way, by analyzing the content of the call, it is possible to disconnect the call only if there is a high possibility of fraud.

[0050] When disconnecting the call, the disconnection unit can make a decision to disconnect the call by taking into account attribute information of the other party of the call. For example, the disconnection unit disconnects the call if the other party is unknown. For example, the disconnection unit continues the call if the other party is a trustworthy person. For example, the disconnection unit disconnects the call if the other party is suspicious. In this way, by taking into account the attribute information of the other party of the call, it is possible to disconnect the call with a more appropriate decision.

[0051] When disconnecting a call, the disconnection unit can determine the priority of disconnection based on the time period of the call. For example, if the call is made at night, the disconnection unit disconnects the call with priority. For example, if the call is made during the day, the disconnection unit disconnects the call with normal priority. For example, if the call is made in the early morning, the disconnection unit disconnects the call with caution. In this way, by determining the priority of disconnection based on the time period of the call, it is possible to disconnect the call at a more appropriate time.

[0052] The text conversion unit can evaluate the voice quality of the call during text conversion and select the optimal text conversion method. For example, if the voice quality of the call is high, the text conversion unit selects a normal text conversion method. For example, if the voice quality of the call is low, the text conversion unit performs noise reduction before converting to text. For example, if the voice quality of the call is medium, the text conversion unit performs appropriate noise reduction before converting to text. In this way, by evaluating the voice quality of the call, the optimal text conversion method can be selected and more accurate text can be generated.

[0053] The text conversion unit can generate more accurate text by taking into account contextual information of the call during text conversion. For example, the text conversion unit analyzes the context of the call and generates text based on a specific scenario. For example, the text conversion unit analyzes the context of the call and generates text by taking into account specific phrases and expressions. For example, the text conversion unit analyzes the context of the call and generates text by taking into account specific conversation flow and developments. In this way, by taking into account contextual information of the call, more accurate text can be generated.

[0054] During the text conversion, the text conversion unit can generate text that supports multiple languages ​​by taking into account language information of the call. For example, if the call is in English, the text conversion unit generates text in English. For example, if the call is in Japanese, the text conversion unit generates text in Japanese. For example, if the call is in multiple languages, the text conversion unit generates text corresponding to each language. In this way, by taking into account language information of the call, text that supports multiple languages ​​can be generated.

[0055] The text conversion unit can adjust the format of the text based on the content of the call during text conversion. For example, if the content of the call is a request for money, the text conversion unit generates text in a format that emphasizes important parts. For example, if the content of the call is a request for personal information, the text conversion unit generates text in a format that emphasizes important parts. For example, if the content of the call is a report of an emergency, the text conversion unit generates text in a format that emphasizes important parts. In this way, by adjusting the format of the text based on the content of the call, it is possible to generate text that emphasizes important information.

[0056] The transmitting unit can evaluate the importance of the call at the time of transmission and prioritize the transmission of important content. For example, if the call content is about an emergency, the transmitting unit prioritizes the transmission of text. For example, if the call content is about important business, the transmitting unit prioritizes the transmission of text. For example, if the call content is about an important family topic, the transmitting unit prioritizes the transmission of text. In this way, by evaluating the importance of the call, important content can be prioritized for transmission.

[0057] The sending unit can analyze the content of the call at the time of sending and select an appropriate destination. For example, if the content of the call is a request for money, the sending unit sends a text to the police. For example, if the content of the call is a request for personal information, the sending unit sends a text to a family member. For example, if the content of the call is a report of an emergency, the sending unit sends a text to an appropriate organization. In this way, by analyzing the content of the call, an appropriate destination can be selected, enabling a more effective response.

[0058] When transmitting, the transmitting unit can make a decision on whether to transmit taking into consideration the attribute information of the other party of the call. For example, if the other party of the call is unknown, the transmitting unit transmits the text to the police. For example, if the other party of the call is a trustworthy person, the transmitting unit transmits the text to a family member. For example, if the other party of the call is suspicious, the transmitting unit transmits the text to an appropriate organization. In this way, by taking into consideration the attribute information of the other party of the call, it is possible to transmit the text with a more appropriate decision.

[0059] The sending unit can determine the priority of sending based on the time period of the call when sending. For example, if the call is at night, the sending unit sends the text with priority. For example, if the call is during the day, the sending unit sends the text with normal priority. For example, if the call is early in the morning, the sending unit sends the text carefully. In this way, by determining the priority of sending based on the time period of the call, it is possible to send the text at a more appropriate time.

[0060] The system according to the embodiment is not limited to the above-described example, and various modifications are possible, for example, as follows.

[0061] The fraud prevention system may further include a voice analysis unit that analyzes the user's tone of voice and speaking patterns. For example, if the user is nervous, the voice analysis unit may detect a tendency for the user's voice tone to become higher, and analyze this as a factor that increases the likelihood of fraud. For example, if the user is relaxed, the voice analysis unit may detect that the user's voice tone is stable, and determine that the likelihood of fraud is low. For example, if the user is impatient, the voice analysis unit may detect that the user's speaking pattern becomes faster, and analyze this as a factor that increases the likelihood of fraud. In this way, by analyzing the user's tone of voice and speaking patterns, the accuracy of fraud detection can be improved.

[0062] The analysis unit can further analyze background sounds of calls to identify factors that increase the likelihood of fraud. For example, if the background sounds of a call are quiet, it determines that the likelihood of fraud is high. If the background sounds of a call are noisy, it determines that the likelihood of fraud is low. If the background sounds of a call include specific environmental sounds (e.g., office sounds or home sounds), it identifies the environmental sounds as factors that increase the likelihood of fraud. Thus, by analyzing background sounds of calls, it is possible to identify factors that increase the likelihood of fraud and improve the accuracy of fraud detection.

[0063] The disconnection unit further evaluates the importance of the call and allows important calls to be continued with priority. For example, if the content of the call is about an emergency, the call is allowed to continue. If the content of the call is about important business, the call is allowed to continue. If the content of the call is about an important family topic, the call is allowed to continue. In this way, by evaluating the importance of the call, important calls can be allowed to be continued with priority.

[0064] The sending unit can further analyze the content of the call and select an appropriate destination. For example, if the content of the call is a request for money, a text message can be sent to the police. If the content of the call is a request for personal information, a text message can be sent to a family member. If the content of the call is to report an emergency, a text message can be sent to an appropriate organization. This allows for more effective response by analyzing the content of the call and selecting an appropriate destination.

