System

Abstract

The system according to the embodiment aims to analyze voice data of a telephone and detect and prevent a fraud in real time.SOLUTION: A system according to an embodiment includes a collection unit, an analysis unit, a determination unit, and a warning unit. The collection unit collects voice data of the telephone. The analysis unit analyzes the voice data collected by the collection unit. The determination unit determines a fraud criterion based on the data analyzed by the analysis unit. The warning unit issues a warning when the criteria for fraud are met by the determination unit.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Conventional techniques have had the problem of making it difficult to detect and prevent fraudulent activities over the phone in real time.

[0005] The system according to the embodiment aims to analyze telephone voice data and detect and prevent fraud in real time. [Means for solving the problem]

[0006] The system according to the embodiment includes a collection unit, an analysis unit, a determination unit, and a warning unit. The collection unit collects telephone voice data. The analysis unit analyzes the voice data collected by the collection unit. The determination unit determines fraud criteria based on the data analyzed by the analysis unit. The warning unit issues a warning when the determination unit determines that the fraud criteria are met. [Effects of the Invention]

[0007] The system according to the embodiment can analyze the voice data of the telephone and detect and prevent fraud in real time. [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 system according to an embodiment of the present invention analyzes telephone conversations to prevent fraud. This system collects voice data, analyzes it using AI, determines whether a call is fraudulent, and issues a warning. For example, the system records the call in real time and saves it as audio data. Next, AI analyzes the collected audio data to determine whether a call is fraudulent. If a call is fraudulent, a warning is issued. This prevents fraudulent activity from occurring. For example, the system automatically starts recording the call as soon as it begins. Next, AI converts the call into text using speech recognition technology and analyzes the text data. For example, it checks whether specific keywords or phrases are included. If a call is determined to be fraudulent, an alarm is sounded or a warning message is displayed after the call ends. This alerts the user to the possibility of fraud. This allows the system to prevent fraudulent activity from occurring. For example, when elderly people are targets of fraud, this system can be used to prevent fraudulent activity from occurring.

[0029] The fraud prevention system according to the embodiment includes a collection unit, an analysis unit, a determination unit, and a warning unit. The collection unit collects telephone voice data. For example, the collection unit records the contents of a call in real time and saves it as voice data. The collection unit can also automatically start recording when a call is initiated. The collection unit can also use noise filtering technology to collect high-quality voice data. The analysis unit analyzes the voice data collected by the collection unit. For example, the analysis unit converts the contents of the call into text using voice recognition technology and analyzes the text data. The analysis unit can also check whether specific keywords or phrases are included. The analysis unit can also have an emotion analysis function that analyzes the tone of the voice and the speaker's emotions. The determination unit determines the possibility of fraud based on the data analyzed by the analysis unit. For example, the determination unit determines the possibility of fraud based on whether specific keywords or phrases are included. The determination unit can also determine the possibility of fraud by analyzing the tone of the voice and the speaker's emotions. The warning unit issues a warning if the determination unit determines that there is a possibility of fraud. For example, the warning unit can sound a warning sound during a call. The warning unit can also display a warning message after the call ends. Furthermore, the warning unit can have a function to automatically notify family members or the police. In this way, the fraud prevention system according to the embodiment can prevent fraudulent acts before they occur. For example, the system can prevent fraud damage by detecting potentially fraudulent calls in real time and issuing a warning to the user.

[0030] The analysis unit can convert the call content into text using speech recognition technology and analyze the text data. Examples of speech recognition technology include speech recognition algorithms using deep learning. The analysis unit can convert the call content into text with high accuracy using speech recognition algorithms. The analysis unit can also extract specific keywords and phrases from the call content using speech recognition technology. Furthermore, the analysis unit can understand the context of the call content and determine the possibility of fraud with high accuracy using speech recognition technology. This improves the accuracy of analyzing the call content by using speech recognition technology. Some or all of the above-described processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input voice data into a generation AI and have the generation AI convert the voice data into text data.

[0031] The determination unit can check whether specific keywords or phrases are included. For example, the determination unit uses a list of keywords related to fraud to check whether specific keywords or phrases are included in the call content. The determination unit can also periodically update the keyword list to respond to new fraudulent methods. Furthermore, the determination unit can determine the possibility of fraud based on the frequency of appearance of keywords and phrases. In this way, by checking for specific keywords and phrases, the possibility of fraud can be determined with high accuracy. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input text data into a generation AI and have the generation AI check for keywords and phrases.

[0032] The determination unit may include an emotion analysis unit that analyzes the tone of the voice and the speaker's emotion. The emotion analysis unit may include, for example, an algorithm for analyzing the tone of the voice and the speaker's emotion. The emotion analysis unit may analyze, for example, the pitch, volume, rhythm, etc. of the voice to estimate the speaker's emotion. The emotion analysis unit may also classify the speaker's emotion and use the resulting data to determine the possibility of fraud. Furthermore, the emotion analysis unit may monitor changes in the speaker's emotion in real time to detect signs of fraud. This allows for more accurate determination of the possibility of fraud by analyzing the tone of the voice and the speaker's emotion. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or a generative AI. The generative AI may be, for example, a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. Some or all of the above-described processing in the determination unit may be performed using, for example, an AI. For example, the determination unit may input voice data to a generative AI and have the generative AI perform emotion analysis.

[0033] The warning unit can sound a warning sound during a call or display a warning message after the call ends. The warning unit, for example, includes a speaker for sounding a warning sound during a call. The warning unit can sound a warning sound during a call, for example, if it determines that there is a possibility of fraud. The warning unit can also include a display for displaying a warning message after the call ends. For example, the warning unit can display a warning message after the call ends to notify the user of the possibility of fraud. The warning unit can also include a function for customizing the content of the warning sound or warning message. This allows the user to be notified of the possibility of fraud by issuing a warning during or after the call ends. Some or all of the above-mentioned processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit can input the determination result to a generation AI and cause the generation AI to generate a warning sound or warning message.

[0034] The warning unit may include a notification unit that automatically notifies family members or the police. The notification unit may include, for example, a communication module for automatically notifying family members or the police when there is a possibility of fraud. The notification unit may automatically send a notification to family members or the police when it determines there is a possibility of fraud. The notification unit may also include a function for customizing the content of the notification. For example, the notification unit may allow the user to set the content of the notification message. Furthermore, the notification unit may also set multiple notification destinations. This enables a prompt response by automatically notifying family members or the police when there is a possibility of fraud. Some or all of the above-mentioned processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit may input the determination result into a generation AI and cause the generation AI to generate a notification message.

[0035] The analysis unit may include a learning management unit that uses past fraud cases as learning data. The learning management unit may, for example, include a database for using past fraud cases as learning data. The learning management unit may, for example, collect past fraud cases and store them in the database. The learning management unit may also analyze the collected fraud cases to learn fraud patterns. Furthermore, the learning management unit may periodically update the learning data to respond to new fraud methods. In this way, using past fraud cases as learning data improves analysis accuracy. Some or all of the above-mentioned processing in the learning management unit may be performed, for example, using AI, or may be performed without using AI. For example, the learning management unit may input fraud case data into a generation AI and cause the generation AI to perform learning.

[0036] The system may include a privacy protection unit that protects user privacy. The privacy protection unit may, for example, include data anonymization technology to protect user privacy. The privacy protection unit may, for example, anonymize collected voice data to protect personal information. The privacy protection unit may also include an access control function to allow only specific users to access the data. Furthermore, the privacy protection unit may also include a function to set a data storage period and automatically delete the data after a certain period. This protects user privacy, allowing the system to be used with peace of mind. Some or all of the above-mentioned processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit may input voice data into a generation AI and have the generation AI perform anonymization processing.

[0037] The collection unit not only automatically collects voice data at the start of a call, but also can collect additional voice data when a specific keyword is detected. The collection unit, for example, has a function for automatically collecting voice data at the start of a call. The collection unit, for example, automatically starts recording when the call starts. The collection unit can also have a function for collecting additional voice data when a specific keyword is detected. For example, the collection unit collects additional voice data when a specific keyword, such as "money" or "transfer," is detected during a call. Furthermore, the collection unit can also immediately collect additional voice data when a highly urgent keyword, such as "help," is detected during a call. In this way, by collecting additional voice data when a specific keyword is detected, important information can be collected without missing any important information. Some or all of the above-described processing by the collection unit may be performed using, or without, AI. For example, the collection unit may input voice data to a generation AI and cause the generation AI to detect specific keywords and collect additional voice data.

[0038] The collection unit can filter background sounds during a call to remove noise when collecting voice data. The collection unit, for example, includes a noise removal algorithm for filtering background sounds during a call to remove noise. The collection unit, for example, filters background sounds during a call (e.g., traffic noise, wind noise) in real time to collect clear voice data. The collection unit can also remove noise during a call (e.g., static, echo) and collect only important conversation portions. Furthermore, the collection unit can filter environmental sounds during a call (e.g., television sounds, other conversations) to preferentially collect conversation portions that are more likely to be fraudulent. This allows clear voice data to be collected by filtering background sounds during a call and removing noise. Some or all of the above-described processing in the collection unit may be performed using, or without, AI. For example, the collection unit can input voice data to a generation AI and have the generation AI perform noise removal.

[0039] When collecting voice data, the collection unit can prioritize collection when the caller is from a specific number. The collection unit, for example, has a function for prioritized collection of voice data when the caller is from a specific number. For example, the collection unit prioritizes collection of voice data when the caller is from a specific number, such as a family member or friend. The collection unit can also prioritize collection of voice data when the caller is from a number suspected of fraud in the past. Furthermore, the collection unit can also prioritize collection of voice data when the caller is from a number registered as an emergency contact. In this way, by prioritizing collection of calls from specific numbers, important call content can be reliably collected. Some or all of the above-described processing in the collection unit may be performed using, for example, AI, or may be performed without AI. For example, the collection unit can input call number data to a generation AI and cause the generation AI to detect specific numbers and prioritize collection of voice data.

[0040] When collecting voice data, the collection unit can prioritize collection of highly relevant calls by taking into account the user's geographical location information. The collection unit, for example, includes a GPS module for acquiring the user's geographical location information. For example, when the user is in a specific area, the collection unit prioritizes collection of calls related to that area. Furthermore, when the user is traveling, the collection unit can also prioritize collection of calls related to the user's travel destination. Furthermore, when the user is at home, the collection unit can also prioritize collection of calls related to information around the user's home. In this way, highly relevant calls can be collected preferentially by taking into account the user's geographical location information. Some or all of the above-described processing in the collection unit may be performed using AI, for example, or may be performed without using AI. For example, the collection unit can input geographical location data to the generation AI and cause the generation AI to prioritize collection of highly relevant calls.

[0041] The collection unit can analyze the user's social media activity and collect related calls when collecting voice data. The collection unit, for example, includes a data analysis algorithm for analyzing the user's social media activity. The collection unit, for example, collects calls related to places where the user has checked in on social media. The collection unit can also analyze the content of the user's social media posts and collect related calls. Furthermore, the collection unit can collect related calls by referring to the activities of the user's friends on social media. In this way, related calls can be efficiently collected by analyzing the user's social media activity. Some or all of the above-mentioned processing in the collection unit may be performed using, for example, AI, or may be performed without using AI. For example, the collection unit can input social media data into a generation AI and cause the generation AI to collect related calls.

[0042] The collection unit can customize the collection method by reflecting the user's past feedback when collecting voice data. The collection unit, for example, has a function for collecting the user's past feedback and storing it in a database. The collection unit, for example, optimizes the collection method based on feedback provided by the user in the past. The collection unit can also prioritize collection of call content that the user has previously determined to be important. Furthermore, the collection unit can also reflect the user's past feedback and adjust the collection timing and collection range. In this way, the collection method can be optimized by reflecting the user's past feedback. Some or all of the above-mentioned processing in the collection unit may be performed using, for example, AI, or may be performed without using AI. For example, the collection unit can input feedback data to a generation AI and cause the generation AI to customize the collection method.

[0043] When analyzing voice data, the analysis unit can improve the accuracy of the analysis by taking into account the context of the call. The analysis unit, for example, includes an algorithm for analyzing the meaning of specific keywords by taking into account the context before and after the call. The analysis unit, for example, grasps the overall flow of the call and identifies important parts. The analysis unit can also prioritize analysis of parts that are more likely to be fraudulent based on the context of the call. This improves the accuracy of the analysis by taking into account the context of the call. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input voice data to a generation AI and have the generation AI perform context analysis.

[0044] When analyzing voice data, the analysis unit can perform the analysis taking into account attribute information of the caller. The analysis unit, for example, includes a database for acquiring attribute information of the caller. For example, if the caller is a family member or friend, the analysis unit performs the analysis taking into account the attribute information. Furthermore, if the caller is a person suspected of fraud in the past, the analysis unit can also perform the analysis taking into account the attribute information. Furthermore, if the caller is an emergency contact, the analysis unit can also perform the analysis taking into account the attribute information. In this way, by taking into account the attribute information of the caller, the accuracy of the analysis is improved. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input the attribute information of the caller into a generation AI and have the generation AI perform the analysis.

