system

Abstract

We provide the system. [Solution] A means of automatically activating an agent when a call starts and collecting information from the other party via the communication device, A means of performing risk assessment by comparing the collected information with existing databases, A means to determine how to process a call based on the risk assessment results, and to transfer the call to the user or display a warning, A system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, 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 order to reduce the damage of nuisance calls and fraud by asking the other party for confirmation and requirements, the conventional telephone system has a problem that efficient and accurate filtering and risk assessment cannot be performed. In particular, there is a need for means to quickly and effectively respond to telephone fraud and nuisance calls targeting the elderly. Against this background, it is necessary to provide a system that allows users to use the telephone with confidence.

Means for Solving the Problems

[0005] This invention provides a means for automatically activating an agent at the start of a call and collecting information from the other party via a communication device. Furthermore, it includes means for comparing the collected information with an existing database to perform a risk assessment. Based on the results of this risk assessment, the invention provides means for determining how to process the call and transferring the call to the user or displaying a warning, thereby minimizing the risk of spam calls and fraud.

[0006] An "agent" is a program designed to automate the initial response to calls and information gathering in a telephone system.

[0007] "Communication devices" refer to electronic devices used to send and receive voice data, such as telephones and smartphones.

[0008] "Means of collecting information" refers to functions for acquiring audio from the other party and analyzing and recording the necessary data.

[0009] An "existing database" is a collection of information that has accumulated patterns of past spam calls and scams, and is used for risk assessment.

[0010] "Risk assessment" is the process of determining the safety of a call using numerical values ​​or ranks based on the information collected.

[0011] "Risk assessment results" refer to safety evaluation indicators calculated based on the content of the call.

[0012] "User" refers to the recipient of a telephone call, and is an individual or legal entity that utilizes the system of the present invention.

[0013] "Transferring a call" means that the agent who initially answered the call takes over the actual call to the user based on the information they obtained.

[0014] "Displaying a warning" means that when the system detects a potential danger in a call, it visually or audibly notifies the user.

Brief Description of the Drawings

[0015] [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 an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Modes for Carrying Out the Invention

[0016] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0017] First, the terms used in the following description will be explained.

[0018] In the following embodiments, the numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0019] In the following embodiments, the numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0020] In the following embodiments, the numbered storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, and the like.

[0021] 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).

[0022] 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."

[0023] [First Embodiment]

[0024] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0025] 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.

[0026] 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).

[0027] 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.

[0028] 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.

[0029] 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.

[0030] 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.

[0031] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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".

[0036] This invention relates to a telephone system that automatically evaluates the risks associated with a call and takes appropriate action. In one embodiment, a server first detects an incoming call and quickly collects information from the caller using an AI agent. The terminal converts the caller's voice into text data using its voice recognition function, and the server performs a risk assessment based on that information.

[0037] During this risk assessment, the server compares the call against a database containing past cases of spam and fraud. This allows the server to evaluate the safety of the call content and determine the next course of action based on the result. Specifically, if the risk is assessed as low, the terminal will smoothly transfer the call to the user. On the other hand, if the risk is determined to be high, the terminal will display a warning to the user to alert them.

[0038] Furthermore, users can provide feedback using options offered after the call ends. The server uses this feedback to improve the AI ​​agent's performance, enabling more accurate risk assessments.

[0039] As a concrete example, when the server detects an incoming call, an agent speaks to the caller saying, "This is the telephone answering system. Please tell me your name and purpose of your call." The terminal transcribes the voice into text in real time and sends it to the server. The server assesses the risk based on this information, and if it determines that the call is secure, the terminal responds, "Please wait a moment. I will connect you to the appropriate person," and connects the call to the user. This entire process helps prevent spam calls and scams, ensuring safe communication.

[0040] The following describes the processing flow.

[0041] Step 1:

[0042] The server detects an incoming call and activates the AI ​​agent. It prepares for the call and initiates the connection with the other party.

[0043] Step 2:

[0044] The device uses voice recognition to automatically ask the person on the other end of the call initial questions such as, "Could you please tell me your name and the purpose of your call?" The collected audio is converted into text data and sent to the server.

[0045] Step 3:

[0046] The server receives the information in text format and inputs it into the risk assessment module. Here, it compares it with existing patterns of spam calls and scams stored in the database and calculates a risk score.

[0047] Step 4:

[0048] The server determines whether the risk is high or low based on the risk assessment. If it is determined to be low risk, the terminal will notify the user with "We will transfer your call" and prepare to connect the call to the user.