[0065] The analysis unit can further analyze the speed and tone of the call and compare it with the characteristics of fraudulent talk. For example, if the speed of the call is fast, it is compared with the characteristics of fraudulent talk. If the tone of the call is emphasized, it is compared with the characteristics of fraudulent talk. The combination of the speed and tone of the call is analyzed, and specific patterns are compared with the characteristics of fraudulent talk. In this way, by analyzing the speed and tone of the call, it is possible to compare it with the characteristics of fraudulent talk and improve the accuracy of fraud detection.

[0066] The processing flow of the first embodiment will be briefly explained below.

[0067] Step 1: The learning unit learns the characteristics of fraudulent speech. For example, data on past fraudulent speech is collected and trained by the generation AI. The generation AI learns the characteristics of fraudulent speech using text generation AI (e.g., LLM) or multimodal generation AI. Step 2: The analysis unit analyzes the call content in real time based on the features learned by the learning unit. For example, the call content is analyzed in real time and compared with the features of fraudulent conversations. The generation AI analyzes the call content and compares it with the features of fraudulent conversations. Step 3: If the analysis unit detects fraud, the guidance unit provides guidance to the victim. For example, it may provide guidance such as, "This may be a fraud. Please hang up the call." Step 4: The disconnection unit disconnects the call after the guidance is played by the guidance unit, for example, after the guidance is played. Step 5: The text conversion unit converts the call content analyzed by the analysis unit into text. For example, the call content is converted into text. The generation AI converts the call content into text using voice recognition technology. Step 6: The sending unit sends the content that has been converted into text by the text conversion unit via SMS. For example, the sending unit sends the content that has been converted into text via SMS.

[0068] (Example 2) A fraud prevention system according to an embodiment of the present invention learns the characteristics of fraudulent conversations and compares them during a call. This fraud prevention system uses a generation AI to learn the characteristics of fraudulent conversations, analyzes the content of the call in real time, and, if fraud is detected, plays a message to the victim and disconnects the call. Furthermore, the call content is converted into text and sent via SMS, which can be used as information for consultation with the police or family. This system can deter successful fraud and instill in consumers the idea of ​​preventing "it's my son" scams. For example, a generation AI is used to learn the characteristics of fraudulent conversations. Data on past fraudulent conversations is collected and trained on the generation AI. The call content is then analyzed in real time. The generation AI analyzes the call content and compares it with the characteristics of fraudulent conversations. If fraud is detected, a message is played to the victim and the call is disconnected. For example, a message such as "This may be a scam. Please hang up." is played. The call content is then converted into text and sent via SMS, which can be used as information for consultation with the police or family. For example, if the content of a call is something like "Please transfer money," the content will be converted to text and sent via SMS. This system will deter successful fraud and establish in consumers' minds the image of preventing "it's my son" fraud. This allows the fraud prevention system to learn the characteristics of fraudulent conversations, detect fraud in real time during the call, provide guidance to the victim, disconnect the call, and convert the content of the call into text and send it, thereby deterring successful fraud and protecting victims.

[0069] The fraud prevention system according to the embodiment includes a learning unit, an analysis unit, a guidance unit, a disconnection unit, a text conversion unit, and a transmission unit. The learning unit learns the characteristics of fraudulent talk. For example, the learning unit collects data on past fraudulent talk and trains the generation AI. The generation AI learns the characteristics of fraudulent talk using a text generation AI (e.g., LLM) or a multimodal generation AI. The analysis unit analyzes the content of the call in real time based on the characteristics learned by the learning unit. For example, the analysis unit analyzes the content of the call in real time and compares it with the characteristics of the fraudulent talk. The generation AI analyzes the content of the call and compares it with the characteristics of the fraudulent talk. The guidance unit plays guidance to the victim when fraud is detected by the analysis unit. For example, the guidance unit plays guidance such as "There is a possibility of fraud. Please hang up the call." The disconnection unit disconnects the call after the guidance unit has played the guidance. For example, the disconnection unit disconnects the call after the guidance has been played. The text conversion unit converts the content of the call analyzed by the analysis unit into text. The text conversion unit, for example, converts the contents of the call into text. The generation AI converts the contents of the call into text using voice recognition technology. The transmission unit sends the contents converted into text by the text conversion unit via SMS. The transmission unit, for example, sends the contents converted into text via SMS. As a result, the fraud prevention system according to the embodiment can learn the characteristics of fraudulent conversations, detect fraud in real time during a call, provide guidance to the victim, disconnect the call, and convert the contents of the call into text and send it, thereby preventing successful fraud and protecting the victim.

[0070] The learning unit collects data on past fraudulent talk and allows the generation AI to learn from it. The learning unit, for example, collects data on past fraudulent talk and allows the generation AI to learn from it. The generation AI learns the characteristics of fraudulent talk using text generation AI (e.g., LLM) or multimodal generation AI. The generation AI, for example, collects data on past fraudulent talk and learns the characteristics of fraudulent talk based on that data. The generation AI, for example, analyzes data on fraudulent talk and learns specific phrases and speaking patterns. The generation AI, for example, analyzes data on fraudulent talk and learns specific voice tones and rhythms. In this way, by collecting data on past fraudulent talk and allowing the generation AI to learn, the characteristics of fraudulent talk can be learned more accurately.

[0071] The analysis unit can analyze the content of the call in real time and compare it with characteristics of fraudulent talk. The analysis unit, for example, analyzes the content of the call in real time and compares it with characteristics of fraudulent talk. The generation AI analyzes the content of the call and compares it with characteristics of fraudulent talk. The generation AI, for example, analyzes the content of the call in real time and compares it with characteristics of fraudulent talk. The generation AI, for example, analyzes the content of the call and compares it with specific phrases or speaking patterns. The generation AI, for example, analyzes the content of the call and compares it with specific voice tones and rhythms. This improves the accuracy of fraud detection by analyzing the content of the call in real time and comparing it with characteristics of fraudulent talk.

[0072] The guidance unit can play guidance such as "Fraud has been detected. Please hang up." when fraud is detected. For example, the guidance unit plays guidance such as "Fraud has been detected. Please hang up." when fraud is detected. For example, the guidance unit plays a voice message when fraud is detected. For example, the guidance unit plays a text message when fraud is detected. For example, the guidance unit plays both voice and text messages when fraud is detected. This allows victims to respond quickly by playing appropriate guidance when fraud is detected.