[0045] When analyzing the voice data, the analysis unit can weight the analysis based on the frequency of calls. The analysis unit includes, for example, an algorithm for evaluating the frequency of calls. The analysis unit, for example, prioritizes analysis of frequently made calls. The analysis unit can also prioritize analysis of calls with frequently called parties. Furthermore, the analysis unit can also prioritize analysis of calls with high importance based on the frequency of calls. In this way, by weighting the analysis based on the frequency of calls, important calls can be prioritized in analysis. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input call frequency data to a generation AI and have the generation AI perform the weighting.

[0046] The analysis unit can take into account the geographical distribution of calls when analyzing voice data. The analysis unit, for example, includes an algorithm for evaluating the geographical distribution of calls. For example, the analysis unit takes into account the geographical distribution of calls and prioritizes analysis of calls related to a specific region. The analysis unit can also prioritize analysis of calls from regions with a high probability of fraud based on the geographical distribution of calls. Furthermore, the analysis unit can analyze the geographical distribution of calls to identify fraud trends by region. This improves the accuracy of the analysis by taking the geographical distribution of calls into account. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input geographical distribution data into a generation AI and have the generation AI perform the analysis.

[0047] When analyzing voice data, the analysis unit can improve the accuracy of the analysis by referring to related literature. The analysis unit, for example, includes a database for referring to related literature. The analysis unit performs analysis based on, for example, the latest information on fraudulent methods. The analysis unit can also analyze the meaning of specific keywords and phrases based on related literature. Furthermore, the analysis unit can also identify call content that is likely to be fraudulent by referring to related literature. In this way, referring to related literature improves the accuracy of the analysis. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input literature data into a generation AI and have the generation AI perform the analysis.

[0048] The analysis unit can take into account the market value of the call when analyzing the voice data. The analysis unit, for example, includes an algorithm for evaluating the market value of the call. The analysis unit, for example, takes into account the market value of the call and prioritizes analysis of important call content. The analysis unit can also identify calls that are likely to be fraudulent based on the market value of the call. Furthermore, the analysis unit can analyze the market value of the call and prioritize analysis of calls that have a high risk of fraud. In this way, by taking the market value of the call into consideration, important call content can be prioritized for analysis. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input market value data to a generation AI and have the generation AI perform the analysis.

[0049] When determining the possibility of fraud, the determination unit can improve the accuracy of the determination by taking into account the interrelationships between calls. The determination unit, for example, includes an algorithm for evaluating the interrelationships between calls. The determination unit, for example, determines the possibility of fraud by taking into account the content of conversations before and after a call. The determination unit can also determine the possibility of fraud by taking into account past call history with the call partner. Furthermore, the determination unit can analyze the interrelationships between calls and identify calls that are likely to be fraudulent. In this way, by taking the interrelationships between calls into account, the possibility of fraud can be determined more accurately. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input call history data into a generation AI and cause the generation AI to analyze the interrelationships and determine the possibility of fraud.

[0050] When determining the possibility of fraud, the determination unit can make the determination taking into account attribute information of the other party of the call. The determination unit, for example, includes a database for acquiring attribute information of the other party of the call. For example, if the other party of the call is a family member or friend, the determination unit determines the possibility of fraud taking into account the attribute information. Furthermore, if the other party of the call is a person suspected of fraud in the past, the determination unit can also determine the possibility of fraud taking into account the attribute information. Furthermore, if the other party of the call is an emergency contact, the determination unit can also determine the possibility of fraud taking into account the attribute information. Thus, by taking into account the attribute information of the other party of the call, the possibility of fraud can be more accurately determined. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input the other party's attribute information into a generation AI and cause the generation AI to determine the possibility of fraud.

[0051] When determining the possibility of fraud, the determination unit can weight the determination based on the frequency of calls. The determination unit includes, for example, an algorithm for evaluating the frequency of calls. The determination unit, for example, prioritizes calls that are made frequently. The determination unit can also prioritize calls with partners who are called frequently. Furthermore, the determination unit can also prioritize calls with high importance based on the frequency of calls. In this way, by weighting the determination based on the frequency of calls, important calls can be prioritized. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input call frequency data to a generation AI and have the generation AI perform the weighting.

[0052] The determination unit can make a determination regarding the possibility of fraud by taking into account the geographic distribution of calls. The determination unit, for example, includes an algorithm for evaluating the geographic distribution of calls. The determination unit, for example, takes into account the geographic distribution of calls and prioritizes calls related to a specific region. The determination unit can also prioritize calls from regions with a high probability of fraud based on the geographic distribution of calls. Furthermore, the determination unit can analyze the geographic distribution of calls and understand fraud trends by region. This allows for a more accurate determination of the possibility of fraud by taking the geographic distribution of calls into account. Some or all of the above-described processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input geographic distribution data into a generation AI and have the generation AI perform the determination.

[0053] When determining the possibility of fraud, the determination unit can improve the accuracy of the determination by referring to related literature. The determination unit, for example, includes a database for referring to related literature. The determination unit makes the determination based on, for example, the latest information on fraudulent methods. The determination unit can also determine the meaning of specific keywords or phrases based on related literature. Furthermore, the determination unit can also identify call content that is likely to be fraudulent by referring to related literature. In this way, the accuracy of the determination is improved by referring to related literature. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input literature data into a generation AI and have the generation AI perform the determination.

[0054] The determination unit can make a determination taking into account the market value of the call when determining the possibility of fraud. The determination unit, for example, includes an algorithm for evaluating the market value of the call. The determination unit, for example, takes into account the market value of the call and prioritizes determining important call content. The determination unit can also identify calls that are likely to be fraudulent based on the market value of the call. Furthermore, the determination unit can analyze the market value of the call and prioritize determining calls that have a high risk of fraud. In this way, important call content can be prioritized by considering the market value of the call. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input market value data to a generation AI and have the generation AI perform the determination.

[0055] When issuing a warning, the warning unit can adjust the level of detail of the warning based on the content of the call. The warning unit, for example, includes an algorithm for evaluating the content of the call. For example, the warning unit issues a detailed warning if the content of the call is related to a specific fraud method. The warning unit can also issue a concise warning if the content of the call indicates the possibility of a general fraud. Furthermore, the warning unit can also issue an immediate warning if the content of the call is urgent. In this way, by adjusting the level of detail of the warning based on the content of the call, a more appropriate warning can be provided. Some or all of the above-mentioned processing in the warning unit may be performed using AI, for example, or may be performed without using AI. For example, the warning unit can input call content data to a generation AI and cause the generation AI to adjust the level of detail of the warning.

[0056] When issuing a warning, the warning unit can take into account attribute information of the other party of the call. The warning unit, for example, includes a database for acquiring attribute information of the other party of the call. For example, if the other party of the call is a family member or a friend, the warning unit can take into account the attribute information when issuing the warning. Furthermore, if the other party of the call is a person suspected of fraud in the past, the warning unit can also take into account the attribute information when issuing the warning. Furthermore, if the other party of the call is an emergency contact, the warning unit can also take into account the attribute information when issuing the warning. In this way, by taking into account the attribute information of the other party of the call, a more appropriate warning can be provided. Some or all of the above-mentioned processing in the warning unit may be performed, for example, using AI, or may be performed without using AI. For example, the warning unit can input the other party's attribute information into a generation AI and have the generation AI execute the warning.

[0057] When issuing a warning, the warning unit can weight the warning based on the frequency of calls. The warning unit includes, for example, an algorithm for evaluating the frequency of calls. The warning unit, for example, weights the warning for calls that are made frequently. The warning unit can also weight the warning for calls with a party with whom the number of calls is high. Furthermore, the warning unit can weight the warning for calls with high importance based on the frequency of calls. In this way, by weighting the warning based on the frequency of calls, important calls can be given priority in being warned. Some or all of the above-mentioned processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit can input call frequency data to a generation AI and have the generation AI perform the weighting.

[0058] The warning unit may issue a warning taking into account the geographical distribution of calls. The warning unit may, for example, include an algorithm for evaluating the geographical distribution of calls. The warning unit may, for example, take into account the geographical distribution of calls and issue a warning for calls related to a specific region. The warning unit may also issue a warning for calls from regions where there is a high possibility of fraud based on the geographical distribution of calls. Furthermore, the warning unit may analyze the geographical distribution of calls, identify fraud trends by region, and issue a warning. In this way, by taking the geographical distribution of calls into account, it becomes easier to identify fraud trends by region. Some or all of the above-described processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit may input geographical distribution data into a generation AI and cause the generation AI to execute the warning.

[0059] When issuing a warning, the warning unit can improve the accuracy of the warning by referring to related literature. The warning unit, for example, includes a database for referring to related literature. The warning unit issues a warning based on, for example, the latest information on fraudulent methods. The warning unit can also reflect the meaning of specific keywords or phrases in the warning based on the related literature. Furthermore, the warning unit can also refer to related literature and issue a warning for call content that is likely to be fraudulent. In this way, referring to related literature improves the accuracy of the warning. Some or all of the above-mentioned processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit can input literature data into a generation AI and have the generation AI execute the warning.

[0060] When issuing a warning, the warning unit can take into account the market value of the call. The warning unit, for example, includes an algorithm for evaluating the market value of the call. The warning unit, for example, takes into account the market value of the call and issues a warning for important call content. The warning unit can also issue a warning for calls that are likely to be fraudulent based on the market value of the call. Furthermore, the warning unit can analyze the market value of the call and issue a warning for calls that are at a high risk of fraud. In this way, by taking the market value of the call into consideration, important call content can be given priority in being warned. Some or all of the above-mentioned processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit can input market value data into a generation AI and have the generation AI execute the warning.

[0061] When performing sentiment analysis, the sentiment analysis unit can improve the accuracy of the analysis by taking into account the context of the call. The sentiment analysis unit, for example, includes an algorithm for analyzing the meaning of a specific emotion by taking into account the context before and after the call. The sentiment analysis unit, for example, grasps the overall flow of the call and identifies important emotional parts. The sentiment analysis unit can also prioritize analysis of emotional parts that are more likely to be fraudulent based on the context of the call. This improves the accuracy of the sentiment analysis by taking into account the context of the call. Some or all of the above-mentioned processing in the sentiment analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the sentiment analysis unit can input voice data to a generation AI and have the generation AI perform context analysis.

[0062] When performing sentiment analysis, the sentiment analysis unit can take into account attribute information of the other party of the call. The sentiment analysis unit, for example, includes a database for acquiring attribute information of the other party of the call. For example, if the other party of the call is a family member or friend, the sentiment analysis unit can take into account the attribute information of the other party of the call into consideration. Furthermore, if the other party of the call is a person suspected of fraud in the past, the sentiment analysis unit can also take into account the attribute information of the other party of the call into consideration when performing sentiment analysis. Furthermore, if the other party of the call is an emergency contact, the sentiment analysis unit can also take into account the attribute information of the other party of the call into consideration when performing sentiment analysis. In this way, taking into account the attribute information of the other party of the call improves the accuracy of the sentiment analysis. Some or all of the above-described processing in the sentiment analysis unit can be performed using, for example, AI, or can be performed without using AI. For example, the sentiment analysis unit can input the other party's attribute information into a generation AI and have the generation AI perform sentiment analysis.

[0063] When performing sentiment analysis, the sentiment analysis unit can weight the analysis based on the frequency of calls. The sentiment analysis unit, for example, includes an algorithm for evaluating the frequency of calls. The sentiment analysis unit, for example, prioritizes sentiment analysis of frequently made calls. The sentiment analysis unit can also prioritize sentiment analysis of calls with frequently called parties. Furthermore, the sentiment analysis unit can also prioritize sentiment analysis of calls with high importance based on the frequency of calls. In this way, by weighting the analysis based on the frequency of calls, important calls can be prioritized in analysis. Some or all of the above-mentioned processing in the sentiment analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the sentiment analysis unit can input call frequency data to a generation AI and have the generation AI perform weighting.

[0064] The sentiment analysis unit may take into account the geographic distribution of calls when performing sentiment analysis. The sentiment analysis unit may, for example, include an algorithm for evaluating the geographic distribution of calls. For example, the sentiment analysis unit may take into account the geographic distribution of calls and prioritize sentiment analysis of calls related to a specific region. The sentiment analysis unit may also prioritize sentiment analysis of calls from regions with a high probability of fraud based on the geographic distribution of calls. Furthermore, the sentiment analysis unit may analyze the geographic distribution of calls to understand trends in fraud by region and perform sentiment analysis. This makes it easier to understand trends in fraud by region by considering the geographic distribution of calls. Some or all of the above-described processing in the sentiment analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the sentiment analysis unit may input geographic distribution data into a generation AI and have the generation AI perform sentiment analysis.

[0065] When performing sentiment analysis, the sentiment analysis unit can improve the accuracy of the analysis by referring to related literature. The sentiment analysis unit, for example, includes a database for referring to related literature. The sentiment analysis unit performs sentiment analysis based on the latest information on fraudulent methods, for example. The sentiment analysis unit can also reflect the meaning of specific keywords and phrases in the sentiment analysis based on related literature. Furthermore, the sentiment analysis unit can also refer to related literature to identify call content that is likely to be fraudulent and perform sentiment analysis. In this way, referring to related literature improves the accuracy of the sentiment analysis. Some or all of the above-mentioned processing in the sentiment analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the sentiment analysis unit can input literature data into a generation AI and have the generation AI perform sentiment analysis.