[0049] Step 5:

[0050] If the device is deemed high-risk, it will display a warning to the user stating, "This call may be suspicious," and prompt the user to take action. If necessary, it will either terminate the call or offer alternative solutions.

[0051] Step 6:

[0052] Users evaluate a call by selecting feedback from options provided after the call.

[0053] Step 7:

[0054] The server collects user feedback and uses it to update the database and improve the AI ​​agent. This feedback process contributes to further improving the accuracy of risk assessments.

[0055] (Example 1)

[0056] 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."

[0057] In modern society, voice calls over communication networks are commonplace, but nuisance calls and fraudulent calls remain a problem. This presents a challenge for users in ensuring safe and reliable communication. Furthermore, existing systems often fail to effectively incorporate user feedback for improvement. Therefore, a system is needed that automatically assesses call risks and responds appropriately to these problems.

[0058] 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.

[0059] In this invention, the server includes means for detecting incoming calls via a communication network and acquiring acoustic data, means for using a terminal device to convert the acquired acoustic data into text data, and means for comparing the text data with an existing data collection and performing a risk assessment. This makes it possible to quickly assess nuisance calls and fraudulent calls and realize safe and reliable communication.

[0060] A "communication network" is the infrastructure for sending and receiving data and voice through digital or analog communication systems.

[0061] "Incoming call" refers to the arrival of a voice call or message sent via a communication network to the recipient's device.

[0062] "Audio data" refers to data that records or transmits audio signals in digital or analog format.

[0063] "Character data" refers to data in which linguistic information is represented in text format, and is usually generated by character recognition technology.

[0064] A "terminal device" is a device that a user can directly operate and is used for converting voice data and establishing communication.

[0065] A "data collection" is a database that compiles past records and case information, and is used to cross-reference it with other data.

[0066] "Risk assessment" is the process of determining whether a particular event or communication poses a risk to the user, based on the information collected.

[0067] "Feedback" refers to the opinions and evaluations provided by users, and is information used to improve and adjust the system.

[0068] This invention relates to a system for automatically evaluating the risks during a call using a communication network and for achieving secure communication. Detailed embodiments thereof are described below.

[0069] The server detects incoming calls via the communication network. A communication protocol with a function to monitor incoming signals is used. Upon detecting an incoming call, the server activates an AI agent to acquire audio data from the other party. The terminal then converts this audio data into text data using speech recognition software. A general-purpose speech recognition engine is used in this process.

[0070] The text data sent from the terminal to the server is matched by the server against a data collection. This matching process queries a large database and evaluates whether it matches known spam calls or scam cases. Based on this evaluation, the server determines the risk level of the call and decides on the necessary actions.

[0071] As a concrete example, when a server receives an incoming call, the AI ​​agent can say, "This is the answering system. Please tell me your name and purpose of your call." The terminal uses speech recognition software to quickly transcribe the received audio into text and send this information to the server. The server uses this information to perform a risk assessment, and if it determines that the call is safe, it can provide the user with a message through the terminal saying, "Please wait a moment. I will connect you to the appropriate person," and smoothly transfer the call.

[0072] Users can provide feedback through an interface provided after a call. The server aggregates this feedback and uses it to improve the accuracy of the AI ​​model. This is expected to make the system's risk assessment increasingly refined over time.

[0073] An example of a prompt would be, "Please describe the process of converting voice data to text used in the telephone answering system. Also, please describe in detail the risk assessment method using historical data." This makes it easier to understand the detailed operation and processes of the system.

[0074] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0075] Step 1:

[0076] The server detects the incoming call.

[0077] Input: Incoming signal from the communication network

[0078] Specific operation: The server continuously monitors incoming signals via the communication protocol. When an incoming signal is received, the server activates the AI ​​agent.

[0079] Output: Ready to collect audio data from the other party.

[0080] Step 2:

[0081] The server uses an AI agent to acquire acoustic data.

[0082] Input: Caller's voice based on incoming call

[0083] Specific operation: The AI ​​agent speaks to the other party saying, "This is the response system. Please tell me your name and how can I help you?" and records the voice signal.

[0084] Output: Acquired acoustic data

[0085] Step 3:

[0086] The device converts the audio data into text data.

[0087] Input: Acoustic data acquired by the AI ​​agent

[0088] Specific operation: The device uses speech recognition software to convert audio data into text data. A phonological analysis algorithm is used in this conversion process.