[0073] The disconnection unit can disconnect the call after the guidance unit has played the guidance. The disconnection unit, for example, disconnects the call after the guidance has been played. The disconnection unit, for example, automatically disconnects the call after the guidance has been played. The disconnection unit, for example, manually disconnects the call after the guidance has been played. The disconnection unit, for example, disconnects the call after a certain time has elapsed after the guidance has been played. In this way, by disconnecting the call after the guidance has been played, it is possible to prevent fraud.

[0074] The text conversion unit can convert the contents of a call into text. The text conversion unit, for example, converts the contents of a call into text. The text conversion unit, for example, converts the contents of a call into text using voice recognition technology. The text conversion unit, for example, converts the contents of a call into text in real time. The text conversion unit, for example, records the contents of a call and converts them into text later. By converting the contents of a call into text, the contents can be more easily checked later.

[0075] The transmitting unit can transmit the content converted into text by the text conversion unit via SMS. The transmitting unit, for example, transmits the text content via SMS. The transmitting unit, for example, transmits the text content to the police. The transmitting unit, for example, transmits the text content to a family member. The transmitting unit, for example, transmits the text content to multiple destinations. In this way, by transmitting the text content via SMS, the content can be used as material for consultation with the police or family members.

[0076] The learning unit can estimate the user's emotions and select training data for fraudulent talk based on the estimated user emotions. For example, if the user is feeling anxious, the learning unit causes the generation AI to prioritize training data from past fraudulent talk that is particularly anxiety-inducing. For example, if the user is relaxed, the learning unit causes the generation AI to select from past fraudulent talk that is likely to be fooled when the user is in a relaxed state. For example, if the learning unit causes the user to be wary, the generation AI selects from past fraudulent talk that will ease the user's guard. This allows for more effective training of fraudulent talk by selecting training data based on the user's emotions.

[0077] During learning, the learning unit can analyze the voice patterns of fraudulent talk and learn voice features. For example, the generation AI analyzes the voice data of fraudulent talk and learns specific voice patterns (e.g., voice tone and rhythm). For example, the generation AI analyzes the voice data of fraudulent talk and learns the emphasis of specific words and pauses. For example, the learning unit analyzes the voice data of fraudulent talk and learns specific emotional expressions (e.g., impatience and relief). In this way, by learning the voice patterns of fraudulent talk, fraud detection based on voice features becomes possible.

[0078] During learning, the learning unit can extract highly accurate features based on contextual information about the fraudulent conversation. For example, the generation AI in the learning unit analyzes the context of the fraudulent conversation and extracts features related to a specific scenario (e.g., a family emergency). For example, the generation AI in the learning unit analyzes the context of the fraudulent conversation and extracts specific phrases and phrasing patterns. For example, the generation AI in the learning unit analyzes the context of the fraudulent conversation and extracts specific patterns of conversation flow and development. This enables more accurate fraud detection by taking into account the contextual information about the fraudulent conversation.

[0079] The learning unit can estimate the user's emotions and adjust the frequency of learning based on the estimated user emotions. For example, if the user is feeling stressed, the generation AI sets the learning frequency low, thereby reducing the burden on the user. For example, if the user is relaxed, the learning unit sets the learning frequency high, thereby promoting efficient learning. For example, if the user is alert, the learning unit sets the learning frequency to a medium level, thereby achieving a balance. In this way, by adjusting the learning frequency based on the user's emotions, effective learning is possible while reducing the burden on the user.

[0080] During learning, the learning unit can take into account the regional characteristics of fraudulent speech and learn the characteristics of each region. For example, the generation AI collects data on fraudulent speech that occurs frequently in a specific region and learns the characteristics unique to that region. For example, the generation AI learns the characteristics of fraudulent speech by taking into account differences in language and dialect in each region. For example, the generation AI learns the characteristics of fraudulent speech by taking into account the culture and customs of each region. In this way, by taking into account regional characteristics, the system can learn the characteristics of fraudulent speech in each region and detect frauds that are unique to that region.

[0081] During learning, the learning unit can analyze the characteristics of fraudulent talk by time of day and learn the patterns for each time of day. For example, the generation AI classifies fraudulent talk data by time of day and learns the characteristics that are common in specific time periods. For example, the generation AI learns the patterns of fraudulent talk by time of day (for example, the difference between daytime and nighttime). For example, the generation AI analyzes the success rate of fraudulent talk during specific time periods and learns the characteristics unique to that time period. In this way, by learning the characteristics for each time period, the accuracy of detecting fraudulent talk during specific time periods is improved.

[0082] The analysis unit can estimate the user's emotions and adjust the accuracy of the analysis based on the estimated user emotions. For example, if the user is feeling anxious, the analysis unit causes the generation AI to increase the accuracy of the analysis, improving the detection rate of fraudulent conversations. For example, if the user is relaxed, the analysis unit causes the generation AI to set the analysis accuracy to normal, achieving a balance. For example, if the user is alert, the analysis unit causes the generation AI to set the analysis accuracy to medium, avoiding excessive warnings. In this way, adjusting the accuracy of the analysis based on the user's emotions improves the detection rate of fraudulent conversations.

[0083] During the analysis, the analysis unit can analyze background sounds of the call and identify elements that increase the likelihood of fraud. For example, the generation AI analyzes the background sounds of the call and identifies specific sounds (e.g., noise or silence) as elements that increase the likelihood of fraud. For example, the generation AI analyzes the background sounds of the call and identifies specific environmental sounds (e.g., office sounds or home sounds) as elements that increase the likelihood of fraud. For example, the analysis unit analyzes the background sounds of the call and identifies specific sound patterns (e.g., intermittent sounds or continuous sounds) as elements that increase the likelihood of fraud. In this way, by analyzing the background sounds of the call, elements that increase the likelihood of fraud are identified, improving the accuracy of fraud detection.

[0084] During analysis, the analysis unit can analyze the speed and tone of the call and compare them with characteristics of fraudulent talk. For example, the generation AI analyzes the speed of the call and compares a specific speed (e.g., fast speaking or slow speaking) with characteristics of fraudulent talk. For example, the generation AI analyzes the tone of the call and compares a specific tone (e.g., emphasized tone or flat tone) with characteristics of fraudulent talk. For example, the generation AI analyzes the combination of speed and tone of the call and compares a specific pattern with characteristics of fraudulent talk. In this way, by analyzing the speed and tone of the call, it can be compared with characteristics of fraudulent talk, improving the accuracy of fraud detection.