[0066] The sentiment analysis unit can take into account the market value of the call when performing sentiment analysis. The sentiment analysis unit, for example, includes an algorithm for evaluating the market value of the call. The sentiment analysis unit, for example, takes into account the market value of the call and prioritizes sentiment analysis of important call content. The sentiment analysis unit can also identify calls that are likely to be fraudulent based on the market value of the call and perform sentiment analysis on calls that are at a high risk of fraud. Furthermore, the sentiment analysis unit can analyze the market value of the call and prioritize sentiment analysis on calls that are at a high risk of fraud. In this way, by taking the market value of the call into consideration, important call content can be prioritized for analysis. Some or all of the above-mentioned processing in the sentiment analysis unit can be performed using, for example, AI, or can be performed without using AI. For example, the sentiment analysis unit can input market value data to a generation AI and have the generation AI perform sentiment analysis.

[0067] When providing a notification, the notification unit can adjust the level of detail of the notification based on the content of the call. The notification unit, for example, includes an algorithm for evaluating the content of the call. For example, the notification unit provides a detailed notification if the content of the call is related to a specific fraud method. The notification unit can also provide a concise notification if the content of the call indicates the possibility of a general fraud. Furthermore, the notification unit can also provide an immediate notification if the content of the call is urgent. In this way, by adjusting the level of detail of the notification based on the content of the call, more appropriate notifications can be provided. Some or all of the above-mentioned processing in the notification unit may be performed using, or without, AI. For example, the notification unit can input call content data to a generation AI and cause the generation AI to adjust the level of detail of the notification.

[0068] When making a notification, the notification unit can take into consideration attribute information of the other party of the call. The notification unit, for example, includes a database for acquiring attribute information of the other party of the call. For example, if the other party of the call is a family member or friend, the notification unit can take into consideration the attribute information when making the notification. Furthermore, if the other party of the call is a person suspected of fraud in the past, the notification unit can also take into consideration the attribute information when making the notification. Furthermore, if the other party of the call is an emergency contact, the notification unit can also take into consideration the attribute information when making the notification. In this way, by taking into consideration the attribute information of the other party of the call, more appropriate notifications can be provided. Some or all of the above-mentioned processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit can input the other party's attribute information into a generation AI and have the generation AI execute the notification.

[0069] When making a notification, the notification unit can weight the notification based on the frequency of the call. The notification unit includes, for example, an algorithm for evaluating the frequency of the call. The notification unit, for example, weights the notification for calls that are made frequently. The notification unit can also weight the notification for calls with a party with which the notification unit makes frequent calls. Furthermore, the notification unit can weight the notification for calls with high importance based on the frequency of the call. In this way, by weighting the notification based on the frequency of the call, important calls can be notified preferentially. Some or all of the above-mentioned processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit can input call frequency data to a generation AI and have the generation AI perform the weighting.

[0070] The notification unit may take into consideration the geographical distribution of calls when making a notification. The notification unit may, for example, include an algorithm for evaluating the geographical distribution of calls. The notification unit may, for example, take into consideration the geographical distribution of calls and make a notification for calls related to a specific region. The notification unit may also, based on the geographical distribution of calls, make a notification for calls in a region where fraud is likely to occur. Furthermore, the notification unit may analyze the geographical distribution of calls, grasp the fraud trends by region, and make a notification. In this way, by taking the geographical distribution of calls into consideration, it becomes easier to grasp the fraud trends by region. Some or all of the above-described processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit may input the geographical distribution data into a generation AI and have the generation AI execute the notification.

[0071] When issuing a notification, the notification unit can improve the accuracy of the notification by referring to related literature. The notification unit, for example, includes a database for referring to related literature. The notification unit issues a notification based on, for example, the latest information on fraudulent methods. The notification unit can also reflect the meaning of specific keywords or phrases in the notification based on the related literature. Furthermore, the notification unit can also refer to related literature and issue a notification for call content that is likely to be fraudulent. In this way, by referring to related literature, the accuracy of the notification is improved. Some or all of the above-mentioned processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit can input literature data into a generation AI and have the generation AI execute the notification.

[0072] When making a notification, the notification unit can make the notification taking into account the market value of the call. The notification unit, for example, includes an algorithm for evaluating the market value of the call. The notification unit, for example, takes into account the market value of the call and makes a notification about important call content. The notification unit can also make a notification about calls that are likely to be fraudulent based on the market value of the call. Furthermore, the notification unit can analyze the market value of the call and make a notification about calls that are at a high risk of fraud. In this way, by taking the market value of the call into consideration, important call content can be given priority in notification. Some or all of the above-mentioned processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit can input market value data into a generation AI and have the generation AI execute the notification.

[0073] During learning, the learning management unit can optimize the learning algorithm by referring to past learning data. The learning management unit, for example, includes a database for referring to past learning data. The learning management unit, for example, optimizes the learning algorithm based on the past learning data. The learning management unit can also improve the accuracy of learning by referring to past learning data. Furthermore, the learning management unit can analyze past learning data and select an optimal learning method. In this way, the accuracy of the learning algorithm is improved by referring to the past learning data. Some or all of the above-mentioned processing in the learning management unit may be performed, for example, using AI, or may be performed without using AI. For example, the learning management unit can input past learning data into the generation AI and cause the generation AI to optimize the learning algorithm.

[0074] The learning management unit can update the learning data by reflecting user feedback during learning. The learning management unit, for example, has a function for collecting user feedback and storing it in a database. The learning management unit, for example, updates the learning data based on user feedback. The learning management unit can also reflect user feedback to improve the accuracy of learning. Furthermore, the learning management unit can analyze user feedback and select the optimal learning method. In this way, the accuracy of the learning data is improved by reflecting user feedback. Some or all of the above-mentioned processing in the learning management unit may be performed, for example, using AI, or may be performed without using AI. For example, the learning management unit can input feedback data into a generation AI and have the generation AI update the learning data.

[0075] During learning, the learning management unit can weight the learning data based on the time the call was submitted. The learning management unit, for example, includes an algorithm for evaluating the time the call was submitted. The learning management unit, for example, prioritizes learning data with a more recent call submission time. The learning management unit can also learn by referring to data with an older call submission time. Furthermore, the learning management unit can weight the learning data based on the time the call was submitted. In this way, by weighting the learning data based on the time the call was submitted, the most recent information can be prioritized for learning. Some or all of the above-mentioned processing in the learning management unit may be performed using, for example, AI, or may be performed without using AI. For example, the learning management unit can input submission time data into a generation AI and have the generation AI perform weighting.

[0076] During learning, the learning management unit can integrate information from different data sources to enrich the learning data. The learning management unit, for example, includes a data analysis algorithm for integrating information from different data sources. The learning management unit, for example, integrates information from different data sources to enrich the learning data. The learning management unit can also refer to different data sources to improve the accuracy of learning. Furthermore, the learning management unit can analyze different data sources and select an optimal learning method. In this way, the accuracy of the learning data is improved by integrating information from different data sources. Some or all of the above-described processing in the learning management unit may be performed, for example, using AI or without AI. For example, the learning management unit can input different data sources into the generation AI and have the generation AI perform data integration and learning.

[0077] When performing privacy protection, the privacy protection unit can adjust the level of detail of protection based on the content of the call. The privacy protection unit, for example, includes an algorithm for evaluating the content of the call. For example, the privacy protection unit performs detailed privacy protection when the content of the call includes specific personal information. The privacy protection unit can also perform simplified privacy protection when the content of the call includes general information. Furthermore, the privacy protection unit can also perform immediate privacy protection when the content of the call is urgent. This enables more appropriate privacy protection by adjusting the level of detail of privacy protection based on the content of the call. Some or all of the above-mentioned processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit can input call content data to a generation AI and cause the generation AI to adjust the level of detail of protection.

[0078] The privacy protection unit can perform privacy protection by taking into account attribute information of the other party of the call. The privacy protection unit, for example, includes a database for acquiring attribute information of the other party of the call. For example, if the other party of the call is a family member or friend, the privacy protection unit can perform privacy protection by taking into account the attribute information. Furthermore, if the other party of the call is a person suspected of fraud in the past, the privacy protection unit can also perform privacy protection by taking into account the attribute information of the other party of the call. Furthermore, if the other party of the call is an emergency contact, the privacy protection unit can also perform privacy protection by taking into account the attribute information of the other party of the call. This enables more appropriate privacy protection by taking into account the attribute information of the other party of the call. Some or all of the above-described processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit can input the other party's attribute information into a generation AI and have the generation AI perform protection.

[0079] When performing privacy protection, the privacy protection unit can weight the protection based on the frequency of calls. The privacy protection unit includes, for example, an algorithm for evaluating the frequency of calls. The privacy protection unit, for example, weights the privacy protection for frequently made calls. The privacy protection unit can also weight the privacy protection for calls with frequently called parties. Furthermore, the privacy protection unit can weight the privacy protection for calls with high call frequency based on the frequency of calls. In this way, by weighting the protection based on the frequency of calls, important calls can be protected preferentially. Some or all of the above-mentioned processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit can input call frequency data to a generation AI and have the generation AI perform the weighting.

[0080] The privacy protection unit may perform privacy protection by taking into account the geographical distribution of calls. The privacy protection unit may, for example, include an algorithm for evaluating the geographical distribution of calls. For example, the privacy protection unit may consider the geographical distribution of calls and perform privacy protection for calls related to a specific region. The privacy protection unit may also perform privacy protection for calls in regions where fraud is more likely to occur based on the geographical distribution of calls. Furthermore, the privacy protection unit may analyze the geographical distribution of calls, identify fraud trends by region, and perform privacy protection. This makes it easier to identify fraud trends by region by taking the geographical distribution of calls into account. Some or all of the above-described processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit may input geographical distribution data into a generation AI and have the generation AI perform protection.

[0081] When performing privacy protection, the privacy protection unit can improve the accuracy of the protection by referring to related literature. The privacy protection unit, for example, includes a database for referring to related literature. The privacy protection unit performs protection based on the latest information on privacy protection, for example. The privacy protection unit can also reflect the meaning of specific keywords or phrases in the privacy protection based on the related literature. Furthermore, the privacy protection unit can also perform privacy protection for call content that is likely to be fraudulent by referring to related literature. In this way, the accuracy of privacy protection is improved by referring to related literature. Some or all of the above-mentioned processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit can input literature data into a generation AI and have the generation AI perform protection.

[0082] The privacy protection unit can perform privacy protection by taking into account the market value of the call. The privacy protection unit, for example, includes an algorithm for evaluating the market value of the call. The privacy protection unit, for example, takes into account the market value of the call and performs privacy protection for important call content. The privacy protection unit can also perform privacy protection for calls with a high probability of fraud based on the market value of the call. Furthermore, the privacy protection unit can analyze the market value of the call and perform privacy protection for calls with a high risk of fraud. In this way, important call content can be protected preferentially by taking into account the market value of the call. Some or all of the above-mentioned processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit can input market value data to a generation AI and have the generation AI perform protection.

[0083] The system according to the embodiment is not limited to the above-described example, and various modifications are possible, for example, as follows.

[0084] The analysis unit can use natural language processing (NLP) technology to understand the context of the call content. For example, the analysis unit can consider the context before and after the call to more accurately analyze the meaning of specific keywords and phrases. The analysis unit can also grasp the overall flow of the call and identify important parts. Furthermore, the analysis unit can prioritize analysis of parts with a high probability of fraud based on the context of the call. This improves the accuracy of the analysis by taking the context of the call into account.

[0085] The determination unit can refer to the user's past call history to determine the possibility of fraud. For example, the determination unit can detect patterns similar to past calls suspected of being fraudulent. The determination unit can also set an alert level for calls with specific parties based on the past call history. Furthermore, the determination unit can analyze the past call history and prioritize calls that are more likely to be fraudulent. This allows for more accurate determination of the possibility of fraud by taking the past call history into consideration.

[0086] The collection unit can collect voice data taking into account the user's geographical location information. For example, if the user is in a specific area, calls related to that area can be collected with priority. Also, if the user is traveling, calls related to the travel destination can be collected with priority. Furthermore, if the user is at home, calls related to information about the area around the user's home can be collected with priority. In this way, by taking into account the user's geographical location information, highly relevant calls can be collected with priority.

[0087] The determination unit can determine the possibility of fraud by taking into account the attribute information of the other party of the call. For example, if the other party of the call is a family member or friend, the determination unit can determine the possibility of fraud by taking into account that attribute information. Also, if the other party of the call is a person suspected of fraud in the past, the determination unit can also determine the possibility of fraud by taking into account that attribute information. Furthermore, if the other party of the call is an emergency contact, the determination unit can also determine the possibility of fraud by taking into account that attribute information. In this way, by taking into account the attribute information of the other party of the call, the possibility of fraud can be determined more accurately.

[0088] The warning unit can adjust the level of detail of the warning based on the content of the call. For example, if the content of the call is related to a specific fraud method, a detailed warning can be issued. Also, if the content of the call indicates the possibility of a general fraud, a brief warning can be issued. Furthermore, if the content of the call is urgent, an immediate warning can be issued. Thus, by adjusting the level of detail of the warning based on the content of the call, more appropriate warnings can be provided.

[0089] The collection unit can analyze the user's social media activity and collect related calls. For example, it can collect calls related to places where the user has checked in on social media. It can also collect related calls by analyzing the content of the user's social media posts. It can also collect related calls by referring to the activities of the user's friends on social media. In this way, it is possible to efficiently collect related calls by analyzing the user's social media activity.