[0089] Output: Transcripted call content as text data

[0090] Step 4:

[0091] The server uses text data to perform a risk assessment.

[0092] Input: Transcripted call content as text data

[0093] Specific operation: The server compares the text data with a data collection and performs an evaluation by applying an algorithm that matches it against known spam calls and fraud cases.

[0094] Output: Results of the call risk assessment

[0095] Step 5:

[0096] The server decides on the next action based on the evaluation results.

[0097] Input: Results of the call risk assessment

[0098] Specific actions: If the risk is low, the server instructs the terminal to transfer the call to the user, and the terminal notifies the user with "Please wait a moment. I will connect you to the appropriate person." If the risk is high, a warning message is displayed on the terminal.

[0099] Output: Call forwarding or warning notification to the user.

[0100] Step 6:

[0101] Users provide feedback after the call.

[0102] Input: User feedback on the call experience

[0103] Specific operation: Users use the provided interface to enter feedback about their call experience. This typically includes simple questions and satisfaction ratings.

[0104] Output: Feedback information sent to the server

[0105] Step 7:

[0106] The server improves the AI ​​model based on the feedback.

[0107] Input: Feedback information received from users

[0108] Specific operation: The server analyzes the feedback and adjusts the AI ​​model's algorithm to improve the accuracy of future risk assessments.

[0109] Output: More accurate risk assessment results from the improved AI agent.

[0110] (Application Example 1)

[0111] 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."

[0112] In today's communication environment, users make a large number of calls daily, including spam and fraudulent calls. This increases the risk of users becoming involved in unintended troubles. Furthermore, the management and subsequent analysis of call information are insufficient, making it difficult to maintain a secure communication environment. To solve this problem, there is a need for technology that can quickly and accurately assess the security of calls and provide a secure communication environment for users.

[0113] 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.

[0114] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via an information processing device; means for comparing the collected information with an existing information aggregation medium and performing a risk assessment; and means for converting speech into text data using speech recognition technology and performing an additional risk assessment using the text data. This makes it possible to prevent nuisance calls and fraudulent calls, and to provide users with a safe and secure communication experience.

[0115] "Call initiation" refers to the moment when communication begins, and this is when the system automatically executes the necessary processes.

[0116] An "agent" is a virtual unit of operation responsible for information processing, and is an entity that collects and analyzes speech and text.

[0117] An "information processing device" is an electronic device that plays a role in collecting, analyzing, and evaluating information, and it forms the core of the entire system.

[0118] "Means of collecting information from the other party" refers to the processes and technologies that function to collect information that can be obtained from the person on the other end of a call.

[0119] An "information aggregation medium" is a database or recording system that stores past call data, patterns of spam calls, and other similar information.

[0120] "Means of risk assessment" refers to an evaluation process or algorithm for determining the safety of a call based on collected information.

[0121] "Speech recognition technology" is a technology that converts speech into text data, processing human voices as digital information.

[0122] "Text data" refers to information in text format converted by speech recognition, and is used for further analysis and evaluation.

[0123] To implement this invention, a server and a terminal must work together. The server automatically activates an agent when a call begins and quickly collects information from the other party via an information processing device. This collection utilizes speech recognition technology, where speech data is converted into text data in real time and transmitted to the server.

[0124] The speech recognition process utilizes tools such as Google's Speech-to-Text API, which uses the smartphone's microphone as hardware. The server compares the received text data with existing data aggregation media, such as a database of past spam calls, and performs a risk assessment using an AI model. The risk assessment algorithm is built using machine learning libraries such as Tensorflow and PyTorch.

[0125] Once the evaluation results are generated, the device immediately displays the results on the information display device, i.e., the smartphone's screen. If the evaluation determines that the call is safe, the call is transferred to the user as usual. On the other hand, if the evaluation determines that the risk is high, the device displays a warning to the user, allowing them to choose a course of action.

[0126] For example, if a call is suspected of being a scam call, the system evaluates it and displays a warning message on the device such as, "This may be a spam call." In this way, the user can decide whether or not to answer the call.

[0127] An example of a prompt message to a generative AI model could be, based on information collected at the start of a call, "What is your name and purpose of this call? Please assess the risk based on the voice pattern." This would allow the server to perform a quick and accurate risk assessment.

[0128] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0129] Step 1:

[0130] The server automatically activates the agent as soon as a call begins. The input is the incoming signal, and the output confirms that the agent has been activated. This action prepares the server to begin collecting information from the other party.