[0085] 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 feeling anxious, the analysis unit causes the generation AI to display the analysis results concisely, reducing the user's burden. For example, if the user is relaxed, the analysis unit causes the generation AI to display the analysis results in detail, providing the user with a lot of information. For example, if the user is alert, the analysis unit causes the generation AI to display the analysis results with a medium level of detail, achieving a balance. In this way, by adjusting the display method of the analysis results based on the user's emotions, the analysis unit provides appropriate information while reducing the user's burden.

[0086] During analysis, the analysis unit can identify region-specific fraudulent talk by taking into account the geographical information of the call. For example, the analysis unit uses a generation AI to analyze the geographical information of the call and identify fraudulent talk that occurs frequently in a specific region. For example, the analysis unit uses a generation AI to analyze the geographical information of the call and identify fraudulent talk by taking into account differences in language and dialect between regions. For example, the analysis unit uses a generation AI to analyze the geographical information of the call and identify fraudulent talk by taking into account the culture and customs of each region. In this way, by taking into account the geographical information of the call, region-specific fraudulent talk can be identified, improving the accuracy of fraud detection.

[0087] During analysis, the analysis unit can evaluate the likelihood of fraud based on contextual information of the call. For example, the generation AI analyzes the context of the call and evaluates the likelihood of fraud related to a specific scenario (e.g., a family emergency). For example, the generation AI analyzes the context of the call and evaluates the likelihood of fraud by referring to specific phrases and phrasing patterns. For example, the generation AI analyzes the context of the call and evaluates the likelihood of fraud by referring to specific patterns of conversation flow and development. This allows for a more accurate evaluation of the likelihood of fraud by referring to the contextual information of the call.

[0088] The guidance unit can estimate the user's emotions and adjust the content of the guidance based on the estimated user's emotions. For example, if the user is feeling anxious, the guidance unit plays the guidance in a calm tone, saying, "This may be a scam. Please hang up." For example, if the user is relaxed, the guidance unit plays the guidance in a normal tone, saying, "This may be a scam. Please hang up." For example, if the user is wary, the guidance unit plays the guidance in a warning tone, saying, "This may be a scam. Please hang up." In this way, by adjusting the content of the guidance based on the user's emotions, more effective guidance can be provided.

[0089] When providing guidance, the guidance unit can refer to the user's past call history to provide optimal guidance. For example, if the user has a history of falling for fraudulent talk in the past, the guidance unit provides strong warning guidance. For example, if the user has a history of ignoring fraudulent talk in the past, the guidance unit provides normal warning guidance. For example, if the user has a history of suspecting fraudulent talk in the past, the guidance unit provides mild warning guidance. In this way, by referring to the user's past call history, more appropriate guidance can be provided.

[0090] The guidance unit can generate different guidance messages depending on the content of the call when providing guidance. For example, if the content of the call is a request for money, the guidance unit plays the guidance, "This may be a scam. Please hang up." For example, if the content of the call is a request for personal information, the guidance unit plays the guidance, "This may be a scam. Please hang up." For example, if the content of the call is a report of an emergency, the guidance unit plays the guidance, "This may be a scam. Please hang up." In this way, by generating guidance messages depending on the content of the call, more effective guidance can be provided.

[0091] The guidance unit can estimate the user's emotions and adjust the timing of the guidance based on the estimated user's emotions. For example, if the user is feeling anxious, the guidance unit plays the guidance early to urge the user to end the call. For example, if the user is relaxed, the guidance unit plays the guidance at a normal timing. For example, if the user is alert, the guidance unit plays the guidance later to urge the user to end the call. In this way, by adjusting the timing of the guidance based on the user's emotions, more effective guidance can be provided.

[0092] The guidance unit can change the way the guidance is presented depending on the age and gender of the user when providing guidance. For example, the guidance unit provides guidance in a slow tone of voice for an elderly person. For example, the guidance unit provides guidance in a normal tone of voice for a young person. For example, the guidance unit provides guidance in a gentle tone of voice for a female person. This allows for more effective guidance to be provided by providing guidance according to the age and gender of the user.

[0093] The guidance unit can adjust the frequency of the guidance based on the length of the call when providing the guidance. For example, if the call continues for a long time, the guidance unit plays the guidance at regular intervals to warn of the possibility of fraud. For example, if the call is short, the guidance unit plays the guidance only once. For example, if the call is medium in length, the guidance unit plays the guidance at appropriate intervals. This allows for more effective guidance to be provided by adjusting the frequency of the guidance based on the length of the call.

[0094] The disconnection unit can estimate the user's emotions and adjust the timing of disconnecting the call based on the estimated user's emotions. For example, the disconnection unit disconnects the call early if the user is feeling anxious. For example, the disconnection unit disconnects the call at a normal timing if the user is relaxed. For example, the disconnection unit disconnects the call late if the user is on guard. In this way, by adjusting the timing of disconnecting the call based on the user's emotions, the call can be disconnected at a more appropriate timing.

[0095] The disconnection unit evaluates the importance of the call at the time of disconnection, and allows important calls to be continued with priority. For example, the disconnection unit allows the call to continue if the content of the call is about an emergency. For example, the disconnection unit allows the call to continue if the content of the call is about important business. For example, the disconnection unit allows the call to continue if the content of the call is about an important family topic. In this way, by evaluating the importance of the call, important calls can be continued with priority.

[0096] The disconnection unit can analyze the content of the call when disconnecting the call and disconnect the call only if there is a high possibility of fraud. For example, the disconnection unit disconnects the call if the content of the call is a request for money. For example, the disconnection unit disconnects the call if the content of the call is a request for personal information. For example, the disconnection unit disconnects the call if the content of the call is a report of an emergency. In this way, by analyzing the content of the call, it is possible to disconnect the call only if there is a high possibility of fraud.

[0097] The disconnection unit can estimate the user's emotions and adjust the call disconnection method based on the estimated user's emotions. For example, the disconnection unit quickly disconnects the call when the user feels anxious. For example, the disconnection unit disconnects the call in a normal manner when the user feels relaxed. For example, the disconnection unit carefully disconnects the call when the user feels alert. In this way, by adjusting the call disconnection method based on the user's emotions, the call can be disconnected in a more appropriate manner.