[0090] The analysis unit can perform analysis taking into account the market value of the call. For example, the analysis unit is provided with an algorithm for evaluating the market value of the call. The analysis unit takes into account the market value of the call and prioritizes analysis of important call content. Furthermore, based on the market value of the call, it can also identify calls that are likely to be fraudulent. Furthermore, it can analyze the market value of the call and prioritize analysis of calls that have a high risk of fraud. In this way, by taking into account the market value of the call, it is possible to prioritize analysis of important call content.

[0091] The processing flow of the first embodiment will be briefly explained below.

[0092] Step 1: The collection unit collects telephone voice data. For example, the collection unit can record the contents of a call in real time and save it as voice data. It can also start recording automatically when a call is initiated. Furthermore, the collection unit can use noise filtering technology to collect voice data at high quality. Step 2: The analysis unit analyzes the voice data collected by the collection unit. For example, the analysis unit may use voice recognition technology to convert the contents of the call into text and analyze the text data. The analysis unit may also check whether specific keywords or phrases are included. It may also have a sentiment analysis function that analyzes the tone of the voice and the speaker's emotions. Step 3: The decision unit determines the possibility of fraud based on the data analyzed by the analysis unit. For example, the decision unit determines the possibility of fraud based on whether specific keywords or phrases are included. The decision unit can also determine the possibility of fraud by analyzing the tone of voice and the speaker's emotions. Step 4: The warning unit issues a warning if the judgment unit determines that there is a possibility of fraud. For example, the warning unit can sound a warning sound during the call. It can also display a warning message after the call ends. Furthermore, the warning unit can have a function to automatically notify family members or the police.

[0093] (Example 2) A system according to an embodiment of the present invention analyzes telephone conversations to prevent fraud. This system collects voice data, analyzes it using AI, determines whether a call is fraudulent, and issues a warning. For example, the system records the call in real time and saves it as audio data. Next, AI analyzes the collected audio data to determine whether a call is fraudulent. If a call is fraudulent, a warning is issued. This prevents fraudulent activity from occurring. For example, the system automatically starts recording the call as soon as it begins. Next, AI converts the call into text using speech recognition technology and analyzes the text data. For example, it checks whether specific keywords or phrases are included. If a call is determined to be fraudulent, an alarm is sounded or a warning message is displayed after the call ends. This alerts the user to the possibility of fraud. This allows the system to prevent fraudulent activity from occurring. For example, when elderly people are targets of fraud, this system can be used to prevent fraudulent activity from occurring.

[0094] The fraud prevention system according to the embodiment includes a collection unit, an analysis unit, a determination unit, and a warning unit. The collection unit collects telephone voice data. For example, the collection unit records the contents of a call in real time and saves it as voice data. The collection unit can also automatically start recording when a call is initiated. The collection unit can also use noise filtering technology to collect high-quality voice data. The analysis unit analyzes the voice data collected by the collection unit. For example, the analysis unit converts the contents of the call into text using voice recognition technology and analyzes the text data. The analysis unit can also check whether specific keywords or phrases are included. The analysis unit can also have an emotion analysis function that analyzes the tone of the voice and the speaker's emotions. The determination unit determines the possibility of fraud based on the data analyzed by the analysis unit. For example, the determination unit determines the possibility of fraud based on whether specific keywords or phrases are included. The determination unit can also determine the possibility of fraud by analyzing the tone of the voice and the speaker's emotions. The warning unit issues a warning if the determination unit determines that there is a possibility of fraud. For example, the warning unit can sound a warning sound during a call. The warning unit can also display a warning message after the call ends. Furthermore, the warning unit can have a function to automatically notify family members or the police. In this way, the fraud prevention system according to the embodiment can prevent fraudulent acts before they occur. For example, the system can prevent fraud damage by detecting potentially fraudulent calls in real time and issuing a warning to the user.

[0095] The analysis unit can convert the call content into text using speech recognition technology and analyze the text data. Examples of speech recognition technology include speech recognition algorithms using deep learning. The analysis unit can convert the call content into text with high accuracy using speech recognition algorithms. The analysis unit can also extract specific keywords and phrases from the call content using speech recognition technology. Furthermore, the analysis unit can understand the context of the call content and determine the possibility of fraud with high accuracy using speech recognition technology. This improves the accuracy of analyzing the call content by using speech recognition technology. Some or all of the above-described processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input voice data into a generation AI and have the generation AI convert the voice data into text data.

[0096] The determination unit can check whether specific keywords or phrases are included. For example, the determination unit uses a list of keywords related to fraud to check whether specific keywords or phrases are included in the call content. The determination unit can also periodically update the keyword list to respond to new fraudulent methods. Furthermore, the determination unit can determine the possibility of fraud based on the frequency of appearance of keywords and phrases. In this way, by checking for specific keywords and phrases, the possibility of fraud can be determined with high accuracy. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input text data into a generation AI and have the generation AI check for keywords and phrases.

[0097] The determination unit may include an emotion analysis unit that analyzes the tone of the voice and the speaker's emotion. The emotion analysis unit may include, for example, an algorithm for analyzing the tone of the voice and the speaker's emotion. The emotion analysis unit may analyze, for example, the pitch, volume, rhythm, etc. of the voice to estimate the speaker's emotion. The emotion analysis unit may also classify the speaker's emotion and use the resulting data to determine the possibility of fraud. Furthermore, the emotion analysis unit may monitor changes in the speaker's emotion in real time to detect signs of fraud. This allows for more accurate determination of the possibility of fraud by analyzing the tone of the voice and the speaker's emotion. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or a generative AI. The generative AI may be, for example, a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. Some or all of the above-described processing in the determination unit may be performed using, for example, an AI. For example, the determination unit may input voice data to a generative AI and have the generative AI perform emotion analysis.

[0098] The warning unit can sound a warning sound during a call or display a warning message after the call ends. The warning unit, for example, includes a speaker for sounding a warning sound during a call. The warning unit can sound a warning sound during a call, for example, if it determines that there is a possibility of fraud. The warning unit can also include a display for displaying a warning message after the call ends. For example, the warning unit can display a warning message after the call ends to notify the user of the possibility of fraud. The warning unit can also include a function for customizing the content of the warning sound or warning message. This allows the user to be notified of the possibility of fraud by issuing a warning during or after the call ends. Some or all of the above-mentioned processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit can input the determination result to a generation AI and cause the generation AI to generate a warning sound or warning message.

[0099] The warning unit may include a notification unit that automatically notifies family members or the police. The notification unit may include, for example, a communication module for automatically notifying family members or the police when there is a possibility of fraud. The notification unit may automatically send a notification to family members or the police when it determines there is a possibility of fraud. The notification unit may also include a function for customizing the content of the notification. For example, the notification unit may allow the user to set the content of the notification message. Furthermore, the notification unit may also set multiple notification destinations. This enables a prompt response by automatically notifying family members or the police when there is a possibility of fraud. Some or all of the above-mentioned processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit may input the determination result into a generation AI and cause the generation AI to generate a notification message.

[0100] The analysis unit may include a learning management unit that uses past fraud cases as learning data. The learning management unit may, for example, include a database for using past fraud cases as learning data. The learning management unit may, for example, collect past fraud cases and store them in the database. The learning management unit may also analyze the collected fraud cases to learn fraud patterns. Furthermore, the learning management unit may periodically update the learning data to respond to new fraud methods. In this way, using past fraud cases as learning data improves analysis accuracy. Some or all of the above-mentioned processing in the learning management unit may be performed, for example, using AI, or may be performed without using AI. For example, the learning management unit may input fraud case data into a generation AI and cause the generation AI to perform learning.

[0101] The system may include a privacy protection unit that protects user privacy. The privacy protection unit may, for example, include data anonymization technology to protect user privacy. The privacy protection unit may, for example, anonymize collected voice data to protect personal information. The privacy protection unit may also include an access control function to allow only specific users to access the data. Furthermore, the privacy protection unit may also include a function to set a data storage period and automatically delete the data after a certain period. This protects user privacy, allowing the system to be used with peace of mind. Some or all of the above-mentioned processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit may input voice data into a generation AI and have the generation AI perform anonymization processing.

[0102] The collection unit can estimate the user's emotions and adjust the timing of collecting voice data based on the estimated user emotions. The collection unit, for example, includes an emotion estimation algorithm for estimating the user's emotions. The collection unit, for example, analyzes the user's facial expressions and voice tone to estimate the emotions. The collection unit can also adjust the timing of collecting voice data based on the estimated emotions. For example, if the user is nervous, voice data can be collected immediately after the start of a call. Alternatively, if the user is relaxed, voice data can be collected when the call content gets into a specific topic. Furthermore, if the user is in a hurry, only important parts of the call can be collected preferentially. This enables more appropriate data collection by adjusting the timing of collecting voice data according to the user's emotions. Emotion estimation is realized using an emotion estimation function, for example, using an emotion engine or a generation AI. The generation AI can be, but is not limited to, a text generation AI (e.g., LLM) or a multimodal generation AI. Some or all of the above-mentioned processing in the collection unit can be performed using, for example, an AI, or without an AI. For example, the collection unit can input the user's facial expression data into the generation AI and have the generation AI perform emotion estimation.

[0103] The collection unit not only automatically collects voice data at the start of a call, but also can collect additional voice data when a specific keyword is detected. The collection unit, for example, has a function for automatically collecting voice data at the start of a call. The collection unit, for example, automatically starts recording when the call starts. The collection unit can also have a function for collecting additional voice data when a specific keyword is detected. For example, the collection unit collects additional voice data when a specific keyword, such as "money" or "transfer," is detected during a call. Furthermore, the collection unit can also immediately collect additional voice data when a highly urgent keyword, such as "help," is detected during a call. In this way, by collecting additional voice data when a specific keyword is detected, important information can be collected without missing any important information. Some or all of the above-described processing by the collection unit may be performed using, or without, AI. For example, the collection unit may input voice data to a generation AI and cause the generation AI to detect specific keywords and collect additional voice data.

[0104] The collection unit can filter background sounds during a call to remove noise when collecting voice data. The collection unit, for example, includes a noise removal algorithm for filtering background sounds during a call to remove noise. The collection unit, for example, filters background sounds during a call (e.g., traffic noise, wind noise) in real time to collect clear voice data. The collection unit can also remove noise during a call (e.g., static, echo) and collect only important conversation portions. Furthermore, the collection unit can filter environmental sounds during a call (e.g., television sounds, other conversations) to preferentially collect conversation portions that are more likely to be fraudulent. This allows clear voice data to be collected by filtering background sounds during a call and removing noise. Some or all of the above-described processing in the collection unit may be performed using, or without, AI. For example, the collection unit can input voice data to a generation AI and have the generation AI perform noise removal.

[0105] When collecting voice data, the collection unit can prioritize collection when the caller is from a specific number. The collection unit, for example, has a function for prioritized collection of voice data when the caller is from a specific number. For example, the collection unit prioritizes collection of voice data when the caller is from a specific number, such as a family member or friend. The collection unit can also prioritize collection of voice data when the caller is from a number suspected of fraud in the past. Furthermore, the collection unit can also prioritize collection of voice data when the caller is from a number registered as an emergency contact. In this way, by prioritizing collection of calls from specific numbers, important call content can be reliably collected. Some or all of the above-described processing in the collection unit may be performed using, for example, AI, or may be performed without AI. For example, the collection unit can input call number data to a generation AI and cause the generation AI to detect specific numbers and prioritize collection of voice data.

[0106] The collection unit can estimate the user's emotions and prioritize the voice data to be collected based on the estimated user emotions. The collection unit, for example, includes an emotion estimation algorithm for estimating the user's emotions. The collection unit, for example, analyzes the user's facial expressions and voice tone to estimate the emotions. The collection unit can also prioritize the voice data to be collected based on the estimated emotions. For example, if the user is nervous, important parts of the conversation can be collected preferentially. Also, if the user is relaxed, the overall conversation content can be collected in a balanced manner. Furthermore, if the user is in a hurry, parts containing important information can be collected preferentially in a short time. In this way, important information can be collected preferentially by prioritizing the voice data according to the user's emotions. Emotion estimation is realized using an emotion estimation function, for example, using an emotion engine or a generation AI. The generation AI can be a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. Some or all of the above-mentioned processing in the collection unit can be performed, for example, using AI or without AI. For example, the collection unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and determine the priority of voice data.

[0107] When collecting voice data, the collection unit can prioritize collection of highly relevant calls by taking into account the user's geographical location information. The collection unit, for example, includes a GPS module for acquiring the user's geographical location information. For example, when the user is in a specific area, the collection unit prioritizes collection of calls related to that area. Furthermore, when the user is traveling, the collection unit can also prioritize collection of calls related to the user's travel destination. Furthermore, when the user is at home, the collection unit can also prioritize collection of calls related to information around the user's home. In this way, highly relevant calls can be collected preferentially by taking into account the user's geographical location information. Some or all of the above-described processing in the collection unit may be performed using AI, for example, or may be performed without using AI. For example, the collection unit can input geographical location data to the generation AI and cause the generation AI to prioritize collection of highly relevant calls.