[0131] Step 2:

[0132] The terminal uses a communication device to record the other party's voice and converts the voice data into text data using speech recognition technology. The input is voice data, and the output is the converted text data. This data is processed using the Google Speech-to-Text API, and the converted text data is sent to the server.

[0133] Step 3:

[0134] The server compares the received text data with a database and performs a risk assessment using an AI model. The input is text data, and the output is the result of the risk assessment. TensorFlow is used to apply the risk assessment algorithm and generate the assessment results.

[0135] Step 4:

[0136] The server sends the risk assessment results back to the terminal. The input is the risk assessment results, and the output is notification data for the user. The information is transmitted to the terminal to prepare for the next action.

[0137] Step 5:

[0138] The device notifies the user of the call's safety based on the received risk assessment results. The input is notification data, and the output is a displayed message. For low-risk calls, the call is transferred; for high-risk calls, a warning is displayed.

[0139] Step 6:

[0140] After a call ends, the user sends feedback to the server via their device. The input is user feedback, and the output is improvement data. Receiving this feedback helps improve the accuracy of the server's AI model.

[0141] Step 7:

[0142] The server uses feedback information as training data to retrain the AI ​​model, improving the accuracy of risk assessment. The input is the improved data, and the output is the updated AI model. This is then reflected in the next risk assessment.

[0143] 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.

[0144] This embodiment provides a telephone system that combines an emotion engine that recognizes the user's emotions and reflects that information in the risk assessment of the call. First, the server detects an incoming call and monitors the call content in real time using an AI agent. At this time, the terminal converts the voice data into text through speech recognition and sends it to the server.

[0145] In addition, the emotion engine detects the user's emotions through voice analysis. By analyzing the tone, tempo, and other voice characteristics, the server evaluates the user's emotional state in real time. This emotion data is used as part of the risk assessment, and if an abnormal emotional response is detected, the risk score of the call is adjusted.

[0146] As a concrete example, after the server detects an incoming call, the terminal uses AI to ask the caller, "What is your name and why are you calling?" This audio data is first fed into the emotion engine. The server analyzes the emotional nuances contained in the caller's response, and if there are significant signs of anxiety or tension, it incorporates this into the risk assessment. If the risk is deemed high, the terminal displays a warning to the user saying, "Please wait a moment. This call requires caution."

[0147] Based on this, the user decides whether to continue or end the call. After the call ends, the server records all voice and emotional data. The accumulated data is used as training material for the system and helps improve the risk assessment algorithm to ensure user safety. This makes it possible to provide users with a safer and more secure telephone environment.

[0148] The following describes the processing flow.

[0149] Step 1:

[0150] The server detects an incoming call and immediately activates the AI ​​agent. It prepares the call connection and starts a session with the other party.

[0151] Step 2:

[0152] The device utilizes voice recognition to ask the caller initial questions such as, "This is an automated response system. Please tell us your name and purpose of your call." It then converts the caller's voice data into text data and sends it to the server.

[0153] Step 3:

[0154] The device simultaneously activates an emotion engine, analyzing emotional characteristics such as tone, pitch, and voice intensity of the audio data, and generating emotion data in real time. It then sends the emotion recognition results to the server.

[0155] Step 4:

[0156] The server retrieves text data and sentiment data, which are then compared against existing databases. The sentiment data is added as a new input to risk assessment, allowing for a quantitative analysis of the user's level of calmness and anxiety.

[0157] Step 5:

[0158] The server compiles the results of the risk assessment, and if it is determined to be high risk, the terminal displays a warning message to the user saying, "This call requires caution." If it is low risk, it will inform the user that "We will transfer your call" and prepare to connect the call to the user.

[0159] Step 6:

[0160] The user reviews the warning and decides whether to continue, transfer, or end the call. Based on the warning, the user reconfirms the safety of the call.

[0161] Step 7:

[0162] After a call ends, the server saves voice and sentiment data as logs, which are then used in subsequent improvement processes. This data is used to train the system and contribute to improving the accuracy of risk assessment in future calls.

[0163] (Example 2)

[0164] 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".

[0165] In modern telephone systems, accurately assessing the risks during calls and appropriately notifying users is crucial. However, mechanisms for analyzing call content in real time, evaluating emotions, and warning users about problematic calls are not yet well-established. As a result, many users may be unable to properly handle risky calls, potentially leading to problems. In light of this situation, there is a need for a system that enables emotional assessment during calls, calculation of risk scores, and subsequent appropriate responses.

