system

Abstract

システムを提供する。【解決手段】発呼者からの情報を受信する手段と、受信した音響データを文字データに変換する手段と、変換された文字データから発信者情報および依頼内容を抽出する手段と、抽出された情報に基づいて発呼者の信頼性を算出する手段と、算出された信頼性に応じて情報の通信を承認または拒絶する手段と、情報の通信継続中に対話の一貫性を監視し、異常検知の場合に情報の通信を切断する手段と、異常な通信が検出された場合に、利用者に警告を通知する手段と、を含むシステム。

Description

Technical Field

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

Background Art

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

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In recent years, the elderly and people with a tendency to dementia are more likely to suffer from special fraud. Also, many companies are faced with the problem of receiving nuisance business calls on a daily basis. As a result, not only are time and resources wasted by unnecessary calls, but actual financial losses and mental stress are also caused. In the conventional telephone system, means for efficiently and reliably preventing these problems are insufficient.

Means for Solving the Problems

[0005] This invention provides a system that receives communications from callers and converts the voice data into text data. The system extracts sender information and requirements from the text data and calculates the caller's trustworthiness based on the extracted information. By approving or rejecting communications according to the trustworthiness level, it prevents fraudulent and unwanted calls. Furthermore, by monitoring the consistency of the conversation during communication and automatically terminating the call if an anomaly is detected, it provides a safer and more reliable calling environment.

[0006] "Sender" refers to the person or system that initiates communication and transmits information.

[0007] "Communication" refers to the act or system of sending and receiving information, including voice and data.

[0008] "Audio data" refers to data that represents audio in a digital format.

[0009] "Text data" refers to data that represents character information in a digital format.

[0010] "Sender information" refers to identifiable information about the sender of a communication.

[0011] "Requirements" refer to information that indicates the purpose or reason that the sender wants to convey.

[0012] "Trustworthiness" is an indicator that evaluates the legitimacy of the sender and quantifies it.

[0013] "Approval" is the decision or action that permits the continuation of communication.

[0014] "Rejection" refers to a decision or action that denies permission to continue communication.

[0015] "Consistency" refers to a state in conversation where the content before and after a statement does not contradict each other and maintains coherence.

[0016] "Abnormality" refers to an unexpected or inappropriate event in normal communication content or flow.

[0017] "Cut-off" refers to an act or state of terminating communication.

Brief Description of Drawings

[0018] [Figure 1] It is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] It is a conceptual diagram showing an example of the main functions of a data processing device and a smart device according to the first embodiment. [Figure 3] It is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] It is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] It is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] It is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which a plurality of emotions are mapped. [Figure 10] It shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when the emotion engine is combined. [Figure 14] This is a sequence diagram showing the processing flow of the data processing system in Application Example 2, which combines an emotion engine. [Modes for carrying out the invention]

[0019] Hereinafter, an example of an embodiment of the system relating to the technology of this disclosure will be described with reference to the attached drawings.

[0020] First, let's explain the terminology used in the following explanation.

[0021] In the following embodiments, the signed processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Furthermore, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include CPU (Central Processing Unit), GPU (Graphics Processing Unit), GPGPU (General-Purpose computing on Graphics Processing Units), and APU (Accelerated Processing Unit).

[0022] In the following embodiments, signed RAM (Random Access Memory) is a memory that temporarily stores information and is used as work memory by the processor.

[0023] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0024] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0025] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0026] [First Embodiment]

[0027] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0028] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0029] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0030] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0031] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0032] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0033] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0034] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0036] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0037] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0038] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0039] This invention is a system that receives calls from callers via a communication network and detects fraudulent calls early using AI-based speech recognition and text analysis. The system functions as follows:

[0040] First, the server detects an incoming call on the communication network. When a call comes in, the AI ​​module activates and sends an automated response message to the caller. This message asks the caller questions such as, "Who is this?" and "What can I help you with?"

[0041] The caller's response is received as audio data, and real-time speech recognition is performed by an AI module on the server. The audio data is converted into text data, and further text analysis is used to extract information and requirements from the caller. The extracted information is analyzed using historical data registered in the system's database and a reliability evaluation algorithm to calculate the caller's reliability.

[0042] If the server determines that the level of trust is above a certain level, it will send a notification to the user's device and authorize the call. Once authorized, the user's device can start the call. The AI ​​module monitors the conversation throughout the call, checking consistency and coherence of the content in real time. If suspicious behavior is detected, the server will immediately disconnect the call and notify the user.

[0043] As a concrete example, consider a system used at a company's reception desk. When a call comes in to a company, the server automatically answers and states that the caller is a representative of a business partner. At this stage, the system transcribes the name and purpose of the call into text and compares it with past transaction data. If the system confirms that the caller is a legitimate business partner, the call is permitted. If any suspicious requests or inconsistencies are detected during the call, the call is automatically terminated. This allows companies to prevent unnecessary calls and spam attacks.

[0044] In this way, the system of the present invention provides secure and efficient communication through caller reliability evaluation and monitoring of call content. The invention can be used in various scenarios, such as in the homes of the elderly or in corporate reception systems, and reduces the risk of fraud and nuisance calls.

[0045] The following describes the processing flow.

[0046] Step 1:

[0047] The server detects incoming calls on the communication network. When a call is received, the AI ​​module is activated.

[0048] Step 2:

[0049] The server sends an automated response to the caller via an AI module, asking questions such as, "Who is this?" and "What can I help you with?"

[0050] Step 3:

[0051] The server receives the voice response from the caller. The voice data is passed to the AI ​​module, which performs voice recognition in real time.

[0052] Step 4:

[0053] An AI module on the server converts the voice data into text data. This converted data is then analyzed to extract sender information and requirements.

[0054] Step 5:

[0055] The server compares the extracted information with an internal database and uses a reliability evaluation algorithm to calculate the reliability of the sender.

[0056] Step 6:

[0057] Based on the confidence level calculated by the server, if a certain standard is exceeded, the user's device will be notified of permission to make the call. If permission is granted, the user will be able to start the call.

[0058] Step 7:

[0059] During a call, the server uses an AI module to monitor the conversation. It checks for consistency and coherence of the content, enabling it to detect anomalies.

[0060] Step 8:

[0061] If the server detects an anomaly, it will immediately disconnect the call and notify the user of the caller's warning.

[0062] Step 9:

[0063] The server records the call results and analysis information in a database, which can then be used to improve future calls.

[0064] (Example 1)

[0065] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0066] Conventional communication systems have faced challenges in preventing spam and fraudulent activities by untrustworthy callers, as it is difficult to detect calls from fraudulent callers in advance. Furthermore, even after a call has started, there is a lack of reliable methods to monitor whether the content is appropriate, making it impossible to respond quickly to abnormal requests or inconsistent information. Therefore, there is a need for a system that can realize secure and efficient communication.

[0067] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0068] In this invention, the server includes means for detecting communication from a caller, means for sending an automatic response, and means for converting received audio data into text data. This makes it possible to evaluate the reliability of the caller before communication begins and to permit only appropriate calls. Furthermore, it is possible to enhance security by monitoring the content in real time during a call and immediately interrupting the call if an anomaly is detected.

[0069] "Means for detecting communications from callers" refers to devices and technologies for identifying incoming phone calls and voice messages from callers in real time on a communication network.

[0070] "Means of sending automated responses" refers to devices or technologies that automatically play or send pre-set messages in response to a caller's inquiry.

[0071] "Means of converting received audio data into text data" refers to speech recognition technology or software used to convert audio information sent by a caller into text information.

[0072] "Means for extracting sender information and requirements" refers to natural language processing technologies and software used to find and extract specific items or important information from text data.

[0073] "Methods for calculating the trustworthiness of a sender" refers to algorithms or systems that quantify or evaluate the trustworthiness of a sender by comparing extracted information with past data and registered information.

[0074] "Means of permitting or denying communication" refers to mechanical or programmatic devices or technologies for determining and executing whether or not to permit the continuation of communication based on a calculated reliability score.

[0075] "Means of notifying the user" refers to devices or systems that inform the user when communication with the caller is permitted.

[0076] "Means of monitoring conversation content and disconnecting calls if suspicious content is detected" refers to technologies or devices that analyze the exchange during a call in real time and automatically terminate the call if an unusual trend is detected.

[0077] This invention provides a system for managing calls over a communication network more securely and efficiently. The system is constructed from basic components such as a server, terminals, and users.

[0078] The server has the functionality to detect incoming calls via the communication network. Specifically, it monitors network communications using VoIP technology and detects incoming calls in real time. It also has an AI module for sending an automated response after an incoming call, which can send a pre-configured message to the caller. The technologies used at this stage include software such as Google® Cloud Speech-to-Text API and IBM Watson® Speech to Text for speech recognition.

[0079] The received audio data is converted into text data within the server. The converted text is analyzed using natural language processing techniques to extract information and requirements from the caller. Techniques such as entity recognition are used in this process. Based on the extracted information, the server calculates the caller's trustworthiness. A historical database and trustworthiness evaluation algorithm are used to calculate trustworthiness, and the server decides whether to allow or deny the communication based on the resulting score.

[0080] Once communication is permitted, the device notifies the user. This notification is delivered visually and audibly, such as through a pop-up message or audio alert. Once a call begins, the server constantly monitors the consistency and integrity of the conversation, and immediately terminates the call if any anomalies are detected. This allows users to quickly respond to fraudulent callers.

[0081] For example, if this system were implemented in a company's reception system, the server would automatically respond to incoming calls it detects. When the caller claims to be a representative of a client company, their name and purpose would be accurately extracted and quickly compared with past transaction data. If the caller is confirmed to be a trusted caller, the call would be permitted. However, if any fraudulent activity or suspicious requests are detected, the call would be immediately terminated.

[0082] Example prompts for describing this system to a generative AI model:

[0083] Please describe a system that detects incoming calls and identifies fraudulent calls using AI-powered voice recognition and text analysis. This system ensures secure communication through automated responses and reliability assessment.

[0084] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0085] Step 1:

[0086] The server detects an incoming call from the caller on the communication network. Specifically, the VoIP system monitors network traffic and identifies incoming packets in real time. In this process, the input is a signal arriving via the network, and the output is an incoming call event. Based on this event, the server initiates the next processing step.

[0087] Step 2:

[0088] When the server detects an incoming call, it sends a pre-recorded automated response message to the caller. This automated response is played by playing an audio file stored within the system. The input here is the incoming call event, and the output is the audio response to the caller. The server prompts the caller for a response through this interaction.

[0089] Step 3:

[0090] When the caller's response is received as audio data, the server activates an AI module to perform speech recognition. In this process, the audio data is passed to the AI ​​model as input, and the processing outputs text data. The Google Cloud Speech-to-Text API is used for speech recognition. The server then prepares the converted text to proceed.

[0091] Step 4:

[0092] The server analyzes the text data and extracts sender information and requirements. Using natural language processing techniques, it performs entity recognition to identify key items. In this step, text data is used as input, and the output is the extracted sender information and requirements. This allows the server to prepare the material for reliability assessment.

[0093] Step 5:

[0094] The server calculates the caller's trustworthiness based on the information it extracts. The server refers to past call records and a caller database, and applies a trustworthiness evaluation algorithm. The input in this step is the extracted information, and the output is the calculated trustworthiness score. Based on this, the server determines whether or not to allow the communication.

[0095] Step 6:

[0096] Once the caller's trustworthiness is verified, the server notifies the terminal, informing it that communication is permitted. The terminal displays the notification as a pop-up or audio message, conveying it to the user. The input is the result of the trustworthiness score evaluation, and the output is the notification to the user. Through this, the user knows that the call can be initiated.

[0097] Step 7:

[0098] Once communication begins, the server continuously monitors the conversation. An AI module performs text analysis to check the consistency and coherence of the conversation. Here, the audio from the conversation is processed as input, and the output is a determination of whether or not an anomaly is detected. The server constantly performs this analysis, ensuring that any suspicious activity can be detected immediately.

[0099] Step 8:

[0100] If suspicious content is detected, the server immediately disconnects the call and sends a warning notification to the user. The input for this step is the anomaly detection result, and the output is the termination of the call and the sending of a warning. This allows users to be quickly protected from attacks and fraudulent activities.

[0101] (Application Example 1)

[0102] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0103] In recent years, with the development of information and communication technology, the risks of fraudulent calls and cyberattacks have increased. In particular, fraudulent activities targeting individuals and organizations have become a social problem, and conventional telephone answering systems are sometimes insufficient to address this. To address this challenge, there is a need for the development of a new system that can detect and block suspicious calls early while ensuring the reliability of communications.

[0104] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0105] In this invention, the server includes means for receiving information from the caller, means for converting the received acoustic data into text data, means for extracting caller information and request details from the converted text data, and means for notifying the user of a warning when abnormal communication is detected. This improves the security of communications, enables immediate detection of fraudulent calls, and allows users to be warned.

[0106] "Caller" refers to the person or device that initiates communication and transmits information.

[0107] "Audio data" refers to information in signal format that records sound.

[0108] "Text data" refers to a data format in which audio or other information is represented as text.

[0109] "Reliability" refers to an indicator used to measure the legitimacy and security of the content transmitted by the other party in a communication.

[0110] "Information communication" refers to the act or process of sending and receiving audio, data, and other information.

[0111] "Consistency" refers to the logical coherence and regularity of the content of a conversation or communication.

[0112] "Anomaly detection" refers to the process of identifying patterns or behaviors that are different from the norm.

[0113] "User" refers to an individual or organization that uses the system or its communication functions.