[0108] The collection unit can analyze the user's social media activity and collect related calls when collecting voice data. The collection unit, for example, includes a data analysis algorithm for analyzing the user's social media activity. The collection unit, for example, collects calls related to places where the user has checked in on social media. The collection unit can also analyze the content of the user's social media posts and collect related calls. Furthermore, the collection unit can collect related calls by referring to the activities of the user's friends on social media. In this way, related calls can be efficiently collected by analyzing the user's social media activity. Some or all of the above-mentioned processing in the collection unit may be performed using, for example, AI, or may be performed without using AI. For example, the collection unit can input social media data into a generation AI and cause the generation AI to collect related calls.

[0109] The collection unit can customize the collection method by reflecting the user's past feedback when collecting voice data. The collection unit, for example, has a function for collecting the user's past feedback and storing it in a database. The collection unit, for example, optimizes the collection method based on feedback provided by the user in the past. The collection unit can also prioritize collection of call content that the user has previously determined to be important. Furthermore, the collection unit can also reflect the user's past feedback and adjust the collection timing and collection range. In this way, the collection method can be optimized by reflecting the user's past feedback. Some or all of the above-mentioned processing in the collection unit may be performed using, for example, AI, or may be performed without using AI. For example, the collection unit can input feedback data to a generation AI and cause the generation AI to customize the collection method.

[0110] The analysis unit can estimate the user's emotion and adjust the presentation method of the analysis based on the estimated user's emotion. The analysis unit, for example, includes an emotion estimation algorithm for estimating the user's emotion. The analysis unit, for example, analyzes the user's facial expression and tone of voice to estimate the emotion. The analysis unit can also adjust the presentation method of the analysis based on the estimated emotion. For example, if the user is nervous, a simple and highly visible analysis result can be provided. On the other hand, if the user is relaxed, a detailed analysis result can be provided. Furthermore, if the user is in a hurry, an analysis result that focuses on the main points can be provided. This allows for adjusting the presentation method of the analysis according to the user's emotion to provide more appropriate analysis results. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or a generation AI. The generation AI can be a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to these examples. Some or all of the above-mentioned processing in the analysis unit can be performed using, for example, AI, or without AI. For example, the analysis unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and adjust the way the analysis is expressed.

[0111] When analyzing voice data, the analysis unit can improve the accuracy of the analysis by taking into account the context of the call. The analysis unit, for example, includes an algorithm for analyzing the meaning of specific keywords by taking into account the context before and after the call. The analysis unit, for example, grasps the overall flow of the call and identifies important parts. The analysis unit can also prioritize analysis of parts that are more likely to be fraudulent based on the context of the call. This improves the accuracy of the analysis by taking into account the context of the call. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input voice data to a generation AI and have the generation AI perform context analysis.

[0112] When analyzing voice data, the analysis unit can perform the analysis taking into account attribute information of the caller. The analysis unit, for example, includes a database for acquiring attribute information of the caller. For example, if the caller is a family member or friend, the analysis unit performs the analysis taking into account the attribute information. Furthermore, if the caller is a person suspected of fraud in the past, the analysis unit can also perform the analysis taking into account the attribute information. Furthermore, if the caller is an emergency contact, the analysis unit can also perform the analysis taking into account the attribute information. In this way, by taking into account the attribute information of the caller, the accuracy of the analysis is improved. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input the attribute information of the caller into a generation AI and have the generation AI perform the analysis.

[0113] When analyzing the voice data, the analysis unit can weight the analysis based on the frequency of calls. The analysis unit includes, for example, an algorithm for evaluating the frequency of calls. The analysis unit, for example, prioritizes analysis of frequently made calls. The analysis unit can also prioritize analysis of calls with frequently called parties. Furthermore, the analysis unit can also prioritize analysis of calls with high importance based on the frequency of calls. In this way, by weighting the analysis based on the frequency of calls, important calls can be prioritized in analysis. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input call frequency data to a generation AI and have the generation AI perform the weighting.

[0114] The analysis unit can estimate the user's emotion and adjust the display order of the analysis results based on the estimated user's emotion. The analysis unit, for example, includes an emotion estimation algorithm for estimating the user's emotion. The analysis unit, for example, analyzes the user's facial expression and tone of voice to estimate the emotion. The analysis unit can also adjust the display order of the analysis results based on the estimated emotion. For example, if the user is nervous, important analysis results can be displayed first. Alternatively, if the user is relaxed, detailed analysis results can be displayed in an orderly manner. Furthermore, if the user is in a hurry, analysis results that highlight the main points can be displayed first. This allows for more appropriate information provision by adjusting the display order of the analysis results according to the user's emotion. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or a generation AI. The generation AI can be, but is not limited to, a text generation AI (e.g., LLM) or a multimodal generation AI. Some or all of the above-described processing in the analysis unit can be performed using, for example, an AI, or without an AI. For example, the analysis unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and adjust the display order of the analysis results.

[0115] The analysis unit can take into account the geographical distribution of calls when analyzing voice data. The analysis unit, for example, includes an algorithm for evaluating the geographical distribution of calls. For example, the analysis unit takes into account the geographical distribution of calls and prioritizes analysis of calls related to a specific region. The analysis unit can also prioritize analysis of calls from regions with a high probability of fraud based on the geographical distribution of calls. Furthermore, the analysis unit can analyze the geographical distribution of calls to identify fraud trends by region. This improves the accuracy of the analysis by taking the geographical distribution of calls into account. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input geographical distribution data into a generation AI and have the generation AI perform the analysis.

[0116] When analyzing voice data, the analysis unit can improve the accuracy of the analysis by referring to related literature. The analysis unit, for example, includes a database for referring to related literature. The analysis unit performs analysis based on, for example, the latest information on fraudulent methods. The analysis unit can also analyze the meaning of specific keywords and phrases based on related literature. Furthermore, the analysis unit can also identify call content that is likely to be fraudulent by referring to related literature. In this way, referring to related literature improves the accuracy of the analysis. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input literature data into a generation AI and have the generation AI perform the analysis.

[0117] The analysis unit can take into account the market value of the call when analyzing the voice data. The analysis unit, for example, includes an algorithm for evaluating the market value of the call. The analysis unit, for example, takes into account the market value of the call and prioritizes analysis of important call content. The analysis unit can also identify calls that are likely to be fraudulent based on the market value of the call. Furthermore, the analysis unit can analyze the market value of the call and prioritize analysis of calls that have a high risk of fraud. In this way, by taking the market value of the call into consideration, important call content can be prioritized for analysis. Some or all of the above-mentioned processing in the analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the analysis unit can input market value data to a generation AI and have the generation AI perform the analysis.

[0118] The determination unit can estimate the user's emotions and adjust the criteria for determining the possibility of fraud based on the estimated user emotions. The determination unit, for example, includes an emotion estimation algorithm for estimating the user's emotions. The determination unit, for example, analyzes the user's facial expressions and tone of voice to estimate the emotions. The determination unit can also adjust the criteria for determining the possibility of fraud based on the estimated emotions. For example, if the user is nervous, the criteria for determining that there is a high possibility of fraud can be tightened. Also, if the user is relaxed, the criteria for determining that there is a low possibility of fraud can be loosened. Furthermore, if the user is in a hurry, the criteria for determining that there is a high possibility of fraud can be tightened. This enables more accurate fraud detection by adjusting the criteria according to the user's emotions. Emotion estimation is realized using an emotion estimation function, for example, an emotion engine or a generation AI. The generation AI can be, but is not limited to, a text generation AI (e.g., LLM) or a multimodal generation AI. Some or all of the above-mentioned processing in the determination unit can be performed using, for example, AI, or without AI. For example, the judgment unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and adjust the judgment criteria.

[0119] When determining the possibility of fraud, the determination unit can improve the accuracy of the determination by taking into account the interrelationships between calls. The determination unit, for example, includes an algorithm for evaluating the interrelationships between calls. The determination unit, for example, determines the possibility of fraud by taking into account the content of conversations before and after a call. The determination unit can also determine the possibility of fraud by taking into account past call history with the call partner. Furthermore, the determination unit can analyze the interrelationships between calls and identify calls that are likely to be fraudulent. In this way, by taking the interrelationships between calls into account, the possibility of fraud can be determined more accurately. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input call history data into a generation AI and cause the generation AI to analyze the interrelationships and determine the possibility of fraud.

[0120] When determining the possibility of fraud, the determination unit can make the determination taking into account attribute information of the other party of the call. The determination unit, for example, includes a database for acquiring attribute information of the other party of the call. For example, if the other party of the call is a family member or friend, the determination unit determines the possibility of fraud taking into account the attribute information. Furthermore, if the other party of the call is a person suspected of fraud in the past, the determination unit can also determine the possibility of fraud taking into account the attribute information. Furthermore, if the other party of the call is an emergency contact, the determination unit can also determine the possibility of fraud taking into account the attribute information. Thus, by taking into account the attribute information of the other party of the call, the possibility of fraud can be more accurately determined. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input the other party's attribute information into a generation AI and cause the generation AI to determine the possibility of fraud.

[0121] When determining the possibility of fraud, the determination unit can weight the determination based on the frequency of calls. The determination unit includes, for example, an algorithm for evaluating the frequency of calls. The determination unit, for example, prioritizes calls that are made frequently. The determination unit can also prioritize calls with partners who are called frequently. Furthermore, the determination unit can also prioritize calls with high importance based on the frequency of calls. In this way, by weighting the determination based on the frequency of calls, important calls can be prioritized. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input call frequency data to a generation AI and have the generation AI perform the weighting.

[0122] The determination unit can estimate the user's emotion and adjust the display method of the determination result based on the estimated user's emotion. The determination unit, for example, includes an emotion estimation algorithm for estimating the user's emotion. The determination unit, for example, analyzes the user's facial expression and voice tone to estimate the emotion. The determination unit can also adjust the display method of the determination result based on the estimated emotion. For example, if the user is nervous, a simple, highly visible display method can be provided. If the user is relaxed, a display method including detailed information can be provided. Furthermore, if the user is in a hurry, a display method that focuses on the main points can be provided. This allows for more appropriate information provision by adjusting the display method of the determination result according to the user's emotion. Emotion estimation is realized using an emotion estimation function, for example, using an emotion engine or a generation AI. The generation AI can be, but is not limited to, a text generation AI (e.g., LLM) or a multimodal generation AI. Some or all of the above-mentioned processing in the determination unit can be performed using, for example, an AI, or without an AI. For example, the judgment unit can input the user's facial expression data into the generation AI and have the generation AI estimate the emotion and adjust the display method.

[0123] The determination unit can make a determination regarding the possibility of fraud by taking into account the geographic distribution of calls. The determination unit, for example, includes an algorithm for evaluating the geographic distribution of calls. The determination unit, for example, takes into account the geographic distribution of calls and prioritizes calls related to a specific region. The determination unit can also prioritize calls from regions with a high probability of fraud based on the geographic distribution of calls. Furthermore, the determination unit can analyze the geographic distribution of calls and understand fraud trends by region. This allows for a more accurate determination of the possibility of fraud by taking the geographic distribution of calls into account. Some or all of the above-described processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input geographic distribution data into a generation AI and have the generation AI perform the determination.

[0124] When determining the possibility of fraud, the determination unit can improve the accuracy of the determination by referring to related literature. The determination unit, for example, includes a database for referring to related literature. The determination unit makes the determination based on, for example, the latest information on fraudulent methods. The determination unit can also determine the meaning of specific keywords or phrases based on related literature. Furthermore, the determination unit can also identify call content that is likely to be fraudulent by referring to related literature. In this way, the accuracy of the determination is improved by referring to related literature. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input literature data into a generation AI and have the generation AI perform the determination.

[0125] The determination unit can make a determination taking into account the market value of the call when determining the possibility of fraud. The determination unit, for example, includes an algorithm for evaluating the market value of the call. The determination unit, for example, takes into account the market value of the call and prioritizes determining important call content. The determination unit can also identify calls that are likely to be fraudulent based on the market value of the call. Furthermore, the determination unit can analyze the market value of the call and prioritize determining calls that have a high risk of fraud. In this way, important call content can be prioritized by considering the market value of the call. Some or all of the above-mentioned processing in the determination unit may be performed using, for example, AI, or may be performed without using AI. For example, the determination unit can input market value data to a generation AI and have the generation AI perform the determination.

[0126] The warning unit can estimate the user's emotion and adjust the warning method based on the estimated user's emotion. The warning unit, for example, includes an emotion estimation algorithm for estimating the user's emotion. The warning unit, for example, analyzes the user's facial expression and voice tone to estimate the emotion. The warning unit can also adjust the warning method based on the estimated emotion. For example, if the user is nervous, the warning can be issued in a calm voice. If the user is relaxed, the warning can be issued in a cheerful voice. Furthermore, if the user is in a hurry, a quick and concise warning can be issued. This allows for adjusting the warning method according to the user's emotion, thereby providing a more appropriate warning. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or a generation AI. The generation AI can be, for example, a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. Some or all of the above-mentioned processing in the warning unit can be performed, for example, using AI or without AI. For example, the warning unit can input the user's facial expression data into the generation AI and have the generation AI estimate the emotion and adjust the warning method.

[0127] When issuing a warning, the warning unit can adjust the level of detail of the warning based on the content of the call. The warning unit, for example, includes an algorithm for evaluating the content of the call. For example, the warning unit issues a detailed warning if the content of the call is related to a specific fraud method. The warning unit can also issue a concise warning if the content of the call indicates the possibility of a general fraud. Furthermore, the warning unit can also issue an immediate warning if the content of the call is urgent. In this way, by adjusting the level of detail of the warning based on the content of the call, a more appropriate warning can be provided. Some or all of the above-mentioned processing in the warning unit may be performed using AI, for example, or may be performed without using AI. For example, the warning unit can input call content data to a generation AI and cause the generation AI to adjust the level of detail of the warning.