[0166] 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.

[0167] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via a communication device; means for converting the collected voice data into text using speech recognition technology and analyzing the text data for sentiment evaluation; and means for calculating a risk score based on the sentiment evaluation results and displaying a warning to the user if the risk is high. This makes it possible to accurately evaluate the risks during a call and alert the user.

[0168] An "agent" is a program that automatically starts up when a call begins, monitors the call content, and collects necessary information.

[0169] "Communication devices" refers to all equipment and devices used to send and receive audio.

[0170] "Speech recognition technology" is a technology that analyzes speech data and converts it into text.

[0171] "Text data" refers to data in character format converted using speech recognition technology.

[0172] "Emotional assessment methods" refer to processes and algorithms that analyze collected text data to identify the emotions of the caller.

[0173] A "risk score" is a numerical value or indicator that shows the potential risk of a phone call, calculated based on the results of an emotional assessment.

[0174] A "means of displaying warnings" refers to a system function that displays a message to alert the user when a high risk is detected.

[0175] A "data recording means" is a function that stores data collected during a call and uses it for subsequent analysis and system improvement.

[0176] A "generative AI model" is an artificial intelligence model that learns from collected data and is used to improve the performance of a system.

[0177] This invention relates to a system that enables real-time sentiment and risk assessment in a voice call system. In this system, the server automatically activates an agent at the start of a call and collects voice data. Specific hardware components include a microphone and communication devices. Upon receiving the voice, the terminal converts this voice data into text data using speech recognition technology (e.g., a common speech recognition API). This text data is immediately transmitted to the server.

[0178] The server analyzes the received text data using sentiment evaluation tools. This process utilizes specific APIs (e.g., general sentiment analysis APIs) to identify the caller's emotional state. Based on the resulting sentiment data, the server calculates a risk score. If a significant emotional abnormality is detected, this risk score is affected and set higher.

[0179] If the risk score is determined to be high, the device will display a warning to alert the user. Specifically, a message such as "Please wait a moment. This call requires attention." will be displayed. Based on this warning, the user will decide whether to continue or end the call.

[0180] After a call ends, the server records all audio and emotional data. This stored data is then analyzed by a generative AI model and used to improve the system. This process enhances the accuracy of the risk assessment algorithm, resulting in a safer calling environment.

[0181] As a concrete example, during a call, the server plays an automated message saying, "Please tell me your name and purpose of your call." If the caller's response indicates anxiety or tension, this is reflected in the risk score. An example of a prompt to the generating AI model would be, "What emotions do you perceive from the tone and tempo of this caller's voice? Please analyze in as much detail as possible and provide information that will help in the risk assessment."

[0182] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0183] Step 1:

[0184] The server detects an incoming call. Once the call begins, the server automatically activates an agent. The agent collects voice data from the communication device. At this point, the input is the incoming call information, and the output is the activation of the agent.

[0185] Step 2:

[0186] The terminal uses a communication device to record call audio in real time. The recorded audio data is converted into text data using speech recognition technology and sent to the server. The input is audio data, and the output of speech recognition is text data. This process utilizes a speech recognition API.

[0187] Step 3:

[0188] The server analyzes the received text data. Using sentiment evaluation tools, it assesses the user's emotions from the text data. The input is text data, and the output is data with sentiment tags attached. A sentiment analysis API is used for this analysis.

[0189] Step 4:

[0190] The server calculates a risk score based on the emotion assessment results. If an abnormal emotion is detected, the risk score is set higher. The input is emotion-tagged data, and the output is the risk score. For example, if strong anger is detected, the risk score will be higher.

[0191] Step 5:

[0192] The server displays a warning to the user on the terminal based on the calculated risk score. If the risk score is particularly high, the terminal displays the message, "Please wait a moment. This call requires caution." The input is the risk score, and the output is the warning display.

[0193] Step 6:

[0194] Based on the warning from the device, the user decides whether to continue or end the call. At this point, the input is a warning message, and the output will be either to continue or end the call, depending on the user's decision.

[0195] Step 7:

[0196] The server records all call audio data and sentiment evaluation results. This data is later analyzed using a generative AI model. The data is used to improve the accuracy of the risk assessment algorithm. The input in this process is audio and sentiment evaluation data, and the output is stored data.

[0197] (Application Example 2)

[0198] 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."