[0114] A "warning" means a notice or message provided to indicate a potential danger or problem.

[0115] The system of this invention ensures the reliability of communication by instantly processing acoustic data from the caller and converting it into text data. The server collects acoustic data using a microphone and communication module to receive it. Next, the acoustic data is instantly converted into text data using the Google Speech Recognition API. The converted text data is then analyzed using text analysis software such as SpaCy or NLTK to extract caller information and request details.

[0116] The server calculates the caller's trustworthiness based on the extracted data and approves or rejects the communication accordingly. When abnormal communication is detected, it uses AWS® Lambda, leveraging cloud functionality, to send a warning notification to the user's device. Upon receiving this notification, the user can quickly take action if the call may be fraudulent.

[0117] For example, when a user receives a call on their smartphone, this system detects calls from suspicious callers and warns the user, "This may be a scam. Do you want to disconnect the call?" This allows users to use communication services with peace of mind and protect themselves from potential dangers.

[0118] An example of a prompt message to input into a generative AI model is, "You have an incoming call. Please activate the speech recognition AI and evaluate the caller's trustworthiness." This prompt allows the server to immediately begin analyzing the voice data and take the necessary actions.

[0119] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0120] Step 1:

[0121] The server detects the incoming signal it receives. The server receives incoming calls from smartphones via the communication network and determines that an incoming call has been received. The input here is the incoming signal, and the output is the start of voice data reception.

[0122] Step 2:

[0123] The server receives audio data in real time and converts it into text data using a speech recognition API. Specifically, the server uses the Google Speech Recognition API to analyze the input audio data and convert it into text. The output is the converted text data.

[0124] Step 3:

[0125] The server analyzes the converted text data using a text analysis tool to extract sender information and request details. This analysis uses either SpaCy or NLTK to interpret the meaning from the input text and obtain the information necessary for reliability evaluation. The output is the extracted sender information and request details.

[0126] Step 4:

[0127] The server uses the extracted data and applies a reliability evaluation algorithm to calculate the caller's reliability. Here, the input data is evaluated based on past communication history and pattern matching, and a reliability score is output.

[0128] Step 5:

[0129] The server determines whether to approve or reject communication based on the confidence score and sends a notification to the user. In particular, if an anomaly is detected, a warning notification is sent via AWS Lambda. Here, the input is the confidence score, and the output is the notification message to the user.

[0130] Step 6:

[0131] The system checks the notifications received by the user and, if necessary, chooses to continue or disconnect the call. The action is determined by the warning message displayed on the user's device. The user's choice is the final output.

[0132] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0133] This invention is a system that receives calls from callers via a communication network and uses AI to perform speech recognition, text analysis, and even emotion recognition, thereby achieving safer and more efficient call management.

[0134] First, the server detects the incoming call on the communication network and sends an automated response to the caller via the AI ​​module. This response asks the caller questions such as, "Who is this?" and "What can I help you with?"

[0135] The server receives the voice response from the caller, and the AI ​​module uses speech recognition to convert it into text data in real time. Furthermore, the converted data is analyzed to extract sender information and requirements. This information is then cross-referenced with the system's database to calculate the caller's trustworthiness.

[0136] Furthermore, this invention incorporates an emotion engine, which allows the server to analyze the caller's emotions in real time from their voice. The results of this emotion analysis are used to adjust the confidence level criteria, and only when the confidence level exceeds a certain standard is it used as a basis for deciding whether to approve the call.

[0137] As an example, consider a use case in a corporate support center. When a customer calls, the server automatically responds, and the customer reports a specific problem. At this time, the system calculates a level of trustworthiness from the customer's voice, while an emotion engine determines the degree of stress and anger. This information becomes important reference information for the operator, forming a foundation for more appropriate responses. For example, if the emotion engine detects a high level of stress, the operator can quickly prioritize the response.

[0138] Furthermore, the server monitors the conversation during the call and automatically disconnects the call if inconsistencies or potential risks are detected, further enhancing security. Additionally, the emotion data recognized by the emotion engine is saved as a log and used for future call analysis and improvement.

[0139] In this way, the system of the present invention aims to provide a higher quality communication environment by comprehensively evaluating the reliability and emotions of the caller. This makes it possible to use it in various use cases, such as safe phone calls for the elderly at home and effective customer service in businesses.

[0140] The following describes the processing flow.

[0141] Step 1:

[0142] The server detects incoming calls on the communication network. Upon receiving a call, it immediately activates the AI ​​module.

[0143] Step 2:

[0144] The server automatically responds to the caller via an AI module, asking questions such as, "Who is this?" and "What can I help you with?"

[0145] Step 3:

[0146] The server receives the caller's voice response. It collects the voice data and passes it to the AI ​​module to start the speech recognition process.

[0147] Step 4:

[0148] An AI module on the server converts audio data into text data in real time. The converted text data is then analyzed to extract sender information and requirements.

[0149] Step 5:

[0150] The server compares the analysis results with the information in the system's database and calculates the sender's trustworthiness. A trustworthiness evaluation algorithm is then applied.

[0151] Step 6:

[0152] The server uses an emotion engine to analyze the emotions in the caller's voice. The analysis results, along with a confidence level, are used to decide whether to approve or reject the call.

[0153] Step 7:

[0154] Based on trust level and sentiment evaluation, the server notifies the user's device whether the call is approved or not. If the call is approved, the user can start the call.

[0155] Step 8:

[0156] During a call, the server uses an AI module and emotion engine to monitor the consistency of the conversation and any changes in the caller's emotions. An anomaly detection process continues.

[0157] Step 9:

[0158] If the server detects inconsistencies or suspicious changes in the conversation content or emotions, it will automatically disconnect the call and notify the user of the problem.

[0159] Step 10:

[0160] The server stores call records and analysis data in a database, which will be used for future improvements and reviews.

[0161] (Example 2)

[0162] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0163] Conventional call management systems have difficulty quickly and accurately assessing the reliability and emotional state of callers, hindering secure and efficient communication management. Furthermore, there has been a lack of effective means to monitor emotional changes and conversational consistency in real time during calls to enhance security; therefore, improvements are needed.

[0164] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0165] In this invention, the server includes means for receiving communications from a caller and converting voice information into text information; means for extracting sender identification data and requirements from the converted text information and calculating trustworthiness; and means for analyzing emotions from the caller's voice in real time and adjusting trustworthiness using the results. This enables a comprehensive evaluation of the caller's trustworthiness and emotional state, allowing for safe and efficient communication management.

[0166] "Caller" refers to the person who sends a call or message via a communication network.

[0167] "Contact" refers to a communication signal in voice or data format from the sender.

[0168] A "server" refers to a computer used to receive, process, and analyze communications from senders.

[0169] "Audio information" refers to the audio signal transmitted by the sender.

[0170] "Text information" refers to character data obtained by converting audio information.

[0171] "Sender identification data" refers to information used to identify the sender.

[0172] "Credibility" refers to the result of quantifying or evaluating the trustworthiness of the sender.

[0173] "Emotion" refers to the psychological and emotional state analyzed from the speaker's voice.

[0174] "Real-time" refers to operations or processes that occur immediately or with very little delay.

[0175] "Monitoring" means continuously monitoring the communication status and collecting and analyzing data.

[0176] "Consistency" refers to a state in which conversations or communications proceed without contradiction.

[0177] "Abnormal" refers to an unusual or irregular situation or condition that differs from the norm.

[0178] The invention will now be described in terms of its embodiments. This system receives calls via a communication network, uses AI technology to convert the caller's voice into text, and analyzes the content to achieve secure and efficient communication. This system primarily operates on a server. The server receives voice from the caller and uses a speech recognition engine to convert the voice data into text data in real time. This speech recognition can utilize general speech recognition software or cloud-based speech recognition APIs.

[0179] Next, the server analyzes the generated text data using a natural language processing engine, such as a common natural language processing tool. This analysis extracts information identifying the sender and their requirements. The extracted information is then compared with an internal database to calculate the sender's credibility.

[0180] The server also features an emotion analysis engine that analyzes the caller's emotional state in real time from the voice data. This emotion analysis uses general emotion recognition software. The analysis results are used to adjust the caller's credibility, and the system approves or rejects the call based on that.

[0181] A concrete example of its use is in a company's customer support center. When a user calls the support center, the system analyzes the voice data and automatically classifies the user's problem. At the same time, the emotion engine determines the user's emotional state, and if the user is stressed, it sends a notification to the operator prompting a quick response. This information contributes to improving the quality of customer service.

[0182] An example of a prompt to a generative AI model is: "Please describe the details of a system that analyzes caller voices and performs real-time emotion recognition to optimize calls in a customer support center."

[0183] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0184] Step 1:

[0185] The server detects an incoming call from the caller via the communication network. The server receives the voice transmitted by the caller and takes this voice data as initial input. Here, the voice data is stored on the server as a digital signal.

[0186] Step 2:

[0187] The server passes this audio data to the speech recognition engine, which converts the digital audio signal into text information. The speech recognition engine analyzes the received audio data, deciphers the audio waveform, and generates corresponding text data. This becomes the output text information.

[0188] Step 3:

[0189] The server, upon receiving the generated text data, passes it to a natural language processing engine. The natural language processing engine analyzes the text data to extract sender identification information and requirements. The output of this step is structured data that shows the sender identification information and requirements.

[0190] Step 4:

[0191] The server then compares specific information with its internal database to calculate the sender's credibility. This credibility is calculated based on the sender's past history and registration information. The output is numerical data indicating trustworthiness.

[0192] Step 5:

[0193] The audio data is then passed back to the emotion analysis engine on the server. The emotion analysis engine analyzes the tone and pitch of the voice and extracts the speaker's emotional state. The output of this step is data that quantitatively represents the speaker's emotional state.

[0194] Step 6:

[0195] The server decides whether to approve or reject the caller's call based on the results of trust and sentiment analysis. If the criteria are met, the call is approved; if not, it is rejected. This output indicates the decision to approve or reject the call.

[0196] Step 7:

[0197] During the call, the server monitors the conversation content and detects inconsistencies or potential risks. If an anomaly is detected, the server automatically disconnects the call. The output of this step indicates the status of whether the call continues or ends.

[0198] Step 8:

[0199] The sentiment analysis results and call history are saved as logs in the database. This can be used for future call analysis and service improvements. The output of this step is the accumulated log information.

[0200] (Application Example 2)

[0201] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0202] In communications, especially when involving the elderly, there is a need to instantly identify the caller, assess their reliability, and determine potential risks during a call, thereby ensuring a safe and stress-free communication environment. Existing systems lack sufficient reliability assessment and sentiment analysis, making it difficult to completely protect against suspicious calls and scams targeting the elderly. Therefore, technology is needed that allows the elderly to communicate safely and securely on their own.

[0203] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0204] In this invention, the server includes means for acquiring the caller's communication, means for converting the acquired voice information into text information, and means for extracting sender identification information and the subject from the converted text information. This enhances the security of calls and allows elderly people to answer the phone with peace of mind.

[0205] "Sender" is a term that refers to the individual or device that initiates a communication.

[0206] "Communication" refers to the act of exchanging information between a sender and a receiver in the form of voice or data.

[0207] "Audio information" refers to data in digital or analog format acquired based on audio signals.

[0208] "Textual information" refers to strings of characters or text data obtained by analyzing audio information.

[0209] "Sender identification information" refers to data used to identify the recipient of a communication, and may include, for example, a phone number or name.

[0210] "Subject" refers to the central content or issue that the sender intends to express in their communication.

[0211] "Reliability" is an indicator of how safe or trustworthy the sender and the content of the communication are.

[0212] "Emotional state" refers to the psychological or emotional state of the speaker and is analyzed from the tone and patterns of their voice.

[0213] "Analysis" refers to the process of examining audio and textual information in detail and extracting or evaluating relevant information.

[0214] This invention is a system that highly manages communications from callers, and is designed in particular to allow elderly people to make calls with peace of mind. The system mainly consists of three elements: a server, a terminal, and a user. The server is connected to a communication network, receives communications from callers, and converts the voice information into text information in real time. This mainly uses the Google Speech-to-Text API. The NLTK library is used to extract sender identification information and subject from the converted text information and to calculate reliability. If the reliability is determined to be sufficient, a notification is sent from the terminal to the user.

[0215] Furthermore, the server analyzes the caller's emotional state based on the acquired audio information. A sentiment analysis model trained using TENSORFLOW® performs this role, evaluating the caller's emotional state. If the emotion is determined to be suspicious, the system automatically rejects the call or issues a warning to the user.

[0216] As a concrete example, consider a scenario where a user is using this system on a mobile device. When a call comes in from an unknown number, the server immediately begins an analysis, and if it detects low reliability or suspicious feelings, it notifies the user that they should not answer the call. This process helps elderly people avoid scams and unwanted contact.

[0217] An example of a prompt sentence to input into a generating AI model is: "Develop a solution that allows elderly people to make phone calls with peace of mind. It will utilize AI technology to evaluate the reliability and sentiment of calls and have a function to identify risky communications."

[0218] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0219] Step 1:

[0220] The server receives communications from the caller via the communication network. In this process, it receives audio data as input. The server is configured to receive the data in an initialized state in order to process the audio data appropriately.

[0221] Step 2:

[0222] The server converts the acquired audio data into text in real time using the Google Speech-to-Text API. The input for this step is audio data, and the output is the converted text. The API analyzes the audio and returns the recognized words as strings.