[0128] When issuing a warning, the warning unit can take into account attribute information of the other party of the call. The warning unit, for example, includes a database for acquiring attribute information of the other party of the call. For example, if the other party of the call is a family member or a friend, the warning unit can take into account the attribute information when issuing the warning. Furthermore, if the other party of the call is a person suspected of fraud in the past, the warning unit can also take into account the attribute information when issuing the warning. Furthermore, if the other party of the call is an emergency contact, the warning unit can also take into account the attribute information when issuing the warning. In this way, by taking into account the attribute information of the other party of the call, a more appropriate warning can be provided. Some or all of the above-mentioned processing in the warning unit may be performed, for example, using AI, or may be performed without using AI. For example, the warning unit can input the other party's attribute information into a generation AI and have the generation AI execute the warning.

[0129] When issuing a warning, the warning unit can weight the warning based on the frequency of calls. The warning unit includes, for example, an algorithm for evaluating the frequency of calls. The warning unit, for example, weights the warning for calls that are made frequently. The warning unit can also weight the warning for calls with a party with whom the number of calls is high. Furthermore, the warning unit can weight the warning for calls with high importance based on the frequency of calls. In this way, by weighting the warning based on the frequency of calls, important calls can be given priority in being warned. Some or all of the above-mentioned processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit can input call frequency data to a generation AI and have the generation AI perform the weighting.

[0130] The warning unit can estimate the user's emotion and determine the priority of warnings based on the estimated user's emotion. The warning unit, for example, includes an emotion estimation algorithm for estimating the user's emotion. The warning unit, for example, analyzes the user's facial expression and tone of voice to estimate the emotion. The warning unit can also determine the priority of warnings based on the estimated emotion. For example, if the user is nervous, an important warning can be issued first. If the user is relaxed, detailed warnings can be issued in an orderly manner. Furthermore, if the user is in a hurry, a warning that covers the main points can be issued first. In this way, by determining the priority of warnings according to the user's emotion, important warnings can be provided first. Emotion estimation is realized using an emotion estimation function, for example, using an emotion engine or a generation AI. The generation AI can be a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. Some or all of the above-mentioned processing in the warning unit can be performed, for example, using AI or without AI. For example, the warning unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and determine the priority of warnings.

[0131] The warning unit may issue a warning taking into account the geographical distribution of calls. The warning unit may, for example, include an algorithm for evaluating the geographical distribution of calls. The warning unit may, for example, take into account the geographical distribution of calls and issue a warning for calls related to a specific region. The warning unit may also issue a warning for calls from regions where there is a high possibility of fraud based on the geographical distribution of calls. Furthermore, the warning unit may analyze the geographical distribution of calls, identify fraud trends by region, and issue a warning. In this way, by taking the geographical distribution of calls into account, it becomes easier to identify fraud trends by region. Some or all of the above-described processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit may input geographical distribution data into a generation AI and cause the generation AI to execute the warning.

[0132] When issuing a warning, the warning unit can improve the accuracy of the warning by referring to related literature. The warning unit, for example, includes a database for referring to related literature. The warning unit issues a warning based on, for example, the latest information on fraudulent methods. The warning unit can also reflect the meaning of specific keywords or phrases in the warning based on the related literature. Furthermore, the warning unit can also refer to related literature and issue a warning for call content that is likely to be fraudulent. In this way, referring to related literature improves the accuracy of the warning. Some or all of the above-mentioned processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit can input literature data into a generation AI and have the generation AI execute the warning.

[0133] When issuing a warning, the warning unit can take into account the market value of the call. The warning unit, for example, includes an algorithm for evaluating the market value of the call. The warning unit, for example, takes into account the market value of the call and issues a warning for important call content. The warning unit can also issue a warning for calls that are likely to be fraudulent based on the market value of the call. Furthermore, the warning unit can analyze the market value of the call and issue a warning for calls that are at a high risk of fraud. In this way, by taking the market value of the call into consideration, important call content can be given priority in being warned. Some or all of the above-mentioned processing in the warning unit may be performed using, for example, AI, or may be performed without using AI. For example, the warning unit can input market value data into a generation AI and have the generation AI execute the warning.

[0134] The emotion analysis unit can estimate a user's emotion and adjust the emotion analysis method based on the estimated user emotion. The emotion analysis unit, for example, includes an emotion estimation algorithm for estimating a user's emotion. The emotion analysis unit, for example, analyzes the user's facial expressions and tone of voice to estimate the emotion. The emotion analysis unit can also adjust the emotion analysis method based on the estimated emotion. For example, if the user is nervous, a simple, highly visible emotion analysis result can be provided. On the other hand, if the user is relaxed, a detailed emotion analysis result can be provided. Furthermore, if the user is in a hurry, a more concise emotion analysis result can be provided. This allows for adjusting the emotion analysis method according to the user's emotion to provide a more appropriate emotion analysis result. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or a generative AI. The generative AI can be, but is not limited to, a text generation AI (e.g., LLM) or a multimodal generation AI. Some or all of the above-described processing in the emotion analysis unit can be performed using, for example, an AI, or without an AI. For example, the emotion analysis unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and adjust the emotion analysis method.

[0135] When performing sentiment analysis, the sentiment analysis unit can improve the accuracy of the analysis by taking into account the context of the call. The sentiment analysis unit, for example, includes an algorithm for analyzing the meaning of a specific emotion by taking into account the context before and after the call. The sentiment analysis unit, for example, grasps the overall flow of the call and identifies important emotional parts. The sentiment analysis unit can also prioritize analysis of emotional parts that are more likely to be fraudulent based on the context of the call. This improves the accuracy of the sentiment analysis by taking into account the context of the call. Some or all of the above-mentioned processing in the sentiment analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the sentiment analysis unit can input voice data to a generation AI and have the generation AI perform context analysis.

[0136] When performing sentiment analysis, the sentiment analysis unit can take into account attribute information of the other party of the call. The sentiment analysis unit, for example, includes a database for acquiring attribute information of the other party of the call. For example, if the other party of the call is a family member or friend, the sentiment analysis unit can take into account the attribute information of the other party of the call into consideration. Furthermore, if the other party of the call is a person suspected of fraud in the past, the sentiment analysis unit can also take into account the attribute information of the other party of the call into consideration when performing sentiment analysis. Furthermore, if the other party of the call is an emergency contact, the sentiment analysis unit can also take into account the attribute information of the other party of the call into consideration when performing sentiment analysis. In this way, taking into account the attribute information of the other party of the call improves the accuracy of the sentiment analysis. Some or all of the above-described processing in the sentiment analysis unit can be performed using, for example, AI, or can be performed without using AI. For example, the sentiment analysis unit can input the other party's attribute information into a generation AI and have the generation AI perform sentiment analysis.

[0137] When performing sentiment analysis, the sentiment analysis unit can weight the analysis based on the frequency of calls. The sentiment analysis unit, for example, includes an algorithm for evaluating the frequency of calls. The sentiment analysis unit, for example, prioritizes sentiment analysis of frequently made calls. The sentiment analysis unit can also prioritize sentiment analysis of calls with frequently called parties. Furthermore, the sentiment analysis unit can also prioritize sentiment analysis of calls with high importance based on the frequency of calls. In this way, by weighting the analysis based on the frequency of calls, important calls can be prioritized in analysis. Some or all of the above-mentioned processing in the sentiment analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the sentiment analysis unit can input call frequency data to a generation AI and have the generation AI perform weighting.

[0138] The emotion analysis unit can estimate a user's emotion and adjust the display order of the emotion analysis results based on the estimated user emotion. The emotion analysis unit, for example, includes an emotion estimation algorithm for estimating the user's emotion. The emotion analysis unit, for example, analyzes the user's facial expressions and voice tone to estimate the emotion. The emotion analysis unit can also adjust the display order of the emotion analysis results based on the estimated emotion. For example, if the user is nervous, important emotion analysis results can be displayed first. Alternatively, if the user is relaxed, detailed emotion analysis results can be displayed in an orderly manner. Furthermore, if the user is in a hurry, emotion analysis results that focus on the main points can be displayed first. This allows for more appropriate information provision by adjusting the display order of the emotion analysis results according to the user's emotion. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or a generation AI. The generation AI can be, but is not limited to, a text generation AI (e.g., LLM) or a multimodal generation AI. Some or all of the above-described processing in the emotion analysis unit can be performed using, for example, an AI, or without an AI. For example, the emotion analysis unit can input the user's facial expression data into the generation AI and have the generation AI estimate the emotion and adjust the display order.

[0139] The sentiment analysis unit may take into account the geographic distribution of calls when performing sentiment analysis. The sentiment analysis unit may, for example, include an algorithm for evaluating the geographic distribution of calls. For example, the sentiment analysis unit may take into account the geographic distribution of calls and prioritize sentiment analysis of calls related to a specific region. The sentiment analysis unit may also prioritize sentiment analysis of calls from regions with a high probability of fraud based on the geographic distribution of calls. Furthermore, the sentiment analysis unit may analyze the geographic distribution of calls to understand trends in fraud by region and perform sentiment analysis. This makes it easier to understand trends in fraud by region by considering the geographic distribution of calls. Some or all of the above-described processing in the sentiment analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the sentiment analysis unit may input geographic distribution data into a generation AI and have the generation AI perform sentiment analysis.

[0140] When performing sentiment analysis, the sentiment analysis unit can improve the accuracy of the analysis by referring to related literature. The sentiment analysis unit, for example, includes a database for referring to related literature. The sentiment analysis unit performs sentiment analysis based on the latest information on fraudulent methods, for example. The sentiment analysis unit can also reflect the meaning of specific keywords and phrases in the sentiment analysis based on related literature. Furthermore, the sentiment analysis unit can also refer to related literature to identify call content that is likely to be fraudulent and perform sentiment analysis. In this way, referring to related literature improves the accuracy of the sentiment analysis. Some or all of the above-mentioned processing in the sentiment analysis unit may be performed using, for example, AI, or may be performed without using AI. For example, the sentiment analysis unit can input literature data into a generation AI and have the generation AI perform sentiment analysis.

[0141] The sentiment analysis unit can take into account the market value of the call when performing sentiment analysis. The sentiment analysis unit, for example, includes an algorithm for evaluating the market value of the call. The sentiment analysis unit, for example, takes into account the market value of the call and prioritizes sentiment analysis of important call content. The sentiment analysis unit can also identify calls that are likely to be fraudulent based on the market value of the call and perform sentiment analysis on calls that are at a high risk of fraud. Furthermore, the sentiment analysis unit can analyze the market value of the call and prioritize sentiment analysis on calls that are at a high risk of fraud. In this way, by taking the market value of the call into consideration, important call content can be prioritized for analysis. Some or all of the above-mentioned processing in the sentiment analysis unit can be performed using, for example, AI, or can be performed without using AI. For example, the sentiment analysis unit can input market value data to a generation AI and have the generation AI perform sentiment analysis.

[0142] The notification unit can estimate the user's emotion and adjust the notification method based on the estimated user's emotion. The notification unit, for example, includes an emotion estimation algorithm for estimating the user's emotion. The notification unit, for example, analyzes the user's facial expression and tone of voice to estimate the emotion. The notification unit can also adjust the notification method based on the estimated emotion. For example, if the user is nervous, the notification can be made in a calm voice. Alternatively, if the user is relaxed, the notification can be made in a cheerful voice. Furthermore, if the user is in a hurry, the notification can be made quickly and concisely. This allows for more appropriate notification by adjusting the notification method according to the user's emotion. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or generation AI. The generation AI can be a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. Some or all of the above-mentioned processing in the notification unit can be performed, for example, using AI or without AI. For example, the notification unit can input the user's facial expression data into the generation AI and have the generation AI estimate the emotion and adjust the notification method.

[0143] When providing a notification, the notification unit can adjust the level of detail of the notification based on the content of the call. The notification unit, for example, includes an algorithm for evaluating the content of the call. For example, the notification unit provides a detailed notification if the content of the call is related to a specific fraud method. The notification unit can also provide a concise notification if the content of the call indicates the possibility of a general fraud. Furthermore, the notification unit can also provide an immediate notification if the content of the call is urgent. In this way, by adjusting the level of detail of the notification based on the content of the call, more appropriate notifications can be provided. Some or all of the above-mentioned processing in the notification unit may be performed using, or without, AI. For example, the notification unit can input call content data to a generation AI and cause the generation AI to adjust the level of detail of the notification.

[0144] When making a notification, the notification unit can take into consideration attribute information of the other party of the call. The notification unit, for example, includes a database for acquiring attribute information of the other party of the call. For example, if the other party of the call is a family member or friend, the notification unit can take into consideration the attribute information when making the notification. Furthermore, if the other party of the call is a person suspected of fraud in the past, the notification unit can also take into consideration the attribute information when making the notification. Furthermore, if the other party of the call is an emergency contact, the notification unit can also take into consideration the attribute information when making the notification. In this way, by taking into consideration the attribute information of the other party of the call, more appropriate notifications can be provided. Some or all of the above-mentioned processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit can input the other party's attribute information into a generation AI and have the generation AI execute the notification.