[0199] Telephone calls present a challenge in responding appropriately to the other party's emotional state. There is a particular need to quickly identify emotional states requiring special attention and ensure call safety. Conventional systems lack risk assessment using emotion recognition, resulting in insufficient quality and security of calls.

[0200] 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.

[0201] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via a communication device, means for comparing the collected information with an existing information set to perform a risk assessment, and means for analyzing voice data to detect emotional characteristics. This makes it possible to reinforce the risk assessment based on emotional data and decide how to handle the call.

[0202] An "agent" is a piece of software or process that automatically performs specific tasks, such as automatically starting up when a call begins and collecting information.

[0203] A "communication device" is an electronic device used to send and receive voice and text data, and capable of processing information in real time.

[0204] An "information collection" is a database or a collection of information that has been collected and organized in the past, and is used as a reference point for risk assessment.

[0205] "Risk assessment" is the process of analyzing collected information, calculating the risks associated with phone calls as numerical values ​​or indicators, and determining safety.

[0206] "Voice data" refers to voice information acquired during a phone call and serves as the basis for analyzing emotional characteristics.

[0207] "Emotional characteristics" are indicators of emotional state determined based on the tone, tempo, and other characteristics of the voice obtained from audio data.

[0208] "Call processing" refers to a series of procedures that determine actions such as continuing, ending, or displaying a warning based on risk assessment results and emotional characteristics.

[0209] The system for implementing the present invention combines a communication method and a signal processing algorithm. In this system, a server first detects the start of a call using a communication device and activates an agent. The agent processes the voice data collected from the other party in real time. The hardware used is assumed to be a smartphone, and the software utilizes a voice recognition service API to convert the voice data into text.

[0210] The server converts the audio data into text data using a speech recognition API. This text data is then analyzed using an emotion analysis library to extract emotional characteristics. Specifically, the tone of voice, tempo, and volume are among the elements analyzed. An example of an emotion analysis library used here is an emotion recognition engine.

[0211] Once voice analysis is complete, the server performs a risk assessment based on emotional characteristics. Here, a generative AI model is used to model the relationship between emotional characteristics and risk, enabling real-time assessment. If the risk is assessed as high, the device displays a warning message to the user. Based on this warning, the user can decide whether to continue or end the call.

[0212] As a concrete example, when a user is on a call with a client who is exhibiting unstable emotions in a work context, this system displays a warning on the smartphone screen saying, "This call requires caution." This warning allows the user to respond flexibly based on the other party's psychological state.

[0213] An example of a prompt message is: "Analyze the emotional state of the person on the other end of this call, and if factors that significantly increase the risk assessment are detected, display a warning to the user."

[0214] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0215] Step 1:

[0216] The server detects the start of a call and automatically activates the agent. At this time, it begins receiving voice data using the communication device. The input is the call start signal, and the output is the activation of the agent.

[0217] Step 2:

[0218] The terminal sends the received audio data to the speech recognition service API. The input is the audio data from the call, which the API analyzes and outputs as text data. Specifically, the process involves converting the audio signal into text.

[0219] Step 3:

[0220] The server sends text data to an emotion analysis library to extract emotional features. The input is transcribed audio data, and the output is emotional features analyzed based on tone, tempo, and volume. Specifically, emotions are identified by quantifying the features.

[0221] Step 4:

[0222] The server uses a generative AI model to assess risk based on emotional characteristics. The input is emotional characteristics, and the output is a risk score. If this risk score exceeds a certain level, the risk is judged to be high. The process involves analyzing emotional data and calculating risk using a learned model based on past data.

[0223] Step 5:

[0224] The device displays warning messages to the user based on a risk score. The input is the risk score, and the output is a warning message displayed on the screen. Specifically, a dialog box appears on the screen to alert the user when a warning is necessary.

[0225] Step 6:

[0226] The user receives a warning message and decides whether to continue or end the call. The input is the warning message displayed on the device, and the output is the user's decision to perform on the call. The action involves choosing whether to continue or disconnect the call.

[0227] 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.

[0228] 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.

[0229] 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.

[0230] [Second Embodiment]

[0231] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0232] 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.

[0233] 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).

[0234] 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.

[0235] 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.

[0236] 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).

[0237] 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.

[0238] 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.

[0239] 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.

[0240] 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.

[0241] 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.

[0242] 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".

[0243] This invention relates to a telephone system that automatically evaluates the risks associated with a call and takes appropriate action. In one embodiment, a server first detects an incoming call and quickly collects information from the caller using an AI agent. The terminal converts the caller's voice into text data using its voice recognition function, and the server performs a risk assessment based on that information.