[0223] Step 3:

[0224] The server uses the NLTK library to extract sender identification information and subject from the converted character data. The input for this step is character data, and the output is the extracted sender identification information and subject. The process involves specific actions to extract important individual information through text analysis of the data.

[0225] Step 4:

[0226] The server calculates the sender's trustworthiness based on the extracted information. The inputs to this step are the extracted identification information and subject, and the output is a trustworthiness score. The process involves calculating safety numerically based on historical data and predefined parameters.

[0227] Step 5:

[0228] The server uses a TensorFlow-based sentiment analysis model to evaluate the caller's emotional state from the acquired audio information. The input for this step is audio data, and the output is data indicating the emotional state. The model analyzes the audio patterns and performs a hierarchical classification of emotions.

[0229] Step 6:

[0230] If the reliability score and emotional state meet certain criteria, the device will notify the user. This notification will include a message indicating sufficient reliability. The input for this step is the reliability score and emotional state, and the output is the notification message. The device will then perform the specific tasks of providing information to the user through display and sound.

[0231] Step 7:

[0232] If the server determines that the caller is unreliable or emotionally unstable, it will either reject the call or send a warning message to the terminal. The input for this step is the result of the reliability and emotional assessment, and the output is the call handling policy (rejection or warning). Database updates and user actions are then performed to deter suspicious communications.

[0233] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0234] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0235] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0236] [Second Embodiment]

[0237] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0238] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0239] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0240] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0241] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0242] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0243] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0244] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0245] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0246] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0247] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0248] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0249] This invention is a system that receives calls from callers via a communication network and detects fraudulent calls early using AI-based speech recognition and text analysis. The system functions as follows:

[0250] First, the server detects an incoming call on the communication network. When a call comes in, the AI ​​module activates and sends an automated response message to the caller. This message asks the caller questions such as, "Who is this?" and "What can I do for you?"

[0251] The caller's response is received as audio data, and real-time speech recognition is performed by an AI module on the server. The audio data is converted into text data, and further text analysis is used to extract information and requirements from the caller. The extracted information is analyzed using historical data registered in the system's database and a reliability evaluation algorithm to calculate the caller's reliability.

[0252] If the server determines that the level of trust is above a certain level, it will send a notification to the user's device and authorize the call. Once authorized, the user's device can start the call. The AI ​​module monitors the conversation throughout the call, checking consistency and coherence of the content in real time. If suspicious behavior is detected, the server will immediately disconnect the call and notify the user.

[0253] As a concrete example, consider a system used at a company's reception desk. When a call comes in to a company, the server automatically answers and states that the caller is a representative of a business partner. At this stage, the system transcribes the name and purpose of the call into text and compares it with past transaction data. If the system confirms that the caller is a legitimate business partner, the call is permitted. If any suspicious requests or inconsistencies are detected during the call, the call is automatically terminated. This allows companies to prevent unnecessary calls and spam attacks.

[0254] In this way, the system of the present invention provides secure and efficient communication through caller reliability evaluation and monitoring of call content. The invention can be used in various scenarios, such as in the homes of the elderly or in corporate reception systems, and reduces the risk of fraud and nuisance calls.

[0255] The following describes the processing flow.

[0256] Step 1:

[0257] The server detects incoming calls on the communication network. When a call is received, the AI ​​module is activated.

[0258] Step 2:

[0259] The server sends an automated response to the caller via an AI module, asking questions such as, "Who is this?" and "What can I help you with?"

[0260] Step 3:

[0261] The server receives the voice response from the caller. The voice data is passed to the AI ​​module, which performs voice recognition in real time.

[0262] Step 4:

[0263] An AI module on the server converts the voice data into text data. This converted data is then analyzed to extract sender information and requirements.

[0264] Step 5:

[0265] The server compares the extracted information with an internal database and uses a reliability evaluation algorithm to calculate the reliability of the sender.

[0266] Step 6:

[0267] Based on the confidence level calculated by the server, if a certain standard is exceeded, the user's device will be notified of permission to make the call. If permission is granted, the user will be able to start the call.

[0268] Step 7:

[0269] During a call, the server uses an AI module to monitor the conversation. It checks for consistency and coherence of the content, enabling it to detect anomalies.

[0270] Step 8:

[0271] If the server detects an anomaly, it will immediately disconnect the call and notify the user of the caller's warning.

[0272] Step 9:

[0273] The server records the call results and analysis information in a database, which can then be used to improve future calls.

[0274] (Example 1)

[0275] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0276] Conventional communication systems have faced challenges in preventing spam and fraudulent activities by untrustworthy callers, as it is difficult to detect calls from fraudulent callers in advance. Furthermore, even after a call has started, there is a lack of reliable methods to monitor whether the content is appropriate, making it impossible to respond quickly to abnormal requests or inconsistent information. Therefore, there is a need for a system that can realize secure and efficient communication.

[0277] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0278] In this invention, the server includes means for detecting communication from a caller, means for sending an automatic response, and means for converting received audio data into text data. This makes it possible to evaluate the reliability of the caller before communication begins and to permit only appropriate calls. Furthermore, it is possible to enhance security by monitoring the content in real time during a call and immediately interrupting the call if an anomaly is detected.

[0279] "Means for detecting communications from callers" refers to devices and technologies for identifying incoming phone calls and voice messages from callers in real time on a communication network.

[0280] The "means for transmitting an automatic response" refers to a device or technology for automatically playing or transmitting a pre-set message in response to an inquiry from a caller.

[0281] The "means for converting received voice data into text data" refers to a voice recognition technology or software for converting voice information sent from a caller into character information.

[0282] The "means for extracting caller information and requirements" refers to a natural language processing technology or software for finding and extracting specific items or important information from text data.

[0283] The "means for calculating the reliability of a caller" refers to an algorithm or system for comparing the extracted information with past data or registered information and quantifying or evaluating the credibility of the caller.

[0284] The "means for permitting or rejecting communication" refers to a mechanical or programmatic device or technology for determining whether to allow the continuation of communication based on the calculated reliability score and executing the determination.

[0285] The "means for notifying a user" refers to a device or system for informing the user of information when communication with the caller is permitted.

[0286] The "means for monitoring conversation content and disconnecting a call if suspicious content is detected" refers to a technology or device for analyzing the exchanges during a call in real time and automatically terminating the call if different tendencies are recognized.

[0287] This invention provides a system for more securely and efficiently managing calls on a communication network. The system is constructed from basic components such as a server, a terminal, and a user.

[0288] The server has the ability to detect incoming calls via the communication network. Specifically, it monitors network communications using VoIP technology and detects incoming calls in real time. It also has an AI module for sending an automated response after an incoming call, which can send a pre-configured message to the caller. The technologies used at this stage include software such as Google Cloud Speech-to-Text API and IBM Watson Speech to Text for speech recognition.

[0289] The received audio data is converted into text data within the server. The converted text is analyzed using natural language processing techniques to extract information and requirements from the caller. Techniques such as entity recognition are used in this process. Based on the extracted information, the server calculates the caller's trustworthiness. A historical database and trustworthiness evaluation algorithm are used to calculate trustworthiness, and the server decides whether to allow or deny the communication based on the resulting score.

[0290] Once communication is permitted, the device notifies the user. This notification is delivered visually and audibly, such as through a pop-up message or audio alert. Once a call begins, the server constantly monitors the consistency and integrity of the conversation, and immediately terminates the call if any anomalies are detected. This allows users to quickly respond to fraudulent callers.

[0291] For example, if this system were implemented in a company's reception system, the server would automatically respond to incoming calls it detects. When the caller claims to be a representative of a client company, their name and purpose would be accurately extracted and quickly compared with past transaction data. If the caller is confirmed to be a trusted caller, the call would be permitted. However, if any fraudulent activity or suspicious requests are detected, the call would be immediately terminated.

[0292] Example prompts for explaining this system to a generative AI model:

[0293] Please describe a system that detects incoming calls and identifies fraudulent calls using AI-powered voice recognition and text analysis. This system ensures secure communication through automated responses and reliability assessment.

[0294] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0295] Step 1:

[0296] The server detects an incoming call from the caller on the communication network. Specifically, the VoIP system monitors network traffic and identifies incoming packets in real time. In this process, the input is a signal arriving via the network, and the output is an incoming call event. Based on this event, the server initiates the next processing step.

[0297] Step 2:

[0298] When the server detects an incoming call, it sends a pre-recorded automated response message to the caller. This automated response is played by playing an audio file stored within the system. The input here is the incoming call event, and the output is the audio response to the caller. The server prompts the caller for a response through this interaction.

[0299] Step 3:

[0300] When the caller's response is received as audio data, the server activates an AI module to perform speech recognition. In this process, the audio data is passed to the AI ​​model as input, and the processing outputs text data. The Google Cloud Speech-to-Text API is used for speech recognition. The server then prepares the converted text to proceed.

[0301] Step 4:

[0302] The server analyzes the text data and extracts the information and requirements of the caller. Using natural language processing technology, entity recognition is performed to identify important items. In this step, text data is used as the input, and the output is the extracted caller information and requirements. This enables the server to prepare materials for reliability evaluation.

[0303] Step 5:

[0304] Based on the information extracted by the server, the reliability of the caller is calculated. The server refers to past call records and the caller database and applies a reliability evaluation algorithm. The input in this step is the extracted information, and the output is the calculated reliability score. Based on this, it is determined whether communication is permitted.

[0305] Step 6:

[0306] When the reliability of the caller is confirmed, the server notifies the terminal and informs it of the permission to communicate. On the terminal, the notification is displayed as a pop-up or voice and conveyed to the user. The input is the evaluation result of the reliability score, and the output is the notification to the user. Through this, the user knows that the call can be started.

[0307] Step 7:

[0308] When communication starts, the server continuously monitors the conversation. The AI module performs text analysis to check the consistency and content integrity of the conversation. Here, the voice during the conversation is processed as the input, and it is determined as the output whether an abnormality is detected. The server always executes this analysis to be in a state where suspicious actions can be detected immediately.

[0309] Step 8:

[0310] If suspicious content is detected, the server immediately disconnects the call and sends a warning notification to the user. The input in this step is the abnormality detection result, and the output is the end of the call and the sending of a warning. This enables the user to be quickly protected from attacks and illegal acts.

[0311] (Application Example 1)

[0312] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0313] In recent years, with the development of information and communication technology, the risks of fraudulent calls and cyberattacks have increased. In particular, fraudulent activities targeting individuals and organizations have become a social problem, and conventional telephone answering systems are sometimes insufficient to address this. To address this challenge, there is a need for the development of a new system that can detect and block suspicious calls early while ensuring the reliability of communications.

[0314] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0315] In this invention, the server includes means for receiving information from the caller, means for converting the received acoustic data into text data, means for extracting caller information and request details from the converted text data, and means for notifying the user of a warning when abnormal communication is detected. This improves the security of communications, enables immediate detection of fraudulent calls, and allows users to be warned.

[0316] "Caller" refers to the person or device that initiates communication and transmits information.

[0317] "Audio data" refers to information in signal format that records sound.

[0318] "Text data" refers to a data format in which audio or other information is represented as text.

[0319] "Reliability" refers to an indicator used to measure the legitimacy and security of the content transmitted by the other party in a communication.

[0320] "Information communication" refers to the act or process of sending and receiving audio, data, and other information.

[0321] "Consistency" refers to the logical coherence and regularity of the content of a conversation or communication.

[0322] "Anomaly detection" refers to the process of identifying patterns or behaviors that are different from the norm.

[0323] "User" refers to an individual or organization that uses the system or its communication functions.

[0324] A "warning" means a notice or message provided to indicate a potential danger or problem.

[0325] The system of this invention ensures the reliability of communication by instantly processing acoustic data from the caller and converting it into text data. The server collects acoustic data using a microphone and communication module to receive it. Next, the acoustic data is instantly converted into text data using the Google Speech Recognition API. The converted text data is then analyzed using text analysis software such as SpaCy or NLTK to extract caller information and request details.

[0326] The server calculates the caller's trustworthiness based on the extracted data and approves or rejects the communication accordingly. When abnormal communication is detected, it uses AWS Lambda, leveraging cloud functionality, to send a warning notification to the user's device. Upon receiving this notification, the user can quickly take action if the call may be fraudulent.

[0327] For example, when a user receives a call on their smartphone, this system detects calls from suspicious callers and warns the user, "This may be a scam. Do you want to disconnect the call?" This allows users to use communication services with peace of mind and protect themselves from potential dangers.

[0328] An example of a prompt message to input into a generative AI model is, "You have an incoming call. Please activate the speech recognition AI and evaluate the caller's trustworthiness." This prompt allows the server to immediately begin analyzing the voice data and take the necessary actions.

[0329] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0330] Step 1:

[0331] The server detects the incoming signal it receives. The server receives incoming calls from smartphones via the communication network and determines that an incoming call has been received. The input here is the incoming signal, and the output is the start of voice data reception.

[0332] Step 2:

[0333] The server receives audio data in real time and converts it into text data using a speech recognition API. Specifically, the server uses the Google Speech Recognition API to analyze the input audio data and convert it into text. The output is the converted text data.

[0334] Step 3:

[0335] The server analyzes the converted text data using a text analysis tool to extract sender information and request details. This analysis uses either SpaCy or NLTK to interpret the meaning from the input text and obtain the information necessary for reliability evaluation. The output is the extracted sender information and request details.

[0336] Step 4:

[0337] The server uses the extracted data and applies a reliability evaluation algorithm to calculate the caller's reliability. Here, the input data is evaluated based on past communication history and pattern matching, and a reliability score is output.