[0145] When making a notification, the notification unit can weight the notification based on the frequency of the call. The notification unit includes, for example, an algorithm for evaluating the frequency of the call. The notification unit, for example, weights the notification for calls that are made frequently. The notification unit can also weight the notification for calls with a party with which the notification unit makes frequent calls. Furthermore, the notification unit can weight the notification for calls with high importance based on the frequency of the call. In this way, by weighting the notification based on the frequency of the call, important calls can be notified preferentially. Some or all of the above-mentioned processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit can input call frequency data to a generation AI and have the generation AI perform the weighting.

[0146] The notification unit can estimate the user's emotion and determine the priority of notifications based on the estimated user's emotion. The notification unit, for example, includes an emotion estimation algorithm for estimating the user's emotion. The notification unit, for example, analyzes the user's facial expression and tone of voice to estimate the emotion. The notification unit can also determine the priority of notifications based on the estimated emotion. For example, if the user is nervous, important notifications can be given priority. Also, if the user is relaxed, detailed notifications can be given in an orderly manner. Furthermore, if the user is in a hurry, notifications that focus on the main points can be given priority. In this way, by determining the priority of notifications according to the user's emotion, important notifications can be given priority. Emotion estimation is realized using an emotion estimation function, for example, using an emotion engine or a generation AI. The generation AI can be a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. Some or all of the above-mentioned processing in the notification unit can be performed, for example, using AI or without AI. For example, the notification unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and determine notification priorities.

[0147] The notification unit may take into consideration the geographical distribution of calls when making a notification. The notification unit may, for example, include an algorithm for evaluating the geographical distribution of calls. The notification unit may, for example, take into consideration the geographical distribution of calls and make a notification for calls related to a specific region. The notification unit may also, based on the geographical distribution of calls, make a notification for calls in a region where fraud is likely to occur. Furthermore, the notification unit may analyze the geographical distribution of calls, grasp the fraud trends by region, and make a notification. In this way, by taking the geographical distribution of calls into consideration, it becomes easier to grasp the fraud trends by region. Some or all of the above-described processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit may input the geographical distribution data into a generation AI and have the generation AI execute the notification.

[0148] When issuing a notification, the notification unit can improve the accuracy of the notification by referring to related literature. The notification unit, for example, includes a database for referring to related literature. The notification unit issues a notification based on, for example, the latest information on fraudulent methods. The notification unit can also reflect the meaning of specific keywords or phrases in the notification based on the related literature. Furthermore, the notification unit can also refer to related literature and issue a notification for call content that is likely to be fraudulent. In this way, by referring to related literature, the accuracy of the notification is improved. Some or all of the above-mentioned processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit can input literature data into a generation AI and have the generation AI execute the notification.

[0149] When making a notification, the notification unit can make the notification taking into account the market value of the call. The notification unit, for example, includes an algorithm for evaluating the market value of the call. The notification unit, for example, takes into account the market value of the call and makes a notification about important call content. The notification unit can also make a notification about calls that are likely to be fraudulent based on the market value of the call. Furthermore, the notification unit can analyze the market value of the call and make a notification about calls that are at a high risk of fraud. In this way, by taking the market value of the call into consideration, important call content can be given priority in notification. Some or all of the above-mentioned processing in the notification unit may be performed using, for example, AI, or may be performed without using AI. For example, the notification unit can input market value data into a generation AI and have the generation AI execute the notification.

[0150] The learning management unit can estimate a user's emotions and select learning data based on the estimated user emotions. The learning management unit, for example, includes an emotion estimation algorithm for estimating a user's emotions. The learning management unit, for example, analyzes the user's facial expressions and voice tone to estimate emotions. The learning management unit can also select learning data based on the estimated emotions. For example, if the user is nervous, important learning data can be prioritized. Also, if the user is relaxed, detailed learning data can be selected. Furthermore, if the user is in a hurry, learning data that covers the main points can be prioritized. This allows for more appropriate learning by selecting learning data according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, using an emotion engine or a generation AI. The generation AI can be a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to these examples. Some or all of the above-described processing in the learning management unit can be performed using, for example, AI, or without AI. For example, the learning management unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and select learning data.

[0151] During learning, the learning management unit can optimize the learning algorithm by referring to past learning data. The learning management unit, for example, includes a database for referring to past learning data. The learning management unit, for example, optimizes the learning algorithm based on the past learning data. The learning management unit can also improve the accuracy of learning by referring to past learning data. Furthermore, the learning management unit can analyze past learning data and select an optimal learning method. In this way, the accuracy of the learning algorithm is improved by referring to the past learning data. Some or all of the above-mentioned processing in the learning management unit may be performed, for example, using AI, or may be performed without using AI. For example, the learning management unit can input past learning data into the generation AI and cause the generation AI to optimize the learning algorithm.

[0152] The learning management unit can update the learning data by reflecting user feedback during learning. The learning management unit, for example, has a function for collecting user feedback and storing it in a database. The learning management unit, for example, updates the learning data based on user feedback. The learning management unit can also reflect user feedback to improve the accuracy of learning. Furthermore, the learning management unit can analyze user feedback and select the optimal learning method. In this way, the accuracy of the learning data is improved by reflecting user feedback. Some or all of the above-mentioned processing in the learning management unit may be performed, for example, using AI, or may be performed without using AI. For example, the learning management unit can input feedback data into a generation AI and have the generation AI update the learning data.

[0153] The learning management unit can estimate the user's emotions and adjust the frequency of learning based on the estimated user emotions. The learning management unit, for example, includes an emotion estimation algorithm for estimating the user's emotions. The learning management unit, for example, analyzes the user's facial expressions and voice tone to estimate the emotions. The learning management unit can also adjust the frequency of learning based on the estimated emotions. For example, if the user is nervous, the learning frequency can be reduced. Also, if the user is relaxed, the learning frequency can be increased. Furthermore, if the user is in a hurry, the learning frequency can be adjusted. This allows for more appropriate learning by adjusting the learning frequency according to the user's emotions. Emotion estimation is achieved using an emotion estimation function, for example, an emotion engine or a generation AI. The generation AI can be a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. Some or all of the above-mentioned processing in the learning management unit can be performed, for example, using AI, or without AI. For example, the learning management unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and adjust the learning frequency.

[0154] During learning, the learning management unit can weight the learning data based on the time the call was submitted. The learning management unit, for example, includes an algorithm for evaluating the time the call was submitted. The learning management unit, for example, prioritizes learning data with a more recent call submission time. The learning management unit can also learn by referring to data with an older call submission time. Furthermore, the learning management unit can weight the learning data based on the time the call was submitted. In this way, by weighting the learning data based on the time the call was submitted, the most recent information can be prioritized for learning. Some or all of the above-mentioned processing in the learning management unit may be performed using, for example, AI, or may be performed without using AI. For example, the learning management unit can input submission time data into a generation AI and have the generation AI perform weighting.

[0155] During learning, the learning management unit can integrate information from different data sources to enrich the learning data. The learning management unit, for example, includes a data analysis algorithm for integrating information from different data sources. The learning management unit, for example, integrates information from different data sources to enrich the learning data. The learning management unit can also refer to different data sources to improve the accuracy of learning. Furthermore, the learning management unit can analyze different data sources and select an optimal learning method. In this way, the accuracy of the learning data is improved by integrating information from different data sources. Some or all of the above-described processing in the learning management unit may be performed, for example, using AI or without AI. For example, the learning management unit can input different data sources into the generation AI and have the generation AI perform data integration and learning.

[0156] The privacy protection unit can estimate a user's emotion and adjust the privacy protection method based on the estimated user's emotion. The privacy protection unit, for example, includes an emotion estimation algorithm for estimating the user's emotion. The privacy protection unit, for example, analyzes the user's facial expression and tone of voice to estimate the emotion. The privacy protection unit can also adjust the privacy protection method based on the estimated emotion. For example, if the user is nervous, the level of privacy protection can be increased. Also, if the user is relaxed, the level of privacy protection can be adjusted. Furthermore, if the user is in a hurry, the privacy protection method can be simplified. This enables more appropriate privacy protection by adjusting the privacy protection method according to the user's emotion. Emotion estimation is realized using an emotion estimation function, for example, an emotion engine or a generation AI. The generation AI can be, but is not limited to, a text generation AI (e.g., LLM) or a multimodal generation AI. Some or all of the above-mentioned processing in the privacy protection unit can be performed, for example, using AI or without AI. For example, the privacy protection unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and adjust privacy protection methods.

[0157] When performing privacy protection, the privacy protection unit can adjust the level of detail of protection based on the content of the call. The privacy protection unit, for example, includes an algorithm for evaluating the content of the call. For example, the privacy protection unit performs detailed privacy protection when the content of the call includes specific personal information. The privacy protection unit can also perform simplified privacy protection when the content of the call includes general information. Furthermore, the privacy protection unit can also perform immediate privacy protection when the content of the call is urgent. This enables more appropriate privacy protection by adjusting the level of detail of privacy protection based on the content of the call. Some or all of the above-mentioned processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit can input call content data to a generation AI and cause the generation AI to adjust the level of detail of protection.

[0158] The privacy protection unit can perform privacy protection by taking into account attribute information of the other party of the call. The privacy protection unit, for example, includes a database for acquiring attribute information of the other party of the call. For example, if the other party of the call is a family member or friend, the privacy protection unit can perform privacy protection by taking into account the attribute information. Furthermore, if the other party of the call is a person suspected of fraud in the past, the privacy protection unit can also perform privacy protection by taking into account the attribute information of the other party of the call. Furthermore, if the other party of the call is an emergency contact, the privacy protection unit can also perform privacy protection by taking into account the attribute information of the other party of the call. This enables more appropriate privacy protection by taking into account the attribute information of the other party of the call. Some or all of the above-described processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit can input the other party's attribute information into a generation AI and have the generation AI perform protection.

[0159] When performing privacy protection, the privacy protection unit can weight the protection based on the frequency of calls. The privacy protection unit includes, for example, an algorithm for evaluating the frequency of calls. The privacy protection unit, for example, weights the privacy protection for frequently made calls. The privacy protection unit can also weight the privacy protection for calls with frequently called parties. Furthermore, the privacy protection unit can weight the privacy protection for calls with high call frequency based on the frequency of calls. In this way, by weighting the protection based on the frequency of calls, important calls can be protected preferentially. Some or all of the above-mentioned processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit can input call frequency data to a generation AI and have the generation AI perform the weighting.

[0160] The privacy protection unit can estimate a user's emotions and determine the priority of privacy protection based on the estimated user emotions. The privacy protection unit includes, for example, an emotion estimation algorithm for estimating a user's emotions. The privacy protection unit can estimate emotions by, for example, analyzing the user's facial expressions and tone of voice. The privacy protection unit can also determine the priority of privacy protection based on the estimated emotions. For example, if the user is nervous, important privacy protection can be prioritized. Also, if the user is relaxed, detailed privacy protection can be prioritized in an orderly manner. Furthermore, if the user is in a hurry, privacy protection that focuses on the main points can be prioritized. In this way, by determining the priority of privacy protection according to the user's emotions, important privacy protection can be prioritized. Emotion estimation is realized using an emotion estimation function, for example, an emotion engine or a generation AI. The generation AI can be, for example, a text generation AI (e.g., LLM) or a multimodal generation AI, but is not limited to such examples. Some or all of the above-mentioned processing in the privacy protection unit can be performed, for example, using AI or without AI. For example, the privacy protection unit can input the user's facial expression data into the generation AI and have the generation AI estimate emotions and determine privacy protection priorities.

[0161] The privacy protection unit may perform privacy protection by taking into account the geographical distribution of calls. The privacy protection unit may, for example, include an algorithm for evaluating the geographical distribution of calls. For example, the privacy protection unit may consider the geographical distribution of calls and perform privacy protection for calls related to a specific region. The privacy protection unit may also perform privacy protection for calls in regions where fraud is more likely to occur based on the geographical distribution of calls. Furthermore, the privacy protection unit may analyze the geographical distribution of calls, identify fraud trends by region, and perform privacy protection. This makes it easier to identify fraud trends by region by taking the geographical distribution of calls into account. Some or all of the above-described processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit may input geographical distribution data into a generation AI and have the generation AI perform protection.

[0162] When performing privacy protection, the privacy protection unit can improve the accuracy of the protection by referring to related literature. The privacy protection unit, for example, includes a database for referring to related literature. The privacy protection unit performs protection based on the latest information on privacy protection, for example. The privacy protection unit can also reflect the meaning of specific keywords or phrases in the privacy protection based on the related literature. Furthermore, the privacy protection unit can also perform privacy protection for call content that is likely to be fraudulent by referring to related literature. In this way, the accuracy of privacy protection is improved by referring to related literature. Some or all of the above-mentioned processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit can input literature data into a generation AI and have the generation AI perform protection.