[0098] When disconnecting the call, the disconnection unit can make a decision to disconnect the call by taking into account attribute information of the other party of the call. For example, the disconnection unit disconnects the call if the other party is unknown. For example, the disconnection unit continues the call if the other party is a trustworthy person. For example, the disconnection unit disconnects the call if the other party is suspicious. In this way, by taking into account the attribute information of the other party of the call, it is possible to disconnect the call with a more appropriate decision.

[0099] When disconnecting a call, the disconnection unit can determine the priority of disconnection based on the time period of the call. For example, if the call is made at night, the disconnection unit disconnects the call with priority. For example, if the call is made during the day, the disconnection unit disconnects the call with normal priority. For example, if the call is made in the early morning, the disconnection unit disconnects the call with caution. In this way, by determining the priority of disconnection based on the time period of the call, it is possible to disconnect the call at a more appropriate time.

[0100] The text conversion unit can estimate the user's emotions and adjust the accuracy of the text conversion based on the estimated user's emotions. For example, if the user is feeling anxious, the text conversion unit increases the accuracy of the text conversion and provides more detailed content. For example, if the user is relaxed, the text conversion unit performs text conversion with normal accuracy. For example, if the user is alert, the text conversion unit sets the accuracy of the text conversion to a medium level to achieve a balance. In this way, by adjusting the accuracy of the text conversion based on the user's emotions, more accurate text can be generated.

[0101] The text conversion unit can evaluate the voice quality of the call during text conversion and select the optimal text conversion method. For example, if the voice quality of the call is high, the text conversion unit selects a normal text conversion method. For example, if the voice quality of the call is low, the text conversion unit performs noise reduction before converting to text. For example, if the voice quality of the call is medium, the text conversion unit performs appropriate noise reduction before converting to text. In this way, by evaluating the voice quality of the call, the optimal text conversion method can be selected and more accurate text can be generated.

[0102] The text conversion unit can generate more accurate text by taking into account contextual information of the call during text conversion. For example, the text conversion unit analyzes the context of the call and generates text based on a specific scenario. For example, the text conversion unit analyzes the context of the call and generates text by taking into account specific phrases and expressions. For example, the text conversion unit analyzes the context of the call and generates text by taking into account specific conversation flow and developments. In this way, by taking into account contextual information of the call, more accurate text can be generated.

[0103] The text conversion unit can estimate the user's emotions and determine the priority of text conversion based on the estimated user's emotions. For example, if the user feels anxious, the text conversion unit sets the priority of text conversion high and generates text quickly. For example, if the user feels relaxed, the text conversion unit performs text conversion with a normal priority. For example, if the user feels alert, the text conversion unit sets the priority of text conversion to a medium level to achieve a balance. In this way, by determining the priority of text conversion based on the user's emotions, text can be generated more quickly.

[0104] During the text conversion, the text conversion unit can generate text that supports multiple languages ​​by taking into account language information of the call. For example, if the call is in English, the text conversion unit generates text in English. For example, if the call is in Japanese, the text conversion unit generates text in Japanese. For example, if the call is in multiple languages, the text conversion unit generates text corresponding to each language. In this way, by taking into account language information of the call, text that supports multiple languages ​​can be generated.

[0105] The text conversion unit can adjust the format of the text based on the content of the call during text conversion. For example, if the content of the call is a request for money, the text conversion unit generates text in a format that emphasizes important parts. For example, if the content of the call is a request for personal information, the text conversion unit generates text in a format that emphasizes important parts. For example, if the content of the call is a report of an emergency, the text conversion unit generates text in a format that emphasizes important parts. In this way, by adjusting the format of the text based on the content of the call, it is possible to generate text that emphasizes important information.

[0106] The sending unit can estimate the user's emotion and adjust the timing of transmission based on the estimated user's emotion. For example, if the user is feeling anxious, the sending unit sends the text quickly. For example, if the user is relaxed, the sending unit sends the text at a normal timing. For example, if the user is on guard, the sending unit sends the text carefully. In this way, by adjusting the timing of transmission based on the user's emotion, it is possible to send the text at a more appropriate timing.

[0107] The transmitting unit can evaluate the importance of the call at the time of transmission and prioritize the transmission of important content. For example, if the call content is about an emergency, the transmitting unit prioritizes the transmission of text. For example, if the call content is about important business, the transmitting unit prioritizes the transmission of text. For example, if the call content is about an important family topic, the transmitting unit prioritizes the transmission of text. In this way, by evaluating the importance of the call, important content can be prioritized for transmission.

[0108] The sending unit can analyze the content of the call at the time of sending and select an appropriate destination. For example, if the content of the call is a request for money, the sending unit sends a text to the police. For example, if the content of the call is a request for personal information, the sending unit sends a text to a family member. For example, if the content of the call is a report of an emergency, the sending unit sends a text to an appropriate organization. In this way, by analyzing the content of the call, an appropriate destination can be selected, enabling a more effective response.

[0109] The sending unit can estimate the user's emotion and adjust the sending method based on the estimated user's emotion. For example, if the user is feeling anxious, the sending unit sends the text quickly. For example, if the user is relaxed, the sending unit sends the text in a normal manner. For example, if the user is alert, the sending unit sends the text carefully. In this way, by adjusting the sending method based on the user's emotion, it is possible to send the text in a more appropriate manner.

[0110] When transmitting, the transmitting unit can make a decision on whether to transmit taking into consideration the attribute information of the other party of the call. For example, if the other party of the call is unknown, the transmitting unit transmits the text to the police. For example, if the other party of the call is a trustworthy person, the transmitting unit transmits the text to a family member. For example, if the other party of the call is suspicious, the transmitting unit transmits the text to an appropriate organization. In this way, by taking into consideration the attribute information of the other party of the call, it is possible to transmit the text with a more appropriate decision.