[0338] Step 5:

[0339] The server determines whether to approve or reject communication based on the confidence score and sends a notification to the user. In particular, if an anomaly is detected, a warning notification is sent via AWS Lambda. Here, the input is the confidence score, and the output is the notification message to the user.

[0340] Step 6:

[0341] The system checks the notifications received by the user and, if necessary, chooses to continue or disconnect the call. The action is determined by the warning message displayed on the user's device. The user's choice is the final output.

[0342] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0343] This invention is a system that receives calls from callers via a communication network and uses AI to perform speech recognition, text analysis, and even emotion recognition, thereby achieving safer and more efficient call management.

[0344] First, the server detects the incoming call on the communication network and sends an automated response to the caller via the AI ​​module. This response asks the caller questions such as, "Who is this?" and "What can I help you with?"

[0345] The server receives the voice response from the caller, and the AI ​​module uses speech recognition to convert it into text data in real time. Furthermore, the converted data is analyzed to extract sender information and requirements. This information is then cross-referenced with the system's database to calculate the caller's trustworthiness.

[0346] Furthermore, this invention incorporates an emotion engine, which allows the server to analyze the caller's emotions in real time from their voice. The results of this emotion analysis are used to adjust the confidence level criteria, and only when the confidence level exceeds a certain standard is it used as a basis for deciding whether to approve the call.

[0347] As an example, consider a use case in a corporate support center. When a customer calls, the server automatically responds, and the customer reports a specific problem. At this time, the system calculates a level of trustworthiness from the customer's voice, while an emotion engine determines the degree of stress and anger. This information becomes important reference information for the operator, forming a foundation for more appropriate responses. For example, if the emotion engine detects a high level of stress, the operator can quickly prioritize the response.

[0348] Furthermore, the server monitors the conversation during the call and automatically disconnects the call if inconsistencies or potential risks are detected, further enhancing security. Additionally, the emotion data recognized by the emotion engine is saved as a log and used for future call analysis and improvement.

[0349] In this way, the system of the present invention aims to provide a higher quality communication environment by comprehensively evaluating the reliability and emotions of the caller. This makes it possible to use it in various use cases, such as safe phone calls for the elderly at home and effective customer service in businesses.

[0350] The following describes the processing flow.

[0351] Step 1:

[0352] The server detects incoming calls on the communication network. Upon receiving a call, it immediately activates the AI ​​module.

[0353] Step 2:

[0354] The server automatically responds to the caller via an AI module, asking questions such as, "Who is this?" and "What can I help you with?"

[0355] Step 3:

[0356] The server receives the caller's voice response. It collects the voice data and passes it to the AI ​​module to start the speech recognition process.

[0357] Step 4:

[0358] An AI module on the server converts audio data into text data in real time. The converted text data is then analyzed to extract sender information and requirements.

[0359] Step 5:

[0360] The server compares the analysis results with the information in the system's database and calculates the sender's trustworthiness. A trustworthiness evaluation algorithm is then applied.

[0361] Step 6:

[0362] The server uses an emotion engine to analyze the emotions in the caller's voice. The analysis results, along with a confidence level, are used to decide whether to approve or reject the call.

[0363] Step 7:

[0364] Based on trust level and sentiment evaluation, the server notifies the user's device whether the call is approved or not. If the call is approved, the user can start the call.

[0365] Step 8:

[0366] During a call, the server uses an AI module and emotion engine to monitor the consistency of the conversation and any changes in the caller's emotions. An anomaly detection process continues.

[0367] Step 9:

[0368] If the server detects inconsistencies or suspicious changes in the conversation content or emotions, it will automatically disconnect the call and notify the user of the problem.

[0369] Step 10:

[0370] The server stores call records and analysis data in a database, which will be used for future improvements and reviews.

[0371] (Example 2)

[0372] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0373] Conventional call management systems have difficulty quickly and accurately assessing the reliability and emotional state of callers, hindering secure and efficient communication management. Furthermore, there has been a lack of effective means to monitor emotional changes and conversational consistency in real time during calls to enhance security; therefore, improvements are needed.

[0374] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0375] In this invention, the server includes means for receiving communications from a caller and converting voice information into text information; means for extracting sender identification data and requirements from the converted text information and calculating trustworthiness; and means for analyzing emotions from the caller's voice in real time and adjusting trustworthiness using the results. This enables a comprehensive evaluation of the caller's trustworthiness and emotional state, allowing for safe and efficient communication management.

[0376] "Caller" refers to the person who sends a call or message via a communication network.

[0377] "Contact" refers to a communication signal in voice or data format from the sender.

[0378] A "server" refers to a computer used to receive, process, and analyze communications from senders.

[0379] "Audio information" refers to the audio signal transmitted by the sender.

[0380] "Text information" refers to character data obtained by converting audio information.

[0381] "Sender identification data" refers to information used to identify the sender.

[0382] "Credibility" refers to the result of quantifying or evaluating the trustworthiness of the sender.

[0383] "Emotion" refers to the psychological and emotional state analyzed from the speaker's voice.

[0384] "Real-time" refers to operations or processes that occur immediately or with very little delay.

[0385] "Monitoring" means continuously monitoring the communication status and collecting and analyzing data.

[0386] "Consistency" refers to a state in which conversations or communications proceed without contradiction.

[0387] "Abnormal" refers to an unusual or irregular situation or condition that differs from the norm.

[0388] The invention will now be described in terms of its embodiments. This system receives calls via a communication network, uses AI technology to convert the caller's voice into text, and analyzes the content to achieve secure and efficient communication. This system primarily operates on a server. The server receives voice from the caller and uses a speech recognition engine to convert the voice data into text data in real time. This speech recognition can utilize general speech recognition software or cloud-based speech recognition APIs.

[0389] Next, the server analyzes the generated text data using a natural language processing engine, such as a common natural language processing tool. This analysis extracts information identifying the sender and their requirements. The extracted information is then compared with an internal database to calculate the sender's credibility.

[0390] The server also features an emotion analysis engine that analyzes the caller's emotional state in real time from the voice data. This emotion analysis uses general emotion recognition software. The analysis results are used to adjust the caller's credibility, and the system approves or rejects the call based on that.

[0391] A concrete example of its use is in a company's customer support center. When a user calls the support center, the system analyzes the voice data and automatically classifies the user's problem. At the same time, the emotion engine determines the user's emotional state, and if the user is stressed, it sends a notification to the operator prompting a quick response. This information contributes to improving the quality of customer service.

[0392] An example of a prompt to a generative AI model is: "Please describe the details of a system that analyzes caller voices and performs real-time emotion recognition to optimize calls in a customer support center."

[0393] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0394] Step 1:

[0395] The server detects an incoming call from the caller via the communication network. The server receives the voice transmitted by the caller and takes this voice data as initial input. Here, the voice data is stored on the server as a digital signal.

[0396] Step 2:

[0397] The server passes this audio data to the speech recognition engine, which converts the digital audio signal into text information. The speech recognition engine analyzes the received audio data, deciphers the audio waveform, and generates corresponding text data. This becomes the output text information.

[0398] Step 3:

[0399] The server, upon receiving the generated text data, passes it to a natural language processing engine. The natural language processing engine analyzes the text data to extract sender identification information and requirements. The output of this step is structured data that shows the sender identification information and requirements.

[0400] Step 4:

[0401] The server then compares specific information with its internal database to calculate the sender's credibility. This credibility is calculated based on the sender's past history and registration information. The output is numerical data indicating trustworthiness.

[0402] Step 5:

[0403] The audio data is then passed back to the emotion analysis engine on the server. The emotion analysis engine analyzes the tone and pitch of the voice and extracts the speaker's emotional state. The output of this step is data that quantitatively represents the speaker's emotional state.

[0404] Step 6:

[0405] The server decides whether to approve or reject the caller's call based on the results of trust and sentiment analysis. If the criteria are met, the call is approved; if not, it is rejected. This output indicates the decision to approve or reject the call.

[0406] Step 7:

[0407] During the call, the server monitors the conversation content and detects inconsistencies or potential risks. If an anomaly is detected, the server automatically disconnects the call. The output of this step indicates the status of whether the call continues or ends.

[0408] Step 8:

[0409] The sentiment analysis results and call history are saved as logs in the database. This can be used for future call analysis and service improvements. The output of this step is the accumulated log information.

[0410] (Application Example 2)

[0411] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0412] In communications, especially when involving the elderly, there is a need to instantly identify the caller, assess their reliability, and determine potential risks during a call, thereby ensuring a safe and stress-free communication environment. Existing systems lack sufficient reliability assessment and sentiment analysis, making it difficult to completely protect against suspicious calls and scams targeting the elderly. Therefore, technology is needed that allows the elderly to communicate safely and securely on their own.

[0413] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0414] In this invention, the server includes means for acquiring the caller's communication, means for converting the acquired voice information into text information, and means for extracting sender identification information and the subject from the converted text information. This enhances the security of calls and allows elderly people to answer the phone with peace of mind.

[0415] "Sender" is a term that refers to the individual or device that initiates a communication.

[0416] "Communication" refers to the act of exchanging information between a sender and a receiver in the form of voice or data.

[0417] "Audio information" refers to data in digital or analog format acquired based on audio signals.

[0418] "Textual information" refers to strings of characters or text data obtained by analyzing audio information.

[0419] "Sender identification information" refers to data used to identify the recipient of a communication, and may include, for example, a phone number or name.

[0420] "Subject" refers to the central content or issue that the sender intends to express in their communication.

[0421] "Reliability" is an indicator of how safe or trustworthy the sender and the content of the communication are.

[0422] "Emotional state" refers to the psychological or emotional state of the speaker and is analyzed from the tone and patterns of their voice.

[0423] "Analysis" refers to the process of examining audio and textual information in detail and extracting or evaluating relevant information.

[0424] This invention is a system that highly manages communications from callers, and is designed in particular to allow elderly people to make calls with peace of mind. The system mainly consists of three elements: a server, a terminal, and a user. The server is connected to a communication network, receives communications from callers, and converts the voice information into text information in real time. This mainly uses the Google Speech-to-Text API. The NLTK library is used to extract sender identification information and subject from the converted text information and to calculate reliability. If the reliability is determined to be sufficient, a notification is sent from the terminal to the user.

[0425] Furthermore, the server analyzes the caller's emotional state based on the acquired audio information. A sentiment analysis model trained using TensorFlow performs this role, evaluating the caller's emotional state. If the emotion is determined to be suspicious, the system automatically rejects the call or issues a warning to the user.

[0426] As a concrete example, consider a scenario where a user is using this system on a mobile device. When a call comes in from an unknown number, the server immediately begins an analysis, and if it detects low reliability or suspicious feelings, it notifies the user that they should not answer the call. This process helps elderly people avoid scams and unwanted contact.

[0427] An example of a prompt sentence to input into a generative AI model is: "Develop a solution that allows elderly people to make phone calls with peace of mind. It will utilize AI technology to evaluate the reliability and sentiment of calls and have a function to identify risky communications."

[0428] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0429] Step 1:

[0430] The server receives communications from the caller via the communication network. In this process, it receives audio data as input. The server is configured to receive the data in an initialized state in order to process the audio data appropriately.

[0431] Step 2:

[0432] The server converts the acquired audio data into text in real time using the Google Speech-to-Text API. The input for this step is audio data, and the output is the converted text. The API analyzes the audio and returns the recognized words as strings.

[0433] Step 3:

[0434] The server uses the NLTK library to extract sender identification information and subject from the converted character data. The input for this step is character data, and the output is the extracted sender identification information and subject. The process involves specific actions to extract important individual information through text analysis of the data.

[0435] Step 4:

[0436] The server calculates the sender's trustworthiness based on the extracted information. The inputs to this step are the extracted identification information and subject, and the output is a trustworthiness score. The process involves calculating safety numerically based on historical data and predefined parameters.

[0437] Step 5:

[0438] The server uses a TensorFlow-based sentiment analysis model to evaluate the caller's emotional state from the acquired audio information. The input for this step is audio data, and the output is data indicating the emotional state. The model analyzes the audio patterns and performs a hierarchical classification of emotions.

[0439] Step 6:

[0440] If the reliability score and emotional state meet certain criteria, the device will notify the user. This notification will include a message indicating sufficient reliability. The input for this step is the reliability score and emotional state, and the output is the notification message. The device will then perform the specific tasks of providing information to the user through display and sound.

[0441] Step 7:

[0442] If the server determines that the caller is unreliable or emotionally unstable, it will either reject the call or send a warning message to the terminal. The input for this step is the result of the reliability and emotional assessment, and the output is the call handling policy (rejection or warning). Database updates and user actions are then performed to deter suspicious communications.

[0443] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0444] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0445] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0446] [Third Embodiment]

[0447] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0448] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0449] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0450] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0451] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0452] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0453] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0454] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0455] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0456] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0457] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0458] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0459] This invention is a system that receives calls from callers via a communication network and detects fraudulent calls early using AI-based speech recognition and text analysis. The system functions as follows:

[0460] First, the server detects an incoming call on the communication network. When a call comes in, the AI ​​module activates and sends an automated response message to the caller. This message asks the caller questions such as, "Who is this?" and "What can I do for you?"

[0461] The caller's response is received as audio data, and real-time speech recognition is performed by an AI module on the server. The audio data is converted into text data, and further text analysis is used to extract information and requirements from the caller. The extracted information is analyzed using historical data registered in the system's database and a reliability evaluation algorithm to calculate the caller's reliability.