[0163] The privacy protection unit can perform privacy protection by taking into account the market value of the call. The privacy protection unit, for example, includes an algorithm for evaluating the market value of the call. The privacy protection unit, for example, takes into account the market value of the call and performs privacy protection for important call content. The privacy protection unit can also perform privacy protection for calls with a high probability of fraud based on the market value of the call. Furthermore, the privacy protection unit can analyze the market value of the call and perform privacy protection for calls with a high risk of fraud. In this way, important call content can be protected preferentially by taking into account the market value of the call. Some or all of the above-mentioned processing in the privacy protection unit may be performed using, for example, AI, or may be performed without using AI. For example, the privacy protection unit can input market value data to a generation AI and have the generation AI perform protection. === Hard Collateral 1-1 === Each of the multiple elements, including the collection unit, analysis unit, determination unit, and warning unit, is realized, for example, by at least one of the smart device 14 and the data processing device 12. For example, the collection unit collects voice data using the camera 42 and microphone 38B of the smart device 14 and records the call content using the control unit 46A. The analysis unit, realized, for example, by the specific processing unit 290 of the data processing device 12, converts the voice data into text and analyzes specific keywords and phrases. The determination unit, realized, for example, by the specific processing unit 290 of the data processing device 12, determines the possibility of fraud based on the analysis results. The warning unit, realized, for example, by the control unit 46A of the smart device 14, sounds an alarm or displays a warning message. The collection unit, for example, analyzes facial expressions using the camera 42 of the smart device 14 to estimate the user's emotions and executes an emotion estimation algorithm. === Hard Collateral 1-2 === Each of the multiple elements, including the collection unit, analysis unit, determination unit, and warning unit, described above, is realized, for example, by at least one of the smart glasses 214 and the data processing device 12. For example, the collection unit collects voice data using the camera 42 and microphone 238 of the smart glasses 214 and records the contents of the call using the control unit 46A. The analysis unit, realized, for example, by the specific processing unit 290 of the data processing device 12, converts the voice data into text and analyzes specific keywords and phrases. The determination unit, realized, for example, by the specific processing unit 290 of the data processing device 12, determines the possibility of fraud based on the analysis results. The warning unit, realized, for example, by the control unit 46A of the smart glasses 214, sounds a warning sound or displays a warning message. The collection unit, for example, analyzes facial expressions using the camera 42 of the smart glasses 214 to estimate the user's emotions and executes an emotion estimation algorithm. === Hard Collateral 1-3 === Each of the multiple elements, including the collection unit, analysis unit, determination unit, and warning unit, is realized, for example, by at least one of the headset-type terminal 314 and the data processing device 12. For example, the collection unit collects voice data using the camera 42 and microphone 238 of the headset-type terminal 314 and records the contents of the call using the control unit 46A. The analysis unit is realized, for example, by the specific processing unit 290 of the data processing device 12, converts the voice data into text, and analyzes specific keywords and phrases. The determination unit is realized, for example, by the specific processing unit 290 of the data processing device 12, and determines the possibility of fraud based on the analysis results. The warning unit is realized, for example, by the control unit 46A of the headset-type terminal 314, and sounds an alarm or displays a warning message. For example, the collection unit analyzes facial expressions using the camera 42 of the headset-type terminal 314 to estimate the user's emotions, and executes an emotion estimation algorithm. === Hard Collateral 1-4 === Each of the multiple elements, including the collection unit, analysis unit, determination unit, and warning unit, is realized, for example, by at least one of the robot 414 and the data processing device 12. For example, the collection unit collects voice data using the camera 42 and microphone 238 of the robot 414 and records the contents of the call using the control unit 46A. The analysis unit is realized, for example, by the specific processing unit 290 of the data processing device 12, converts the voice data into text, and analyzes specific keywords and phrases. The determination unit is realized, for example, by the specific processing unit 290 of the data processing device 12, and determines the possibility of fraud based on the analysis results. The warning unit is realized, for example, by the control unit 46A of the robot 414, and sounds an alarm or displays a warning message. For example, the collection unit analyzes facial expressions using the camera 42 of the robot 414 to estimate the user's emotions and executes an emotion estimation algorithm.

[0164] The system according to the embodiment is not limited to the above-described example, and various modifications are possible, for example, as follows.

[0165] The analysis unit can use natural language processing (NLP) technology to understand the context of the call content. For example, the analysis unit can consider the context before and after the call to more accurately analyze the meaning of specific keywords and phrases. The analysis unit can also grasp the overall flow of the call and identify important parts. Furthermore, the analysis unit can prioritize analysis of parts with a high probability of fraud based on the context of the call. This improves the accuracy of the analysis by taking the context of the call into account.

[0166] The determination unit can refer to the user's past call history to determine the possibility of fraud. For example, the determination unit can detect patterns similar to past calls suspected of being fraudulent. The determination unit can also set an alert level for calls with specific parties based on the past call history. Furthermore, the determination unit can analyze the past call history and prioritize calls that are more likely to be fraudulent. This allows for more accurate determination of the possibility of fraud by taking the past call history into consideration.

[0167] The warning unit can estimate the user's emotions and adjust the warning method based on the estimated user's emotions. For example, if the user is nervous, the warning can be issued in a calm voice. If the user is relaxed, the warning can be issued in a cheerful voice. Furthermore, if the user is in a hurry, the warning can be issued in a quick and concise voice. In this way, by adjusting the warning method according to the user's emotions, more appropriate warnings can be provided.

[0168] The collection unit can collect voice data taking into account the user's geographical location information. For example, if the user is in a specific area, calls related to that area can be collected with priority. Also, if the user is traveling, calls related to the travel destination can be collected with priority. Furthermore, if the user is at home, calls related to information about the area around the user's home can be collected with priority. In this way, by taking into account the user's geographical location information, highly relevant calls can be collected with priority.

[0169] The analysis unit can estimate the user's emotions and adjust the way the analysis is presented based on the estimated user emotions. For example, if the user is nervous, it can provide simple, highly visible analysis results. If the user is relaxed, it can also provide detailed analysis results. Furthermore, if the user is in a hurry, it can also provide analysis results that focus on the main points. In this way, by adjusting the way the analysis is presented according to the user's emotions, it is possible to provide more appropriate analysis results.

[0170] The determination unit can determine the possibility of fraud by taking into account the attribute information of the other party of the call. For example, if the other party of the call is a family member or friend, the determination unit can determine the possibility of fraud by taking into account that attribute information. Also, if the other party of the call is a person suspected of fraud in the past, the determination unit can also determine the possibility of fraud by taking into account that attribute information. Furthermore, if the other party of the call is an emergency contact, the determination unit can also determine the possibility of fraud by taking into account that attribute information. In this way, by taking into account the attribute information of the other party of the call, the possibility of fraud can be determined more accurately.

[0171] The warning unit can adjust the level of detail of the warning based on the content of the call. For example, if the content of the call is related to a specific fraud method, a detailed warning can be issued. Also, if the content of the call indicates the possibility of a general fraud, a brief warning can be issued. Furthermore, if the content of the call is urgent, an immediate warning can be issued. Thus, by adjusting the level of detail of the warning based on the content of the call, more appropriate warnings can be provided.

[0172] The collection unit can analyze the user's social media activity and collect related calls. For example, it can collect calls related to places where the user has checked in on social media. It can also collect related calls by analyzing the content of the user's social media posts. It can also collect related calls by referring to the activities of the user's friends on social media. In this way, it is possible to efficiently collect related calls by analyzing the user's social media activity.

[0173] The analysis unit can perform analysis taking into account the market value of the call. For example, the analysis unit is provided with an algorithm for evaluating the market value of the call. The analysis unit takes into account the market value of the call and prioritizes analysis of important call content. Furthermore, based on the market value of the call, it can also identify calls that are likely to be fraudulent. Furthermore, it can analyze the market value of the call and prioritize analysis of calls that have a high risk of fraud. In this way, by taking into account the market value of the call, it is possible to prioritize analysis of important call content.

[0174] The determination unit can estimate the user's emotions and adjust the criteria for determining the possibility of fraud based on the estimated user emotions. For example, if the user is nervous, the criteria for determining that there is a high possibility of fraud can be tightened. Also, if the user is relaxed, the criteria for determining that there is a low possibility of fraud can be loosened. Furthermore, if the user is in a hurry, the criteria for determining that there is a high possibility of fraud can be tightened. In this way, by adjusting the determination criteria according to the user's emotions, more accurate fraud determination can be achieved.

[0175] The processing flow of the second embodiment will be briefly explained below.

[0176] Step 1: The collection unit collects telephone voice data. For example, the collection unit can record the contents of a call in real time and save it as voice data. It can also start recording automatically when a call is initiated. Furthermore, the collection unit can use noise filtering technology to collect voice data at high quality. Step 2: The analysis unit analyzes the voice data collected by the collection unit. For example, the analysis unit may use voice recognition technology to convert the contents of the call into text and analyze the text data. The analysis unit may also check whether specific keywords or phrases are included. It may also have a sentiment analysis function that analyzes the tone of the voice and the speaker's emotions. Step 3: The decision unit determines the possibility of fraud based on the data analyzed by the analysis unit. For example, the decision unit determines the possibility of fraud based on whether specific keywords or phrases are included. The decision unit can also determine the possibility of fraud by analyzing the tone of voice and the speaker's emotions. Step 4: The warning unit issues a warning if the judgment unit determines that there is a possibility of fraud. For example, the warning unit can sound a warning sound during the call. It can also display a warning message after the call ends. Furthermore, the warning unit can have a function to automatically notify family members or the police.

[0177] 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.

[0178] 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 generative AIs include the data generation model 58, such as a neural network model (e.g., a neural network model), and a neural network model (e.g., a neural network model). The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating speech, text data indicating text, and image data indicating an image is also input to the data generation model 58. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. 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 can perform various processes, but is not limited to these examples. The AI ​​may also be an AI agent. When the processes of each of the above-mentioned parts are performed by AI, the processes may be performed in part or entirely by AI, but are not limited to these examples. The processes performed by AI, including the generative AI, may be replaced with rule-based processes.

[0179] 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.

[0180] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0181] [Second embodiment] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0182] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0183] 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.

[0184] 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.

[0185] 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.

[0186] 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).

[0187] 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.

[0188] 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.

[0189] 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.

[0190] 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.

[0191] 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.

[0192] 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.

[0193] 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.

[0194] 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 containing an instruction, as well as inference data such as voice data representing speech, text data representing text, and image data representing an image. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. 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 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 can perform various processes, but is not limited to these examples. The AI ​​may also be an AI agent. When the processes of each of the above-mentioned parts are performed by AI, the processes may be performed in part or entirely by AI, but are not limited to these examples. The processes performed by AI, including the generative AI, may be replaced with rule-based processes.

[0195] 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.

[0196] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0197] [Third embodiment] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0198] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0199] 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.

[0200] 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.

[0201] 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.

[0202] 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).

[0203] 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.

[0204] 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.

[0205] 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.

[0206] 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.

[0207] 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.

[0208] 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.

[0209] 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.

[0210] 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 containing an instruction, as well as inference data such as voice data representing speech, text data representing text, and image data representing an image. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. 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 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 can perform various processes, but is not limited to these examples. The AI ​​may also be an AI agent. When the processes of each of the above-mentioned parts are performed by AI, the processes may be performed in part or entirely by AI, but are not limited to these examples. The processes performed by AI, including the generative AI, may be replaced with rule-based processes.

[0211] 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.

[0212] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0213] [Fourth embodiment] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0214] 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0215] 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.

[0216] 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.

[0217] 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.

[0218] 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).

[0219] 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.

[0220] 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.

[0221] 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.

[0222] 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.

[0223] 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.

[0224] 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.

[0225] 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.

[0226] 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.

[0227] 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 containing an instruction, as well as inference data such as voice data representing speech, text data representing text, and image data representing an image. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. 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 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 can perform various processes, but is not limited to these examples. The AI ​​may also be an AI agent. When the processes of each of the above-mentioned parts are performed by AI, the processes may be performed in part or entirely by AI, but are not limited to these examples. The processes performed by AI, including the generative AI, may be replaced with rule-based processes.

[0228] 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.

[0229] The correspondence between each part and the device or control part is not limited to the example described above, and various modifications are possible.

[0230] 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.

[0231] 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.

[0232] 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.

[0233] 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).

[0234] 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.

[0235] 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."

[0236] 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.

[0237] 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.

[0238] 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.

[0239] 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.

[0240] 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.

[0241] The hardware resource for executing a specific process can be any of the following types of processors: A processor, for example, is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. A processor also includes a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[0242] 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.

[0243] 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.

[0244] 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.

[0245] 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.

[0246] 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.

[0247] 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.

[0248] [Explanation of symbols]

[0249] 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 collection unit that collects telephone voice data; an analysis unit that analyzes the voice data collected by the collection unit; a determination unit that determines fraud criteria based on the data analyzed by the analysis unit; a warning unit that issues a warning when the fraud criteria are met by the determination unit. A system characterized by:

2. The analysis unit Convert the contents of the call into text using voice recognition technology and analyze the text data.

2. The system of claim 1.

3. The determination unit Check for specific keywords or phrases 2. The system of claim 1.

4. The determination unit Equipped with an emotion analysis unit that analyzes the tone of the voice and the speaker's emotions 2. The system of claim 1.

5. The warning unit Clarify the conditions under which a warning sound is played during a call or a warning message is displayed after a call ends.

2. The system of claim 1.

6. The warning unit Equipped with a notification function that automatically notifies family members and the police 2. The system of claim 1.

7. The analysis unit Equipped with a learning management department that uses past fraud cases as learning data 2. The system of claim 1.

8. The system comprises: Equipped with a privacy protection unit that protects user privacy 2. The system of claim 1.

Images