[0111] The sending unit can determine the priority of sending based on the time period of the call when sending. For example, if the call is at night, the sending unit sends the text with priority. For example, if the call is during the day, the sending unit sends the text with normal priority. For example, if the call is early in the morning, the sending unit sends the text carefully. In this way, by determining the priority of sending based on the time period of the call, it is possible to send the text at a more appropriate time. === Hard Collateral 1-1 === Each of the multiple elements including the learning unit, analysis unit, guidance unit, disconnection unit, text conversion unit, and transmission unit described above is realized, for example, by at least one of the smart device 14 and the data processing device 12. For example, the learning unit is realized by the control unit 46A of the smart device 14 or the specific processing unit 290 of the data processing device 12. The analysis unit is realized, for example, by the specific processing unit 290 of the data processing device 12. The guidance unit is realized, for example, by the control unit 46A of the smart device 14. The disconnection unit is realized, for example, by the control unit 46A of the smart device 14. The text conversion unit is realized, for example, by the specific processing unit 290 of the data processing device 12. The transmission unit is realized, for example, by the control unit 46A of the smart device 14. === Hard Collateral 1-2 === Each of the multiple elements including the above-mentioned learning unit, analysis unit, guidance unit, cutting unit, text conversion unit, and transmission unit is realized, for example, by at least one of the smart glasses 214 and the data processing device 12. For example, the learning unit is realized by the control unit 46A of the smart glasses 214 or the specific processing unit 290 of the data processing device 12. The analysis unit is realized, for example, by the specific processing unit 290 of the data processing device 12. The guidance unit is realized, for example, by the control unit 46A of the smart glasses 214. The cutting unit is realized, for example, by the control unit 46A of the smart glasses 214. The text conversion unit is realized, for example, by the specific processing unit 290 of the data processing device 12. The transmission unit is realized, for example, by the control unit 46A of the smart glasses 214. === Hard Collateral 1-3 === Each of the multiple elements including the above-mentioned learning unit, analyzing unit, guidance unit, disconnecting unit, text conversion unit, and transmitting unit is realized, for example, by at least one of the headset type terminal 314 and the data processing device 12. For example, the learning unit is realized by the control unit 46A of the headset type terminal 314 or the specific processing unit 290 of the data processing device 12. The analyzing unit is realized, for example, by the specific processing unit 290 of the data processing device 12. The guidance unit is realized, for example, by the control unit 46A of the headset type terminal 314. The disconnecting unit is realized, for example, by the control unit 46A of the headset type terminal 314. The text conversion unit is realized, for example, by the specific processing unit 290 of the data processing device 12. The transmitting unit is realized, for example, by the control unit 46A of the headset type terminal 314. === Hard Collateral 1-4 === Each of the multiple elements including the above-mentioned learning unit, analyzing unit, guidance unit, cutting unit, text conversion unit, and transmitting unit is realized, for example, by at least one of the robot 414 and the data processing device 12. For example, the learning unit is realized by the control unit 46A of the robot 414 or the specific processing unit 290 of the data processing device 12. The analyzing unit is realized, for example, by the specific processing unit 290 of the data processing device 12. The guidance unit is realized, for example, by the control unit 46A of the robot 414. The cutting unit is realized, for example, by the control unit 46A of the robot 414. The text conversion unit is realized, for example, by the specific processing unit 290 of the data processing device 12. The transmitting unit is realized, for example, by the control unit 46A of the robot 414.

[0112] The system according to the embodiment is not limited to the above-described example, and various modifications are possible, for example, as follows.

[0113] The fraud prevention system may further include a voice analysis unit that analyzes the user's tone of voice and speaking patterns. For example, if the user is nervous, the voice analysis unit may detect a tendency for the user's voice tone to become higher, and analyze this as a factor that increases the likelihood of fraud. For example, if the user is relaxed, the voice analysis unit may detect that the user's voice tone is stable, and determine that the likelihood of fraud is low. For example, if the user is impatient, the voice analysis unit may detect that the user's speaking pattern becomes faster, and analyze this as a factor that increases the likelihood of fraud. In this way, by analyzing the user's tone of voice and speaking patterns, the accuracy of fraud detection can be improved.

[0114] The learning unit can further estimate the user's emotions and adjust the priority of the learning data based on the estimated user emotions. For example, if the user is feeling anxious, the generation AI will prioritize learning data on scam talk that will incite anxiety. If the user is relaxed, the generation AI will prioritize learning data on scam talk that is easy to fall for when the user is relaxed. If the user is wary, the generation AI will prioritize learning data on scam talk that will ease the user's guard. This allows for more effective learning of scam talk by adjusting the priority of the learning data based on the user's emotions.

[0115] The analysis unit can further analyze background sounds of calls to identify factors that increase the likelihood of fraud. For example, if the background sounds of a call are quiet, it determines that the likelihood of fraud is high. If the background sounds of a call are noisy, it determines that the likelihood of fraud is low. If the background sounds of a call include specific environmental sounds (e.g., office sounds or home sounds), it identifies the environmental sounds as factors that increase the likelihood of fraud. Thus, by analyzing background sounds of calls, it is possible to identify factors that increase the likelihood of fraud and improve the accuracy of fraud detection.

[0116] The guidance unit can further estimate the user's emotions and adjust the content of the guidance based on the estimated user emotions. For example, if the user is feeling anxious, the guidance unit can play a calming tone of "This may be a scam. Please hang up." If the user is relaxed, the guidance unit can play a normal tone of "This may be a scam. Please hang up." If the user is wary, the guidance unit can play a warning tone of "This may be a scam. Please hang up." This allows the provision of more effective guidance by adjusting the content of the guidance based on the user's emotions.

[0117] The disconnection unit further evaluates the importance of the call and allows important calls to be continued with priority. For example, if the content of the call is about an emergency, the call is allowed to continue. If the content of the call is about important business, the call is allowed to continue. If the content of the call is about an important family topic, the call is allowed to continue. In this way, by evaluating the importance of the call, important calls can be allowed to be continued with priority.

[0118] The text conversion unit can further estimate the user's emotions and adjust the accuracy of the text conversion based on the estimated user emotions. For example, if the user feels anxious, the text conversion unit increases the accuracy and provides more detailed content. If the user feels relaxed, the text conversion unit performs text conversion with normal accuracy. If the user feels alert, the text conversion unit sets the accuracy to a medium level to achieve a balance. This allows for more accurate text to be generated by adjusting the accuracy of the text conversion based on the user's emotions.

[0119] The sending unit can further analyze the content of the call and select an appropriate destination. For example, if the content of the call is a request for money, a text message can be sent to the police. If the content of the call is a request for personal information, a text message can be sent to a family member. If the content of the call is to report an emergency, a text message can be sent to an appropriate organization. This allows for more effective response by analyzing the content of the call and selecting an appropriate destination.