[0462] If the server determines that the level of trust is above a certain level, it will send a notification to the user's device and authorize the call. Once authorized, the user's device can start the call. The AI ​​module monitors the conversation throughout the call, checking consistency and coherence of the content in real time. If suspicious behavior is detected, the server will immediately disconnect the call and notify the user.

[0463] As a concrete example, consider a system used at a company's reception desk. When a call comes in to a company, the server automatically answers and states that the caller is a representative of a business partner. At this stage, the system transcribes the name and purpose of the call into text and compares it with past transaction data. If the system confirms that the caller is a legitimate business partner, the call is permitted. If any suspicious requests or inconsistencies are detected during the call, the call is automatically terminated. This allows companies to prevent unnecessary calls and spam attacks.

[0464] In this way, the system of the present invention provides secure and efficient communication through caller reliability evaluation and monitoring of call content. The invention can be used in various scenarios, such as in the homes of the elderly or in corporate reception systems, and reduces the risk of fraud and nuisance calls.

[0465] The following describes the processing flow.

[0466] Step 1:

[0467] The server detects incoming calls on the communication network. When a call is received, the AI ​​module is activated.

[0468] Step 2:

[0469] The server sends an automated response to the caller via an AI module, asking questions such as, "Who is this?" and "What can I help you with?"

[0470] Step 3:

[0471] The server receives the voice response from the caller. The voice data is passed to the AI ​​module, which performs voice recognition in real time.

[0472] Step 4:

[0473] An AI module on the server converts the voice data into text data. This converted data is then analyzed to extract sender information and requirements.

[0474] Step 5:

[0475] The server compares the extracted information with an internal database and uses a reliability evaluation algorithm to calculate the reliability of the sender.

[0476] Step 6:

[0477] Based on the confidence level calculated by the server, if a certain standard is exceeded, the user's device will be notified of permission to make the call. If permission is granted, the user will be able to start the call.

[0478] Step 7:

[0479] During a call, the server uses an AI module to monitor the conversation. It checks for consistency and coherence of the content, enabling it to detect anomalies.

[0480] Step 8:

[0481] If the server detects an anomaly, it will immediately disconnect the call and notify the user of the caller's warning.

[0482] Step 9:

[0483] The server records the call results and analysis information in a database, which can then be used to improve future calls.

[0484] (Example 1)

[0485] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0486] Conventional communication systems have faced challenges in preventing spam and fraudulent activities by untrustworthy callers, as it is difficult to detect calls from fraudulent callers in advance. Furthermore, even after a call has started, there is a lack of reliable methods to monitor whether the content is appropriate, making it impossible to respond quickly to abnormal requests or inconsistent information. Therefore, there is a need for a system that can realize secure and efficient communication.

[0487] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0488] In this invention, the server includes means for detecting communication from a caller, means for sending an automatic response, and means for converting received audio data into text data. This makes it possible to evaluate the reliability of the caller before communication begins and to permit only appropriate calls. Furthermore, it is possible to enhance security by monitoring the content in real time during a call and immediately interrupting the call if an anomaly is detected.

[0489] "Means for detecting communications from callers" refers to devices and technologies for identifying incoming phone calls and voice messages from callers in real time on a communication network.

[0490] "Means of sending automated responses" refers to devices or technologies that automatically play or send pre-set messages in response to a caller's inquiry.

[0491] "Means of converting received audio data into text data" refers to speech recognition technology or software used to convert audio information sent by a caller into text information.

[0492] "Means for extracting sender information and requirements" refers to natural language processing technologies and software used to find and extract specific items or important information from text data.

[0493] "Methods for calculating the trustworthiness of a sender" refers to algorithms or systems that quantify or evaluate the trustworthiness of a sender by comparing extracted information with past data and registered information.

[0494] "Means of permitting or denying communication" refers to mechanical or programmatic devices or technologies for determining and executing whether or not to permit the continuation of communication based on a calculated reliability score.

[0495] "Means of notifying the user" refers to devices or systems that inform the user when communication with the caller is permitted.

[0496] "Means of monitoring conversation content and disconnecting calls if suspicious content is detected" refers to technologies or devices that analyze the exchange during a call in real time and automatically terminate the call if an unusual trend is detected.

[0497] This invention provides a system for managing calls over a communication network more securely and efficiently. The system is constructed from basic components such as a server, terminals, and users.

[0498] The server has the ability to detect incoming calls via the communication network. Specifically, it monitors network communications using VoIP technology and detects incoming calls in real time. It also has an AI module for sending an automated response after an incoming call, which can send a pre-configured message to the caller. The technologies used at this stage include software such as Google Cloud Speech-to-Text API and IBM Watson Speech to Text for speech recognition.

[0499] The received audio data is converted into text data within the server. The converted text is analyzed using natural language processing techniques to extract information and requirements from the caller. Techniques such as entity recognition are used in this process. Based on the extracted information, the server calculates the caller's trustworthiness. A historical database and trustworthiness evaluation algorithm are used to calculate trustworthiness, and the server decides whether to allow or deny the communication based on the resulting score.

[0500] Once communication is permitted, the device notifies the user. This notification is delivered visually and audibly, such as through a pop-up message or audio alert. Once a call begins, the server constantly monitors the consistency and integrity of the conversation, and immediately terminates the call if any anomalies are detected. This allows users to quickly respond to fraudulent callers.

[0501] For example, if this system were implemented in a company's reception system, the server would automatically respond to incoming calls it detects. When the caller claims to be a representative of a client company, their name and purpose would be accurately extracted and quickly compared with past transaction data. If the caller is confirmed to be a trusted caller, the call would be permitted. However, if any fraudulent activity or suspicious requests are detected, the call would be immediately terminated.

[0502] Example prompts for explaining this system to a generative AI model:

[0503] Please describe a system that detects incoming calls and identifies fraudulent calls using AI-powered voice recognition and text analysis. This system ensures secure communication through automated responses and reliability assessment.

[0504] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0505] Step 1:

[0506] The server detects an incoming call from the caller on the communication network. Specifically, the VoIP system monitors network traffic and identifies incoming packets in real time. In this process, the input is a signal arriving via the network, and the output is an incoming call event. Based on this event, the server initiates the next processing step.

[0507] Step 2:

[0508] When the server detects an incoming call, it sends a pre-recorded automated response message to the caller. This automated response is played by playing an audio file stored within the system. The input here is the incoming call event, and the output is the audio response to the caller. The server prompts the caller for a response through this interaction.

[0509] Step 3:

[0510] When the caller's response is received as audio data, the server activates an AI module to perform speech recognition. In this process, the audio data is passed to the AI ​​model as input, and the processing outputs text data. The Google Cloud Speech-to-Text API is used for speech recognition. The server then prepares the converted text to proceed.

[0511] Step 4:

[0512] The server analyzes the text data and extracts sender information and requirements. Using natural language processing techniques, it performs entity recognition to identify key items. In this step, text data is used as input, and the output is the extracted sender information and requirements. This allows the server to prepare the material for reliability assessment.

[0513] Step 5:

[0514] The server calculates the caller's trustworthiness based on the information it extracts. The server refers to past call records and a caller database, and applies a trustworthiness evaluation algorithm. The input in this step is the extracted information, and the output is the calculated trustworthiness score. Based on this, the server determines whether or not to allow the communication.

[0515] Step 6:

[0516] Once the caller's trustworthiness is verified, the server notifies the terminal, informing it that communication is permitted. The terminal displays the notification as a pop-up or audio message, conveying it to the user. The input is the result of the trustworthiness score evaluation, and the output is the notification to the user. Through this, the user knows that the call can be initiated.

[0517] Step 7:

[0518] Once communication begins, the server continuously monitors the conversation. An AI module performs text analysis to check the consistency and coherence of the conversation. Here, the audio from the conversation is processed as input, and the output is a determination of whether or not an anomaly is detected. The server constantly performs this analysis, ensuring that any suspicious activity can be detected immediately.

[0519] Step 8:

[0520] If suspicious content is detected, the server immediately disconnects the call and sends a warning notification to the user. The input for this step is the anomaly detection result, and the output is the termination of the call and the sending of a warning. This allows users to be quickly protected from attacks and fraudulent activities.

[0521] (Application Example 1)

[0522] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0523] In recent years, with the development of information and communication technology, the risks of fraudulent calls and cyberattacks have increased. In particular, fraudulent activities targeting individuals and organizations have become a social problem, and conventional telephone answering systems are sometimes insufficient to address this. To address this challenge, there is a need for the development of a new system that can detect and block suspicious calls early while ensuring the reliability of communications.

[0524] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0525] In this invention, the server includes means for receiving information from the caller, means for converting the received acoustic data into text data, means for extracting caller information and request details from the converted text data, and means for notifying the user of a warning when abnormal communication is detected. This improves the security of communications, enables immediate detection of fraudulent calls, and allows users to be warned.

[0526] "Caller" refers to the person or device that initiates communication and transmits information.

[0527] "Audio data" refers to information in signal format that records sound.

[0528] "Text data" refers to a data format in which audio or other information is represented as text.

[0529] "Reliability" refers to an indicator used to measure the legitimacy and security of the content transmitted by the other party in a communication.

[0530] "Information communication" refers to the act or process of sending and receiving audio, data, and other information.

[0531] "Consistency" refers to the logical coherence and regularity of the content of a conversation or communication.

[0532] "Anomaly detection" refers to the process of identifying patterns or behaviors that are different from the norm.

[0533] "User" refers to an individual or organization that uses the system or its communication functions.

[0534] A "warning" means a notice or message provided to indicate a potential danger or problem.

[0535] The system of this invention ensures the reliability of communication by instantly processing acoustic data from the caller and converting it into text data. The server collects acoustic data using a microphone and communication module to receive it. Next, the acoustic data is instantly converted into text data using the Google Speech Recognition API. The converted text data is then analyzed using text analysis software such as SpaCy or NLTK to extract caller information and request details.

[0536] The server calculates the caller's trustworthiness based on the extracted data and approves or rejects the communication accordingly. When abnormal communication is detected, it uses AWS Lambda, leveraging cloud functionality, to send a warning notification to the user's device. Upon receiving this notification, the user can quickly take action if the call may be fraudulent.

[0537] For example, when a user receives a call on their smartphone, this system detects calls from suspicious callers and warns the user, "This may be a scam. Do you want to disconnect the call?" This allows users to use communication services with peace of mind and protect themselves from potential dangers.

[0538] An example of a prompt message to input into a generative AI model is, "You have an incoming call. Please activate the speech recognition AI and evaluate the caller's trustworthiness." This prompt allows the server to immediately begin analyzing the voice data and take the necessary actions.

[0539] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0540] Step 1:

[0541] The server detects the incoming signal it receives. The server receives incoming calls from smartphones via the communication network and determines that an incoming call has been received. The input here is the incoming signal, and the output is the start of voice data reception.

[0542] Step 2:

[0543] The server receives audio data in real time and converts it into text data using a speech recognition API. Specifically, the server uses the Google Speech Recognition API to analyze the input audio data and convert it into text. The output is the converted text data.

[0544] Step 3:

[0545] The server analyzes the converted text data using a text analysis tool to extract sender information and request details. This analysis uses either SpaCy or NLTK to interpret the meaning from the input text and obtain the information necessary for reliability evaluation. The output is the extracted sender information and request details.

[0546] Step 4:

[0547] The server uses the extracted data and applies a reliability evaluation algorithm to calculate the caller's reliability. Here, the input data is evaluated based on past communication history and pattern matching, and a reliability score is output.

[0548] Step 5:

[0549] The server determines whether to approve or reject communication based on the confidence score and sends a notification to the user. In particular, if an anomaly is detected, a warning notification is sent via AWS Lambda. Here, the input is the confidence score, and the output is the notification message to the user.

[0550] Step 6:

[0551] The system checks the notifications received by the user and, if necessary, chooses to continue or disconnect the call. The action is determined by the warning message displayed on the user's device. The user's choice is the final output.

[0552] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0553] This invention is a system that receives calls from callers via a communication network and uses AI to perform speech recognition, text analysis, and even emotion recognition, thereby achieving safer and more efficient call management.

[0554] First, the server detects the incoming call on the communication network and sends an automated response to the caller via the AI ​​module. This response asks the caller questions such as, "Who is this?" and "What can I help you with?"

[0555] The server receives the voice response from the caller, and the AI ​​module uses speech recognition to convert it into text data in real time. Furthermore, the converted data is analyzed to extract sender information and requirements. This information is then cross-referenced with the system's database to calculate the caller's trustworthiness.

[0556] Furthermore, this invention incorporates an emotion engine, which allows the server to analyze the caller's emotions in real time from their voice. The results of this emotion analysis are used to adjust the confidence level criteria, and only when the confidence level exceeds a certain standard is it used as a basis for deciding whether to approve the call.

[0557] As an example, consider a use case in a corporate support center. When a customer calls, the server automatically responds, and the customer reports a specific problem. At this time, the system calculates a level of trustworthiness from the customer's voice, while an emotion engine determines the degree of stress and anger. This information becomes important reference information for the operator, forming a foundation for more appropriate responses. For example, if the emotion engine detects a high level of stress, the operator can quickly prioritize the response.

[0558] Furthermore, the server monitors the conversation during the call and automatically disconnects the call if inconsistencies or potential risks are detected, further enhancing security. Additionally, the emotion data recognized by the emotion engine is saved as a log and used for future call analysis and improvement.

[0559] In this way, the system of the present invention aims to provide a higher quality communication environment by comprehensively evaluating the reliability and emotions of the caller. This makes it possible to use it in various use cases, such as safe phone calls for the elderly at home and effective customer service in businesses.