[0244] During this risk assessment, the server compares the call against a database containing past cases of spam and fraud. This allows the server to evaluate the safety of the call content and determine the next course of action based on the result. Specifically, if the risk is assessed as low, the terminal will smoothly transfer the call to the user. On the other hand, if the risk is determined to be high, the terminal will display a warning to the user to alert them.

[0245] Furthermore, users can provide feedback using options offered after the call ends. The server uses this feedback to improve the AI ​​agent's performance, enabling more accurate risk assessments.

[0246] As a concrete example, when the server detects an incoming call, an agent speaks to the caller saying, "This is the telephone answering system. Please tell me your name and purpose of your call." The terminal transcribes the voice into text in real time and sends it to the server. The server assesses the risk based on this information, and if it determines that the call is secure, the terminal responds, "Please wait a moment. I will connect you to the appropriate person," and connects the call to the user. This entire process helps prevent spam calls and scams, ensuring safe communication.

[0247] The following describes the processing flow.

[0248] Step 1:

[0249] The server detects an incoming call and activates the AI ​​agent. It prepares for the call and initiates the connection with the other party.

[0250] Step 2:

[0251] The device uses voice recognition to automatically ask the person on the other end of the call initial questions such as, "Could you please tell me your name and the purpose of your call?" The collected audio is converted into text data and sent to the server.

[0252] Step 3:

[0253] The server receives the information in text format and inputs it into the risk assessment module. Here, it compares it with existing patterns of spam calls and scams stored in the database and calculates a risk score.

[0254] Step 4:

[0255] The server determines whether the risk is high or low based on the risk assessment. If it is determined to be low risk, the terminal will notify the user with "We will transfer your call" and prepare to connect the call to the user.

[0256] Step 5:

[0257] If the device is deemed high-risk, it will display a warning to the user stating, "This call may be suspicious," and prompt the user to take action. If necessary, it will either terminate the call or offer alternative solutions.

[0258] Step 6:

[0259] Users evaluate a call by selecting feedback from options provided after the call.

[0260] Step 7:

[0261] The server collects user feedback and uses it to update the database and improve the AI ​​agent. This feedback process contributes to further improving the accuracy of risk assessments.

[0262] (Example 1)

[0263] 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."

[0264] In modern society, voice calls over communication networks are commonplace, but nuisance calls and fraudulent calls remain a problem. This presents a challenge for users in ensuring safe and reliable communication. Furthermore, existing systems often fail to effectively incorporate user feedback for improvement. Therefore, a system is needed that automatically assesses call risks and responds appropriately to these problems.

[0265] 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.

[0266] In this invention, the server includes means for detecting incoming calls via a communication network and acquiring acoustic data, means for using a terminal device to convert the acquired acoustic data into text data, and means for comparing the text data with an existing data collection and performing a risk assessment. This makes it possible to quickly assess nuisance calls and fraudulent calls and realize safe and reliable communication.

[0267] A "communication network" is the infrastructure for sending and receiving data and voice through digital or analog communication systems.

[0268] "Incoming call" refers to the arrival of a voice call or message sent via a communication network to the recipient's device.

[0269] "Audio data" refers to data that records or transmits audio signals in digital or analog format.

[0270] "Character data" refers to data in which linguistic information is represented in text format, and is usually generated by character recognition technology.

[0271] A "terminal device" is a device that a user can directly operate and is used for converting voice data and establishing communication.

[0272] A "data collection" is a database that compiles past records and case information, and is used to cross-reference it with other data.

[0273] "Risk assessment" is the process of determining whether a particular event or communication poses a risk to the user, based on the information collected.

[0274] "Feedback" refers to the opinions and evaluations provided by users, and is information used to improve and adjust the system.

[0275] This invention relates to a system for automatically evaluating the risks during a call using a communication network and for achieving secure communication. Detailed embodiments thereof are described below.

[0276] The server detects incoming calls via the communication network. A communication protocol with a function to monitor incoming signals is used. Upon detecting an incoming call, the server activates an AI agent to acquire audio data from the other party. The terminal then converts this audio data into text data using speech recognition software. A general-purpose speech recognition engine is used in this process.

[0277] The text data sent from the terminal to the server is matched by the server against a data collection. This matching process queries a large database and evaluates whether it matches known spam calls or scam cases. Based on this evaluation, the server determines the risk level of the call and decides on the necessary actions.