[0120] The learning unit can further estimate the user's emotions and adjust the learning frequency based on the estimated user emotions. For example, if the user is feeling stressed, the generation AI will set the learning frequency low, reducing the user's burden. If the user is relaxed, the generation AI will set the learning frequency high, allowing for efficient learning. If the user is alert, the generation AI will set the learning frequency to a medium level to achieve a balance. In this way, adjusting the learning frequency based on the user's emotions enables effective learning while reducing the user's burden.

[0121] The analysis unit can further analyze the speed and tone of the call and compare it with the characteristics of fraudulent talk. For example, if the speed of the call is fast, it is compared with the characteristics of fraudulent talk. If the tone of the call is emphasized, it is compared with the characteristics of fraudulent talk. The combination of the speed and tone of the call is analyzed, and specific patterns are compared with the characteristics of fraudulent talk. In this way, by analyzing the speed and tone of the call, it is possible to compare it with the characteristics of fraudulent talk and improve the accuracy of fraud detection.

[0122] The sending unit can further estimate the user's emotions and adjust the timing of transmission based on the estimated user's emotions. For example, if the user is feeling anxious, the sending unit sends the text quickly. If the user is relaxed, the sending unit sends the text at a normal timing. If the user is alert, the sending unit sends the text carefully. In this way, by adjusting the timing of transmission based on the user's emotions, the text can be sent at a more appropriate timing.

[0123] The processing flow of the second embodiment will be briefly explained below.

[0124] Step 1: The learning unit learns the characteristics of fraudulent speech. For example, data on past fraudulent speech is collected and trained by the generation AI. The generation AI learns the characteristics of fraudulent speech using text generation AI (e.g., LLM) or multimodal generation AI. Step 2: The analysis unit analyzes the call content in real time based on the features learned by the learning unit. For example, the call content is analyzed in real time and compared with the features of fraudulent conversations. The generation AI analyzes the call content and compares it with the features of fraudulent conversations. Step 3: If the analysis unit detects fraud, the guidance unit provides guidance to the victim. For example, it may provide guidance such as, "This may be a fraud. Please hang up the call." Step 4: The disconnection unit disconnects the call after the guidance is played by the guidance unit, for example, after the guidance is played. Step 5: The text conversion unit converts the call content analyzed by the analysis unit into text. For example, the call content is converted into text. The generation AI converts the call content into text using voice recognition technology. Step 6: The sending unit sends the content that has been converted into text by the text conversion unit via SMS. For example, the sending unit sends the content that has been converted into text via SMS.

[0125] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0126] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> Examples of the generative AI include a neural network (NN) and a neural network (NN). The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 receives a prompt containing an instruction, as well as inference data such as voice data representing speech, text data representing text, and image data representing an image (e.g., still image data or video data). The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in one or more data formats of voice data, text data, image data, etc. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The specification processing unit 290 performs the above-mentioned specification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation model 58 includes AIs other than the generative AI. The AI ​​other than the generative AI may be, for example, linear regression, logistic regression, decision tree, random forest, support vector machine (SVM), k-means clustering, convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), or naive Bayes, and may perform various processes, but is not limited to these examples. The AI ​​may also be an AI agent. When the processing of each of the above-mentioned parts is performed by an AI, the processing may be performed in part or entirely by the AI, but is not limited to these examples. The processing performed by an AI including the generative AI may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by an AI including the generative AI.

[0127] Furthermore, the processing by the data processing system 10 described above is executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart device 14, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart device 14. Furthermore, 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.

[0128] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0129] [Second embodiment] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0130] 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.

[0131] 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, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN and / or a LAN.

[0132] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0133] The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to instructions from the processor 46.

[0134] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the user's surroundings (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0135] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0136] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0137] The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0138] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0139] In the smart glasses 214, the specific processing is performed by the processor 46. A specific processing program 60 is stored in the storage 50. 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 the control unit 46A in accordance with the specific processing program 60 executed on the RAM 48. The smart glasses 214 also have a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can perform processing similar to that of the specific processing unit 290 using these models.

[0140] Note that a device other than the data processing device 12 may 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 communicates with the server device having the data generation model 58 to obtain a processing result (such as a prediction result) using the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device (for example, a mobile phone, a robot, a home appliance, etc.) owned by a user.

[0141] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0142] The data generation model 58 is a so-called generative AI. An example of the 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 receives a prompt including an instruction, as well as inference data such as audio data indicating speech, text data indicating text, and image data indicating an image (e.g., still image data or video data). The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in one or more data formats, such as audio data, text data, and image data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The identification processing unit 290 performs the above-mentioned identification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation models 58 include AIs other than the generation AI. Examples of AIs other than the generation AI include, but are not limited to, 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), and naive Bayes. These AIs can perform various types of processing, but are not limited to these examples. The AI ​​may also be an AI agent. When the processing of each of the above-described parts is performed by an AI, the processing may be performed in part or entirely by the AI, but is not limited to these examples. Processing performed by an AI, including the generation AI, may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by an AI, including the generation AI.

[0143] 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 executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the smart glasses 214, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the smart glasses 214. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the smart glasses 214 or an external device, etc., and the smart glasses 214 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

[0144] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0145] [Third embodiment] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0146] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0147] 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, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN and / or a LAN.

[0148] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0149] The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to instructions from the processor 46.

[0150] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the user's surroundings (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0151] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0152] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0153] The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0154] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0155] In the headset type terminal 314, the identification process is performed by the processor 46. A identification program 60 is stored in the storage 50. The processor 46 reads the identification program 60 from the storage 50 and executes the read identification program 60 on the RAM 48. The identification process is realized by the processor 46 operating as a control unit 46A in accordance with the identification program 60 executed on the RAM 48. Note that the headset type terminal 314 has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can also perform processing similar to that of the identification processing unit 290 using these models.

[0156] Note that a device other than the data processing device 12 may 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 communicates with the server device having the data generation model 58 to obtain a processing result (such as a prediction result) using the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device (for example, a mobile phone, a robot, a home appliance, etc.) owned by a user.