[0560] The following describes the processing flow.

[0561] Step 1:

[0562] The server detects incoming calls on the communication network. Upon receiving a call, it immediately activates the AI ​​module.

[0563] Step 2:

[0564] The server automatically responds to the caller via an AI module, asking questions such as, "Who is this?" and "What can I help you with?"

[0565] Step 3:

[0566] The server receives the caller's voice response. It collects the voice data and passes it to the AI ​​module to start the speech recognition process.

[0567] Step 4:

[0568] An AI module on the server converts audio data into text data in real time. The converted text data is then analyzed to extract sender information and requirements.

[0569] Step 5:

[0570] The server compares the analysis results with the information in the system's database and calculates the sender's trustworthiness. A trustworthiness evaluation algorithm is then applied.

[0571] Step 6:

[0572] The server uses an emotion engine to analyze the emotions in the caller's voice. The analysis results, along with a confidence level, are used to decide whether to approve or reject the call.

[0573] Step 7:

[0574] Based on trust level and sentiment evaluation, the server notifies the user's device whether the call is approved or not. If the call is approved, the user can start the call.

[0575] Step 8:

[0576] During a call, the server uses an AI module and emotion engine to monitor the consistency of the conversation and any changes in the caller's emotions. An anomaly detection process continues.

[0577] Step 9:

[0578] If the server detects inconsistencies or suspicious changes in the conversation content or emotions, it will automatically disconnect the call and notify the user of the problem.

[0579] Step 10:

[0580] The server stores call records and analysis data in a database, which will be used for future improvements and reviews.

[0581] (Example 2)

[0582] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0583] Conventional call management systems have difficulty quickly and accurately assessing the reliability and emotional state of callers, hindering secure and efficient communication management. Furthermore, there has been a lack of effective means to monitor emotional changes and conversational consistency in real time during calls to enhance security; therefore, improvements are needed.

[0584] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0585] In this invention, the server includes means for receiving communications from a caller and converting voice information into text information; means for extracting sender identification data and requirements from the converted text information and calculating trustworthiness; and means for analyzing emotions from the caller's voice in real time and adjusting trustworthiness using the results. This enables a comprehensive evaluation of the caller's trustworthiness and emotional state, allowing for safe and efficient communication management.

[0586] "Caller" refers to the person who sends a call or message via a communication network.

[0587] "Contact" refers to a communication signal in voice or data format from the sender.

[0588] A "server" refers to a computer used to receive, process, and analyze communications from senders.

[0589] "Audio information" refers to the audio signal transmitted by the sender.

[0590] "Text information" refers to character data obtained by converting audio information.

[0591] "Sender identification data" refers to information used to identify the sender.

[0592] "Credibility" refers to the result of quantifying or evaluating the trustworthiness of the sender.

[0593] "Emotion" refers to the psychological and emotional state analyzed from the speaker's voice.

[0594] "Real-time" refers to operations or processes that occur immediately or with very little delay.

[0595] "Monitoring" means continuously monitoring the communication status and collecting and analyzing data.

[0596] "Consistency" refers to a state in which conversations or communications proceed without contradiction.

[0597] "Abnormal" refers to an unusual or irregular situation or condition that differs from the norm.

[0598] The invention will now be described in terms of its embodiments. This system receives calls via a communication network, uses AI technology to convert the caller's voice into text, and analyzes the content to achieve secure and efficient communication. This system primarily operates on a server. The server receives voice from the caller and uses a speech recognition engine to convert the voice data into text data in real time. This speech recognition can utilize general speech recognition software or cloud-based speech recognition APIs.

[0599] Next, the server analyzes the generated text data using a natural language processing engine, such as a common natural language processing tool. This analysis extracts information identifying the sender and their requirements. The extracted information is then compared with an internal database to calculate the sender's credibility.

[0600] The server also features an emotion analysis engine that analyzes the caller's emotional state in real time from the voice data. This emotion analysis uses general emotion recognition software. The analysis results are used to adjust the caller's credibility, and the system approves or rejects the call based on that.

[0601] A concrete example of its use is in a company's customer support center. When a user calls the support center, the system analyzes the voice data and automatically classifies the user's problem. At the same time, the emotion engine determines the user's emotional state, and if the user is stressed, it sends a notification to the operator prompting a quick response. This information contributes to improving the quality of customer service.

[0602] An example of a prompt to a generative AI model is: "Please describe the details of a system that analyzes caller voices and performs real-time emotion recognition to optimize calls in a customer support center."

[0603] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0604] Step 1:

[0605] The server detects an incoming call from the caller via the communication network. The server receives the voice transmitted by the caller and takes this voice data as initial input. Here, the voice data is stored on the server as a digital signal.

[0606] Step 2:

[0607] The server passes this audio data to the speech recognition engine, which converts the digital audio signal into text information. The speech recognition engine analyzes the received audio data, deciphers the audio waveform, and generates corresponding text data. This becomes the output text information.

[0608] Step 3:

[0609] The server, upon receiving the generated text data, passes it to a natural language processing engine. The natural language processing engine analyzes the text data to extract sender identification information and requirements. The output of this step is structured data that shows the sender identification information and requirements.

[0610] Step 4:

[0611] The server then compares specific information with its internal database to calculate the sender's credibility. This credibility is calculated based on the sender's past history and registration information. The output is numerical data indicating trustworthiness.

[0612] Step 5:

[0613] The audio data is then passed back to the emotion analysis engine on the server. The emotion analysis engine analyzes the tone and pitch of the voice and extracts the speaker's emotional state. The output of this step is data that quantitatively represents the speaker's emotional state.

[0614] Step 6:

[0615] The server decides whether to approve or reject the caller's call based on the results of trust and sentiment analysis. If the criteria are met, the call is approved; if not, it is rejected. This output indicates the decision to approve or reject the call.

[0616] Step 7:

[0617] During the call, the server monitors the conversation content and detects inconsistencies or potential risks. If an anomaly is detected, the server automatically disconnects the call. The output of this step indicates the status of whether the call continues or ends.

[0618] Step 8:

[0619] The sentiment analysis results and call history are saved as logs in the database. This can be used for future call analysis and service improvements. The output of this step is the accumulated log information.

[0620] (Application Example 2)

[0621] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0622] In communications, especially when involving the elderly, there is a need to instantly identify the caller, assess their reliability, and determine potential risks during a call, thereby ensuring a safe and stress-free communication environment. Existing systems lack sufficient reliability assessment and sentiment analysis, making it difficult to completely protect against suspicious calls and scams targeting the elderly. Therefore, technology is needed that allows the elderly to communicate safely and securely on their own.

[0623] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0624] In this invention, the server includes means for acquiring the caller's communication, means for converting the acquired voice information into text information, and means for extracting sender identification information and the subject from the converted text information. This enhances the security of calls and allows elderly people to answer the phone with peace of mind.

[0625] "Sender" is a term that refers to the individual or device that initiates a communication.

[0626] "Communication" refers to the act of exchanging information between a sender and a receiver in the form of voice or data.

[0627] "Audio information" refers to data in digital or analog format acquired based on audio signals.

[0628] "Textual information" refers to strings of characters or text data obtained by analyzing audio information.

[0629] "Sender identification information" refers to data used to identify the recipient of a communication, and may include, for example, a phone number or name.

[0630] "Subject" refers to the central content or issue that the sender intends to express in their communication.

[0631] "Reliability" is an indicator of how safe or trustworthy the sender and the content of the communication are.

[0632] "Emotional state" refers to the psychological or emotional state of the speaker and is analyzed from the tone and patterns of their voice.

[0633] "Analysis" refers to the process of examining audio and textual information in detail and extracting or evaluating relevant information.

[0634] This invention is a system that highly manages communications from callers, and is designed in particular to allow elderly people to make calls with peace of mind. The system mainly consists of three elements: a server, a terminal, and a user. The server is connected to a communication network, receives communications from callers, and converts the voice information into text information in real time. This mainly uses the Google Speech-to-Text API. The NLTK library is used to extract sender identification information and subject from the converted text information and to calculate reliability. If the reliability is determined to be sufficient, a notification is sent from the terminal to the user.

[0635] Furthermore, the server analyzes the caller's emotional state based on the acquired audio information. A sentiment analysis model trained using TensorFlow performs this role, evaluating the caller's emotional state. If the emotion is determined to be suspicious, the system automatically rejects the call or issues a warning to the user.

[0636] As a concrete example, consider a scenario where a user is using this system on a mobile device. When a call comes in from an unknown number, the server immediately begins an analysis, and if it detects low reliability or suspicious feelings, it notifies the user that they should not answer the call. This process helps elderly people avoid scams and unwanted contact.

[0637] An example of a prompt sentence to input into a generative AI model is: "Develop a solution that allows elderly people to make phone calls with peace of mind. It will utilize AI technology to evaluate the reliability and sentiment of calls and have a function to identify risky communications."

[0638] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0639] Step 1:

[0640] The server receives communications from the caller via the communication network. In this process, it receives audio data as input. The server is configured to receive the data in an initialized state in order to process the audio data appropriately.

[0641] Step 2:

[0642] The server converts the acquired audio data into text in real time using the Google Speech-to-Text API. The input for this step is audio data, and the output is the converted text. The API analyzes the audio and returns the recognized words as strings.

[0643] Step 3:

[0644] The server uses the NLTK library to extract sender identification information and subject from the converted character data. The input for this step is character data, and the output is the extracted sender identification information and subject. The process involves specific actions to extract important individual information through text analysis of the data.

[0645] Step 4:

[0646] The server calculates the sender's trustworthiness based on the extracted information. The inputs to this step are the extracted identification information and subject, and the output is a trustworthiness score. The process involves calculating safety numerically based on historical data and predefined parameters.

[0647] Step 5:

[0648] The server uses a TensorFlow-based sentiment analysis model to evaluate the caller's emotional state from the acquired audio information. The input for this step is audio data, and the output is data indicating the emotional state. The model analyzes the audio patterns and performs a hierarchical classification of emotions.

[0649] Step 6:

[0650] If the reliability score and emotional state meet certain criteria, the device will notify the user. This notification will include a message indicating sufficient reliability. The input for this step is the reliability score and emotional state, and the output is the notification message. The device will then perform the specific tasks of providing information to the user through display and sound.

[0651] Step 7:

[0652] If the server determines that the caller is unreliable or emotionally unstable, it will either reject the call or send a warning message to the terminal. The input for this step is the result of the reliability and emotional assessment, and the output is the call handling policy (rejection or warning). Database updates and user actions are then performed to deter suspicious communications.

[0653] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0654] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0655] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0656] [Fourth Embodiment]

[0657] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0658] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0659] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0660] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0661] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0662] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0663] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0664] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0665] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0666] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0667] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0668] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0669] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0670] This invention is a system that receives calls from callers via a communication network and detects fraudulent calls early using AI-based speech recognition and text analysis. The system functions as follows:

[0671] First, the server detects an incoming call on the communication network. When a call comes in, the AI ​​module activates and sends an automated response message to the caller. This message asks the caller questions such as, "Who is this?" and "What can I do for you?"

[0672] The caller's response is received as audio data, and real-time speech recognition is performed by an AI module on the server. The audio data is converted into text data, and further text analysis is used to extract information and requirements from the caller. The extracted information is analyzed using historical data registered in the system's database and a reliability evaluation algorithm to calculate the caller's reliability.

[0673] If the server determines that the level of trust is above a certain level, it will send a notification to the user's device and authorize the call. Once authorized, the user's device can start the call. The AI ​​module monitors the conversation throughout the call, checking consistency and coherence of the content in real time. If suspicious behavior is detected, the server will immediately disconnect the call and notify the user.

[0674] As a concrete example, consider a system used at a company's reception desk. When a call comes in to a company, the server automatically answers and states that the caller is a representative of a business partner. At this stage, the system transcribes the name and purpose of the call into text and compares it with past transaction data. If the system confirms that the caller is a legitimate business partner, the call is permitted. If any suspicious requests or inconsistencies are detected during the call, the call is automatically terminated. This allows companies to prevent unnecessary calls and spam attacks.

[0675] In this way, the system of the present invention provides secure and efficient communication through caller reliability evaluation and monitoring of call content. The invention can be used in various scenarios, such as in the homes of the elderly or in corporate reception systems, and reduces the risk of fraud and nuisance calls.

[0676] The following describes the processing flow.

[0677] Step 1:

[0678] The server detects incoming calls on the communication network. When a call is received, the AI ​​module is activated.

[0679] Step 2:

[0680] The server sends an automated response to the caller via an AI module, asking questions such as, "Who is this?" and "What can I help you with?"

[0681] Step 3:

[0682] The server receives the voice response from the caller. The voice data is passed to the AI ​​module, which performs voice recognition in real time.

[0683] Step 4:

[0684] An AI module on the server converts the voice data into text data. This converted data is then analyzed to extract sender information and requirements.

[0685] Step 5:

[0686] The server compares the extracted information with an internal database and uses a reliability evaluation algorithm to calculate the reliability of the sender.

[0687] Step 6:

[0688] Based on the confidence level calculated by the server, if a certain standard is exceeded, the user's device will be notified of permission to make the call. If permission is granted, the user will be able to start the call.

[0689] Step 7:

[0690] During a call, the server uses an AI module to monitor the conversation. It checks for consistency and coherence of the content, enabling it to detect anomalies.

[0691] Step 8:

[0692] If the server detects an anomaly, it will immediately disconnect the call and notify the user of the caller's warning.