[0157] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0158] The data generation model 58 is a so-called generative AI. An example of the 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 receives a prompt including an instruction, as well as inference data such as audio data indicating speech, text data indicating text, and image data indicating an image (e.g., still image data or video data). The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in one or more data formats, such as audio data, text data, and image data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The identification processing unit 290 performs the above-mentioned identification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation models 58 include AIs other than the generation AI. Examples of AIs other than the generation AI include, but are not limited to, 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), and naive Bayes. These AIs can perform various types of processing, but are not limited to these examples. The AI ​​may also be an AI agent. When the processing of each of the above-described parts is performed by an AI, the processing may be performed in part or entirely by the AI, but is not limited to these examples. Processing performed by an AI, including the generation AI, may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by an AI, including the generation AI.

[0159] 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 executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the headset type terminal 314, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the headset type terminal 314. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the headset type terminal 314 or an external device, etc., and the headset type terminal 314 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

[0160] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0161] [Fourth embodiment] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0162] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0163] 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, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN and / or a LAN.

[0164] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[0165] The microphone 238 receives instructions and the like from the user by receiving voice uttered by the user. The microphone 238 captures the voice uttered by the user, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to instructions from the processor 46.

[0166] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS image sensor or a CCD image sensor, and captures images of the user's surroundings (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0167] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0168] The control 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 emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[0169] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0170] The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0171] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. The identification processing unit 290 can estimate a user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion. The emotion estimation function (emotion identification function) using the emotion identification model 59 performs various estimations and predictions regarding the user's emotion, including estimation and prediction of the user's emotion, but is not limited to these examples. Furthermore, the estimation and prediction of emotion also includes, for example, emotion analysis.

[0172] In the robot 414, the processor 46 performs the identification process. The storage 50 stores the identification program 60. The processor 46 reads the identification program 60 from the storage 50 and executes the read identification program 60 on the RAM 48. The identification process is realized by the processor 46 operating as the control unit 46A in accordance with the identification program 60 executed on the RAM 48. The robot 414 also has a data generation model and an emotion identification model similar to the data generation model 58 and the emotion identification model 59, and can perform the same process as the identification processing unit 290 using these models.

[0173] Note that a device other than the data processing device 12 may 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 communicates with the server device having the data generation model 58 to obtain a processing result (such as a prediction result) using the data generation model 58. Furthermore, the data processing device 12 may be a server device, or may be a terminal device (for example, a mobile phone, a robot, a home appliance, etc.) owned by a user.

[0174] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[0175] The data generation model 58 is a so-called generative AI. An example of the 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 receives a prompt including an instruction, as well as inference data such as audio data indicating speech, text data indicating text, and image data indicating an image (e.g., still image data or video data). The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in one or more data formats, such as audio data, text data, and image data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. The identification processing unit 290 performs the above-mentioned identification processing using the data generation model 58. The data generation model 58 may be a fine-tuned model so as to output an inference result from a prompt that does not include an instruction. In this case, the data generation model 58 can output an inference result from a prompt that does not include an instruction. The data processing device 12 and the like include multiple types of data generation models 58, and the data generation models 58 include AIs other than the generation AI. Examples of AIs other than the generation AI include, but are not limited to, 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), and naive Bayes. These AIs can perform various types of processing, but are not limited to these examples. The AI ​​may also be an AI agent. When the processing of each of the above-described parts is performed by an AI, the processing may be performed in part or entirely by the AI, but is not limited to these examples. Processing performed by an AI, including the generation AI, may be replaced with rule-based processing, and rule-based processing may be replaced with processing performed by an AI, including the generation AI.

[0176] 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 executed by the specific processing unit 290 of the data processing device 12 or the control unit 46A of the robot 414, but may also be executed by the specific processing unit 290 of the data processing device 12 and the control unit 46A of the robot 414. Furthermore, the specific processing unit 290 of the data processing device 12 acquires or collects information required for processing from the robot 414 or an external device, etc., and the robot 414 acquires or collects information required for processing from the data processing device 12 or an external device, etc.

[0177] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0178] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0179] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion encompasses both emotions and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[0180] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[0181] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[0182] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is expressed, and when they approach the ideal, a state of pleasure is expressed. Emotions can also be created for robots, cars, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is expressed, and when they approach the ideal, a state of pleasure is expressed. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on speech emotion recognition and brain physiological signal analysis systems for emotions, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the area called "reaction," where sensation is dominant. The right half of the emotion map lists emotions belonging to the area called "situation," where situational awareness is dominant.

[0183] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[0184] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[0185] In the above embodiment, an example was given in which a specific process is performed by one computer 22, but the technology disclosed herein is not limited to this, and distributed processing of the specific process may be performed by multiple computers including computer 22.

[0186] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[0187] 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.

[0188] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[0189] The hardware resource for executing a specific process can be any of the following processors: A CPU is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. A dedicated electrical circuit, such as a field-programmable gate array (FPGA), a programmable logic device (PLD), or an application-specific integrated circuit (ASIC), is a processor with a circuit configuration specifically designed to execute a specific process. Each processor has built-in or connected memory, and uses the memory to execute the specific process.

[0190] The hardware resource that executes the specific process may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific process may be a single processor.

[0191] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[0192] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[0193] In the above example, the first to fourth embodiments have been described separately, but some or all of these embodiments may be combined. The smart device 14, smart glasses 214, headset terminal 314, and robot 414 are merely examples, and they may be combined, or other devices may be used. In the above example, the first and second embodiments have been described separately, but they may be combined.

[0194] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[0195] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[0196] [Explanation of symbols]

[0197] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot

Claims

1. A learning department that learns the characteristics of fraudulent talk, an analysis unit that analyzes the content of the call in real time based on the features learned by the learning unit; a guidance unit that provides guidance to a victim when fraud is detected by the analysis unit; a disconnection unit that disconnects the call after the guidance unit has played the guidance; a text conversion unit that converts the call content analyzed by the analysis unit into text; a transmitting unit that transmits the content converted into text by the text conversion unit via SMS; Equipped with A system characterized by:

2. The learning unit Collect data on past fraudulent conversations and train the AI ​​to generate them 2. The system of claim 1.

3. The analysis unit Analyze call content in real time and compare it with characteristics of fraudulent conversations 2. The system of claim 1.

4. The cutting portion is The call is disconnected after the guidance is played by the guidance unit.

2. The system of claim 1.

5. The text conversion unit Convert call content to text 2. The system of claim 1.

6. The transmission unit The content converted into text by the text conversion unit is sent via SMS.

2. The system of claim 1.

7. The learning unit Estimate user emotions and select fraudulent speech training data based on the estimated user emotions.

2. The system of claim 1.

Images