[0693] Step 9:

[0694] The server records the call results and analysis information in a database, which can then be used to improve future calls.

[0695] (Example 1)

[0696] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0697] Conventional communication systems have faced challenges in preventing spam and fraudulent activities by untrustworthy callers, as it is difficult to detect calls from fraudulent callers in advance. Furthermore, even after a call has started, there is a lack of reliable methods to monitor whether the content is appropriate, making it impossible to respond quickly to abnormal requests or inconsistent information. Therefore, there is a need for a system that can realize secure and efficient communication.

[0698] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0699] In this invention, the server includes means for detecting communication from a caller, means for sending an automatic response, and means for converting received audio data into text data. This makes it possible to evaluate the reliability of the caller before communication begins and to permit only appropriate calls. Furthermore, it is possible to enhance security by monitoring the content in real time during a call and immediately interrupting the call if an anomaly is detected.

[0700] "Means for detecting communications from callers" refers to devices and technologies for identifying incoming phone calls and voice messages from callers in real time on a communication network.

[0701] "Means of sending automated responses" refers to devices or technologies that automatically play or send pre-set messages in response to a caller's inquiry.

[0702] "Means of converting received audio data into text data" refers to speech recognition technology or software used to convert audio information sent by a caller into text information.

[0703] "Means for extracting sender information and requirements" refers to natural language processing technologies and software used to find and extract specific items or important information from text data.

[0704] "Methods for calculating the trustworthiness of a sender" refers to algorithms or systems that quantify or evaluate the trustworthiness of a sender by comparing extracted information with past data and registered information.

[0705] "Means of permitting or denying communication" refers to mechanical or programmatic devices or technologies for determining and executing whether or not to permit the continuation of communication based on a calculated reliability score.

[0706] "Means of notifying the user" refers to devices or systems that inform the user when communication with the caller is permitted.

[0707] "Means of monitoring conversation content and disconnecting calls if suspicious content is detected" refers to technologies or devices that analyze the exchange during a call in real time and automatically terminate the call if an unusual trend is detected.

[0708] This invention provides a system for managing calls over a communication network more securely and efficiently. The system is constructed from basic components such as a server, terminals, and users.

[0709] The server has the ability to detect incoming calls via the communication network. Specifically, it monitors network communications using VoIP technology and detects incoming calls in real time. It also has an AI module for sending an automated response after an incoming call, which can send a pre-configured message to the caller. The technologies used at this stage include software such as Google Cloud Speech-to-Text API and IBM Watson Speech to Text for speech recognition.

[0710] The received audio data is converted into text data within the server. The converted text is analyzed using natural language processing techniques to extract information and requirements from the caller. Techniques such as entity recognition are used in this process. Based on the extracted information, the server calculates the caller's trustworthiness. A historical database and trustworthiness evaluation algorithm are used to calculate trustworthiness, and the server decides whether to allow or deny the communication based on the resulting score.

[0711] Once communication is permitted, the device notifies the user. This notification is delivered visually and audibly, such as through a pop-up message or audio alert. Once a call begins, the server constantly monitors the consistency and integrity of the conversation, and immediately terminates the call if any anomalies are detected. This allows users to quickly respond to fraudulent callers.

[0712] For example, if this system were implemented in a company's reception system, the server would automatically respond to incoming calls it detects. When the caller claims to be a representative of a client company, their name and purpose would be accurately extracted and quickly compared with past transaction data. If the caller is confirmed to be a trusted caller, the call would be permitted. However, if any fraudulent activity or suspicious requests are detected, the call would be immediately terminated.

[0713] Example prompts for explaining this system to a generative AI model:

[0714] Please describe a system that detects incoming calls and identifies fraudulent calls using AI-powered voice recognition and text analysis. This system ensures secure communication through automated responses and reliability assessment.

[0715] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0716] Step 1:

[0717] The server detects an incoming call from the caller on the communication network. Specifically, the VoIP system monitors network traffic and identifies incoming packets in real time. In this process, the input is a signal arriving via the network, and the output is an incoming call event. Based on this event, the server initiates the next processing step.

[0718] Step 2:

[0719] When the server detects an incoming call, it sends a pre-recorded automated response message to the caller. This automated response is played by playing an audio file stored within the system. The input here is the incoming call event, and the output is the audio response to the caller. The server prompts the caller for a response through this interaction.

[0720] Step 3:

[0721] When the caller's response is received as audio data, the server activates an AI module to perform speech recognition. In this process, the audio data is passed to the AI ​​model as input, and the processing outputs text data. The Google Cloud Speech-to-Text API is used for speech recognition. The server then prepares the converted text to proceed.

[0722] Step 4:

[0723] The server analyzes the text data and extracts sender information and requirements. Using natural language processing techniques, it performs entity recognition to identify key items. In this step, text data is used as input, and the output is the extracted sender information and requirements. This allows the server to prepare the material for reliability assessment.

[0724] Step 5:

[0725] The server calculates the caller's trustworthiness based on the information it extracts. The server refers to past call records and a caller database, and applies a trustworthiness evaluation algorithm. The input in this step is the extracted information, and the output is the calculated trustworthiness score. Based on this, the server determines whether or not to allow the communication.

[0726] Step 6:

[0727] Once the caller's trustworthiness is verified, the server notifies the terminal, informing it that communication is permitted. The terminal displays the notification as a pop-up or audio message, conveying it to the user. The input is the result of the trustworthiness score evaluation, and the output is the notification to the user. Through this, the user knows that the call can be initiated.

[0728] Step 7:

[0729] Once communication begins, the server continuously monitors the conversation. An AI module performs text analysis to check the consistency and coherence of the conversation. Here, the audio from the conversation is processed as input, and the output is a determination of whether or not an anomaly is detected. The server constantly performs this analysis, ensuring that any suspicious activity can be detected immediately.

[0730] Step 8:

[0731] If suspicious content is detected, the server immediately disconnects the call and sends a warning notification to the user. The input for this step is the anomaly detection result, and the output is the termination of the call and the sending of a warning. This allows users to be quickly protected from attacks and fraudulent activities.

[0732] (Application Example 1)

[0733] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0734] In recent years, with the development of information and communication technology, the risks of fraudulent calls and cyberattacks have increased. In particular, fraudulent activities targeting individuals and organizations have become a social problem, and conventional telephone answering systems are sometimes insufficient to address this. To address this challenge, there is a need for the development of a new system that can detect and block suspicious calls early while ensuring the reliability of communications.

[0735] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0736] In this invention, the server includes means for receiving information from the caller, means for converting the received acoustic data into text data, means for extracting caller information and request details from the converted text data, and means for notifying the user of a warning when abnormal communication is detected. This improves the security of communications, enables immediate detection of fraudulent calls, and allows users to be warned.

[0737] "Caller" refers to the person or device that initiates communication and transmits information.

[0738] "Audio data" refers to information in signal format that records sound.

[0739] "Text data" refers to a data format in which audio or other information is represented as text.

[0740] "Reliability" refers to an indicator used to measure the legitimacy and security of the content transmitted by the other party in a communication.

[0741] "Information communication" refers to the act or process of sending and receiving audio, data, and other information.

[0742] "Consistency" refers to the logical coherence and regularity of the content of a conversation or communication.

[0743] "Anomaly detection" refers to the process of identifying patterns or behaviors that are different from the norm.

[0744] "User" refers to an individual or organization that uses the system or its communication functions.

[0745] A "warning" means a notice or message provided to indicate a potential danger or problem.

[0746] The system of this invention ensures the reliability of communication by instantly processing acoustic data from the caller and converting it into text data. The server collects acoustic data using a microphone and communication module to receive it. Next, the acoustic data is instantly converted into text data using the Google Speech Recognition API. The converted text data is then analyzed using text analysis software such as SpaCy or NLTK to extract caller information and request details.

[0747] The server calculates the caller's trustworthiness based on the extracted data and approves or rejects the communication accordingly. When abnormal communication is detected, it uses AWS Lambda, leveraging cloud functionality, to send a warning notification to the user's device. Upon receiving this notification, the user can quickly take action if the call may be fraudulent.

[0748] For example, when a user receives a call on their smartphone, this system detects calls from suspicious callers and warns the user, "This may be a scam. Do you want to disconnect the call?" This allows users to use communication services with peace of mind and protect themselves from potential dangers.

[0749] An example of a prompt message to input into a generative AI model is, "You have an incoming call. Please activate the speech recognition AI and evaluate the caller's trustworthiness." This prompt allows the server to immediately begin analyzing the voice data and take the necessary actions.

[0750] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0751] Step 1:

[0752] The server detects the incoming signal it receives. The server receives incoming calls from smartphones via the communication network and determines that an incoming call has been received. The input here is the incoming signal, and the output is the start of voice data reception.

[0753] Step 2:

[0754] The server receives audio data in real time and converts it into text data using a speech recognition API. Specifically, the server uses the Google Speech Recognition API to analyze the input audio data and convert it into text. The output is the converted text data.

[0755] Step 3:

[0756] The server analyzes the converted text data using a text analysis tool to extract sender information and request details. This analysis uses either SpaCy or NLTK to interpret the meaning from the input text and obtain the information necessary for reliability evaluation. The output is the extracted sender information and request details.

[0757] Step 4:

[0758] The server uses the extracted data and applies a reliability evaluation algorithm to calculate the caller's reliability. Here, the input data is evaluated based on past communication history and pattern matching, and a reliability score is output.

[0759] Step 5:

[0760] The server determines whether to approve or reject communication based on the confidence score and sends a notification to the user. In particular, if an anomaly is detected, a warning notification is sent via AWS Lambda. Here, the input is the confidence score, and the output is the notification message to the user.

[0761] Step 6:

[0762] The system checks the notifications received by the user and, if necessary, chooses to continue or disconnect the call. The action is determined by the warning message displayed on the user's device. The user's choice is the final output.

[0763] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0764] This invention is a system that receives calls from callers via a communication network and uses AI to perform speech recognition, text analysis, and even emotion recognition, thereby achieving safer and more efficient call management.

[0765] First, the server detects the incoming call on the communication network and sends an automated response to the caller via the AI ​​module. This response asks the caller questions such as, "Who is this?" and "What can I help you with?"

[0766] The server receives the voice response from the caller, and the AI ​​module uses speech recognition to convert it into text data in real time. Furthermore, the converted data is analyzed to extract sender information and requirements. This information is then cross-referenced with the system's database to calculate the caller's trustworthiness.

[0767] Furthermore, this invention incorporates an emotion engine, which allows the server to analyze the caller's emotions in real time from their voice. The results of this emotion analysis are used to adjust the confidence level criteria, and only when the confidence level exceeds a certain standard is it used as a basis for deciding whether to approve the call.

[0768] As an example, consider a use case in a corporate support center. When a customer calls, the server automatically responds, and the customer reports a specific problem. At this time, the system calculates a level of trustworthiness from the customer's voice, while an emotion engine determines the degree of stress and anger. This information becomes important reference information for the operator, forming a foundation for more appropriate responses. For example, if the emotion engine detects a high level of stress, the operator can quickly prioritize the response.

[0769] Furthermore, the server monitors the conversation during the call and automatically disconnects the call if inconsistencies or potential risks are detected, further enhancing security. Additionally, the emotion data recognized by the emotion engine is saved as a log and used for future call analysis and improvement.

[0770] In this way, the system of the present invention aims to provide a higher quality communication environment by comprehensively evaluating the reliability and emotions of the caller. This makes it possible to use it in various use cases, such as safe phone calls for the elderly at home and effective customer service in businesses.

[0771] The following describes the processing flow.

[0772] Step 1:

[0773] The server detects incoming calls on the communication network. Upon receiving a call, it immediately activates the AI ​​module.

[0774] Step 2:

[0775] The server automatically responds to the caller via an AI module, asking questions such as, "Who is this?" and "What can I help you with?"

[0776] Step 3:

[0777] The server receives the caller's voice response. It collects the voice data and passes it to the AI ​​module to start the speech recognition process.

[0778] Step 4:

[0779] An AI module on the server converts audio data into text data in real time. The converted text data is then analyzed to extract sender information and requirements.

[0780] Step 5:

[0781] The server compares the analysis results with the information in the system's database and calculates the sender's trustworthiness. A trustworthiness evaluation algorithm is then applied.

[0782] Step 6:

[0783] The server uses an emotion engine to analyze the emotions in the caller's voice. The analysis results, along with a confidence level, are used to decide whether to approve or reject the call.

[0784] Step 7:

[0785] Based on trust level and sentiment evaluation, the server notifies the user's device whether the call is approved or not. If the call is approved, the user can start the call.

[0786] Step 8:

[0787] During a call, the server uses an AI module and emotion engine to monitor the consistency of the conversation and any changes in the caller's emotions. An anomaly detection process continues.

[0788] Step 9:

[0789] If the server detects inconsistencies or suspicious changes in the conversation content or emotions, it will automatically disconnect the call and notify the user of the problem.

[0790] Step 10:

[0791] The server stores call records and analysis data in a database, which will be used for future improvements and reviews.

[0792] (Example 2)

[0793] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0794] Conventional call management systems have difficulty quickly and accurately assessing the reliability and emotional state of callers, hindering secure and efficient communication management. Furthermore, there has been a lack of effective means to monitor emotional changes and conversational consistency in real time during calls to enhance security; therefore, improvements are needed.

[0795] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0796] In this invention, the server includes means for receiving communications from a caller and converting voice information into text information; means for extracting sender identification data and requirements from the converted text information and calculating trustworthiness; and means for analyzing emotions from the caller's voice in real time and adjusting trustworthiness using the results. This enables a comprehensive evaluation of the caller's trustworthiness and emotional state, allowing for safe and efficient communication management.

[0797] "Caller" refers to the person who sends a call or message via a communication network.

[0798] "Contact" refers to a communication signal in voice or data format from the sender.

[0799] A "server" refers to a computer used to receive, process, and analyze communications from senders.

[0800] "Audio information" refers to the audio signal transmitted by the sender.

[0801] "Text information" refers to character data obtained by converting audio information.

[0802] "Sender identification data" refers to information used to identify the sender.

[0803] "Credibility" refers to the result of quantifying or evaluating the trustworthiness of the sender.

[0804] "Emotion" refers to the psychological and emotional state analyzed from the speaker's voice.

[0805] "Real-time" refers to operations or processes that occur immediately or with very little delay.

[0806] "Monitoring" means continuously monitoring the communication status and collecting and analyzing data.

[0807] "Consistency" refers to a state in which conversations or communications proceed without contradiction.

[0808] "Abnormal" refers to an unusual or irregular situation or condition that differs from the norm.

[0809] The invention will now be described in terms of its embodiments. This system receives calls via a communication network, uses AI technology to convert the caller's voice into text, and analyzes the content to achieve secure and efficient communication. This system primarily operates on a server. The server receives voice from the caller and uses a speech recognition engine to convert the voice data into text data in real time. This speech recognition can utilize general speech recognition software or cloud-based speech recognition APIs.

[0810] Next, the server analyzes the generated text data using a natural language processing engine, such as a common natural language processing tool. This analysis extracts information identifying the sender and their requirements. The extracted information is then compared with an internal database to calculate the sender's credibility.

[0811] The server also features an emotion analysis engine that analyzes the caller's emotional state in real time from the voice data. This emotion analysis uses general emotion recognition software. The analysis results are used to adjust the caller's credibility, and the system approves or rejects the call based on that.

[0812] A concrete example of its use is in a company's customer support center. When a user calls the support center, the system analyzes the voice data and automatically classifies the user's problem. At the same time, the emotion engine determines the user's emotional state, and if the user is stressed, it sends a notification to the operator prompting a quick response. This information contributes to improving the quality of customer service.

[0813] An example of a prompt to a generative AI model is: "Please describe the details of a system that analyzes caller voices and performs real-time emotion recognition to optimize calls in a customer support center."

[0814] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0815] Step 1:

[0816] The server detects an incoming call from the caller via the communication network. The server receives the voice transmitted by the caller and takes this voice data as initial input. Here, the voice data is stored on the server as a digital signal.

[0817] Step 2:

[0818] The server passes this audio data to the speech recognition engine, which converts the digital audio signal into text information. The speech recognition engine analyzes the received audio data, deciphers the audio waveform, and generates corresponding text data. This becomes the output text information.

[0819] Step 3:

[0820] The server, upon receiving the generated text data, passes it to a natural language processing engine. The natural language processing engine analyzes the text data to extract sender identification information and requirements. The output of this step is structured data that shows the sender identification information and requirements.

[0821] Step 4:

[0822] The server then compares specific information with its internal database to calculate the sender's credibility. This credibility is calculated based on the sender's past history and registration information. The output is numerical data indicating trustworthiness.

[0823] Step 5:

[0824] The audio data is then passed back to the emotion analysis engine on the server. The emotion analysis engine analyzes the tone and pitch of the voice and extracts the speaker's emotional state. The output of this step is data that quantitatively represents the speaker's emotional state.

[0825] Step 6:

[0826] The server decides whether to approve or reject the caller's call based on the results of trust and sentiment analysis. If the criteria are met, the call is approved; if not, it is rejected. This output indicates the decision to approve or reject the call.

[0827] Step 7:

[0828] During the call, the server monitors the conversation content and detects inconsistencies or potential risks. If an anomaly is detected, the server automatically disconnects the call. The output of this step indicates the status of whether the call continues or ends.

[0829] Step 8:

[0830] The sentiment analysis results and call history are saved as logs in the database. This can be used for future call analysis and service improvements. The output of this step is the accumulated log information.

[0831] (Application Example 2)

[0832] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0833] In communications, especially when involving the elderly, there is a need to instantly identify the caller, assess their reliability, and determine potential risks during a call, thereby ensuring a safe and stress-free communication environment. Existing systems lack sufficient reliability assessment and sentiment analysis, making it difficult to completely protect against suspicious calls and scams targeting the elderly. Therefore, technology is needed that allows the elderly to communicate safely and securely on their own.

[0834] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0835] In this invention, the server includes means for acquiring the caller's communication, means for converting the acquired voice information into text information, and means for extracting sender identification information and the subject from the converted text information. This enhances the security of calls and allows elderly people to answer the phone with peace of mind.

[0836] "Sender" is a term that refers to the individual or device that initiates a communication.

[0837] "Communication" refers to the act of exchanging information between a sender and a receiver in the form of voice or data.

[0838] "Audio information" refers to data in digital or analog format acquired based on audio signals.

[0839] "Textual information" refers to strings of characters or text data obtained by analyzing audio information.

[0840] "Sender identification information" refers to data used to identify the recipient of a communication, and may include, for example, a phone number or name.

[0841] "Subject" refers to the central content or issue that the sender intends to express in their communication.

[0842] "Reliability" is an indicator of how safe or trustworthy the sender and the content of the communication are.

[0843] "Emotional state" refers to the psychological or emotional state of the speaker and is analyzed from the tone and patterns of their voice.

[0844] "Analysis" refers to the process of examining audio and textual information in detail and extracting or evaluating relevant information.

[0845] This invention is a system that highly manages communications from callers, and is designed in particular to allow elderly people to make calls with peace of mind. The system mainly consists of three elements: a server, a terminal, and a user. The server is connected to a communication network, receives communications from callers, and converts the voice information into text information in real time. This mainly uses the Google Speech-to-Text API. The NLTK library is used to extract sender identification information and subject from the converted text information and to calculate reliability. If the reliability is determined to be sufficient, a notification is sent from the terminal to the user.

[0846] Furthermore, the server analyzes the caller's emotional state based on the acquired audio information. A sentiment analysis model trained using TensorFlow performs this role, evaluating the caller's emotional state. If the emotion is determined to be suspicious, the system automatically rejects the call or issues a warning to the user.

[0847] As a concrete example, consider a scenario where a user is using this system on a mobile device. When a call comes in from an unknown number, the server immediately begins an analysis, and if it detects low reliability or suspicious feelings, it notifies the user that they should not answer the call. This process helps elderly people avoid scams and unwanted contact.

[0848] An example of a prompt sentence to input into a generative AI model is: "Develop a solution that allows elderly people to make phone calls with peace of mind. It will utilize AI technology to evaluate the reliability and sentiment of calls and have a function to identify risky communications."

[0849] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0850] Step 1:

[0851] The server receives communications from the caller via the communication network. In this process, it receives audio data as input. The server is configured to receive the data in an initialized state in order to process the audio data appropriately.

[0852] Step 2:

[0853] The server converts the acquired audio data into text in real time using the Google Speech-to-Text API. The input for this step is audio data, and the output is the converted text. The API analyzes the audio and returns the recognized words as strings.

[0854] Step 3:

[0855] The server uses the NLTK library to extract sender identification information and subject from the converted character data. The input for this step is character data, and the output is the extracted sender identification information and subject. The process involves specific actions to extract important individual information through text analysis of the data.

[0856] Step 4:

[0857] The server calculates the sender's trustworthiness based on the extracted information. The inputs to this step are the extracted identification information and subject, and the output is a trustworthiness score. The process involves calculating safety numerically based on historical data and predefined parameters.

[0858] Step 5:

[0859] The server uses a TensorFlow-based sentiment analysis model to evaluate the caller's emotional state from the acquired audio information. The input for this step is audio data, and the output is data indicating the emotional state. The model analyzes the audio patterns and performs a hierarchical classification of emotions.

[0860] Step 6:

[0861] If the reliability score and emotional state meet certain criteria, the device will notify the user. This notification will include a message indicating sufficient reliability. The input for this step is the reliability score and emotional state, and the output is the notification message. The device will then perform the specific tasks of providing information to the user through display and sound.

[0862] Step 7:

[0863] If the server determines that the caller is unreliable or emotionally unstable, it will either reject the call or send a warning message to the terminal. The input for this step is the result of the reliability and emotional assessment, and the output is the call handling policy (rejection or warning). Database updates and user actions are then performed to deter suspicious communications.

[0864] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0865] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0866] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0867] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0868] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0869] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0870] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0871] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0872] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0873] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0874] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0875] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0876] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0877] 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.

[0878] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0879] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0880] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0881] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0882] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0883] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0884] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[0885] The following is further disclosed regarding the embodiments described above.

[0886] (Claim 1)

[0887] A means of receiving communications from the sender,

[0888] A means of converting received audio data into text data,

[0889] A means for extracting sender information and requirements from the converted text data,

[0890] A means for calculating the reliability of the sender based on the extracted information,

[0891] A means of approving or rejecting communications based on a calculated confidence level,

[0892] A means to monitor the consistency of the conversation while communication is ongoing and to disconnect the call if an anomaly is detected,

[0893] A system that includes this.

[0894] (Claim 2)

[0895] The system according to claim 1, comprising means for processing received audio data in real time.

[0896] (Claim 3)

[0897] The system according to claim 1, comprising means for notifying the user when approving a communication.

[0898] "Example 1"

[0899] (Claim 1)

[0900] A means of detecting communication from the caller,

[0901] A means for sending an automatic response to detected communications,

[0902] A means of converting received audio data into text data,

[0903] A means for extracting sender information and requirements from converted text data,

[0904] A means for calculating the trustworthiness of the sender based on the extracted information,

[0905] A means of permitting or denying communication based on the calculated confidence level,

[0906] A means of notifying the user when communication is permitted,

[0907] A means of monitoring the content of the conversation while communication is ongoing and disconnecting the call if suspicious content is detected,

[0908] A system that includes this.

[0909] (Claim 2)

[0910] The system according to claim 1, comprising means for processing audio data in real time, converting it into text data, and then quickly calculating the reliability of the caller.

[0911] (Claim 3)

[0912] The system according to claim 1, comprising means for constantly monitoring the consistency and coherence of the content of a conversation during communication and for automatically suspending the call if an anomaly is detected.

[0913] "Application Example 1"

[0914] (Claim 1)

[0915] A means of receiving information from the caller,

[0916] A means of converting received audio data into text data,

[0917] A means for extracting sender information and request details from the converted text data,

[0918] A means for calculating the reliability of the caller based on the extracted information,

[0919] A means of approving or rejecting the communication of information based on the calculated reliability,

[0920] A means to monitor the consistency of the dialogue while information is being transmitted and to disconnect the information transmission if an anomaly is detected,

[0921] A means of notifying the user of a warning when abnormal communication is detected,

[0922] A system that includes this.

[0923] (Claim 2)

[0924] The system according to claim 1, comprising means for immediately processing received acoustic data.

[0925] (Claim 3)

[0926] The system according to claim 1, comprising means for notifying the user when approving the communication of information.

[0927] "Example 2 of combining an emotion engine"

[0928] (Claim 1)

[0929] Means of receiving contact from the sender,

[0930] A means of converting received audio information into text information,

[0931] A means for extracting sender identification data and requirements from the converted text information,

[0932] A means for calculating the credibility of a sender based on extracted specific data,

[0933] A method for analyzing emotions from the caller's voice in real time and using the results to adjust the caller's credibility,

[0934] A means of permitting or denying communication when the level of trustworthiness exceeds a certain standard,

[0935] A means of monitoring the consistency of the conversation while communication is ongoing and interrupting the call if an anomaly is detected,

[0936] A system that includes this.

[0937] (Claim 2)

[0938] The system according to claim 1, comprising means for processing received audio information in real time and analyzing the emotions of the caller in real time.

[0939] (Claim 3)

[0940] The system according to claim 1, further comprising means for notifying the operator of the sentiment analysis results when granting permission for communication, and for suggesting countermeasures according to the status of the caller.

[0941] "Application example 2 when combining with an emotional engine"

[0942] (Claim 1)

[0943] A device that acquires communications from the caller,

[0944] A device that converts acquired audio information into text information,

[0945] A device for extracting sender identification information and subject from converted character information,

[0946] A device that calculates the reliability of the sender based on extracted information,

[0947] A device that authorizes or denies communications based on calculated reliability,

[0948] A device that monitors the consistency of conversation content during communication and terminates communication if an anomaly is detected,

[0949] A device that analyzes the emotional state of the sender and adjusts the reliability calculation criteria based on the results,

[0950] A system that includes this.

[0951] (Claim 2)

[0952] The system according to claim 1, comprising a device for analyzing acquired audio information in real time.

[0953] (Claim 3)

[0954] The system according to claim 1, further comprising a device that notifies the user when authorizing a communication. [Explanation of Symbols]

[0955] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of receiving information from the caller, A means of converting received audio data into text data, A means for extracting sender information and request details from the converted text data, A means for calculating the reliability of the caller based on the extracted information, A means of approving or rejecting the communication of information based on the calculated reliability, A means to monitor the consistency of the dialogue while information is being transmitted and to disconnect the information transmission if an anomaly is detected, A means of notifying the user of a warning when abnormal communication is detected, A system that includes this.

2. The system according to claim 1, further comprising means for immediately processing received acoustic data.

3. The system according to claim 1, further comprising means for notifying the user when approving the communication of information.

Images