[0278] As a specific example, when the server receives an incoming call, the AI agent can say, "This is the response system. Please tell me your name and the matter." On the terminal, the voice recognition software quickly converts the received voice into text and transmits this information to the server. The server uses this information to perform a risk assessment. If it is determined to be safe, the server provides the user with a message "Please wait a moment. I will connect you to the person in charge." through the terminal and can smoothly transfer the call.

[0279] The user can input feedback through the interface provided after the call. The server aggregates this feedback and uses it to improve the accuracy of the AI model. It is expected that as time passes, the risk assessment of the system will become increasingly refined.

[0280] Examples of prompt sentences include "Please explain the text conversion process of the voice data used in the telephone response system. Also, please describe in detail the method of risk assessment using past data." This makes it easier to understand the detailed operations and processes of the system.

[0281] The flow of the specific process in Example 1 will be described using FIG. 11.

[0282] Step 1:

[0283] The server detects an incoming call.

[0284] Input: Incoming signal from the communication network

[0285] Specific operation: The server continuously monitors the incoming signal through the communication protocol. When the incoming signal is received, the server activates the AI agent.

[0286] Output: A state where preparations are complete to collect acoustic data from the other party

[0287] Step 2:

[0288] The server uses an AI agent to acquire acoustic data.

[0289] Input: Caller's voice based on incoming call

[0290] Specific operation: The AI ​​agent speaks to the other party saying, "This is the response system. Please tell me your name and how can I help you?" and records the voice signal.

[0291] Output: Acquired acoustic data

[0292] Step 3:

[0293] The device converts the audio data into text data.

[0294] Input: Acoustic data acquired by the AI ​​agent

[0295] Specific operation: The device uses speech recognition software to convert audio data into text data. A phonological analysis algorithm is used in this conversion process.

[0296] Output: Transcripted call content as text data

[0297] Step 4:

[0298] The server uses text data to perform a risk assessment.

[0299] Input: Transcripted call content as text data

[0300] Specific operation: The server compares the text data with a data collection and performs an evaluation by applying an algorithm that matches it against known spam calls and fraud cases.

[0301] Output: Results of the call risk assessment

[0302] Step 5:

[0303] The server determines the next action based on the evaluation result.

[0304] Input: Result of call risk assessment

[0305] Specific action: If the risk is low, the server instructs the terminal to transfer the call to the user, and the terminal notifies the user with "Please wait a moment. I will connect you to the responsible person." If the risk is high, a warning message is displayed on the terminal.

[0306] Output: Notification of call transfer or warning to the user

[0307] Step 6:

[0308] The user provides feedback after the call.

[0309] Input: User's opinion on the call experience

[0310] Specific action: The user uses the provided interface to input feedback on the call experience. This generally includes simple questions and satisfaction evaluations.

[0311] Output: Feedback information sent to the server

[0312] Step 7:

[0313] The server improves the AI model based on the feedback.

[0314] Input: Feedback information received from the user

[0315] Specific action: The server analyzes the feedback and adjusts the algorithm of the AI model to improve the accuracy of future risk assessments.

[0316] Output: More accurate risk assessment results from the improved AI agent.

[0317] (Application Example 1)

[0318] 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."

[0319] In today's communication environment, users make a large number of calls daily, including spam and fraudulent calls. This increases the risk of users becoming involved in unintended troubles. Furthermore, the management and subsequent analysis of call information are insufficient, making it difficult to maintain a secure communication environment. To solve this problem, there is a need for technology that can quickly and accurately assess the security of calls and provide a secure communication environment for users.

[0320] 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.

[0321] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via an information processing device; means for comparing the collected information with an existing information aggregation medium and performing a risk assessment; and means for converting speech into text data using speech recognition technology and performing an additional risk assessment using the text data. This makes it possible to prevent nuisance calls and fraudulent calls, and to provide users with a safe and secure communication experience.

[0322] "Call initiation" refers to the moment when communication begins, and this is when the system automatically executes the necessary processes.

[0323] An "agent" is a virtual unit of operation responsible for information processing, and is an entity that collects and analyzes speech and text.

[0324] An "information processing device" is an electronic device that plays a role in collecting, analyzing, and evaluating information, and it forms the core of the entire system.

[0325] "Means of collecting information from the other party" refers to the processes and technologies that function to collect information that can be obtained from the person on the other end of a call.

[0326] An "information aggregation medium" is a database or recording system that stores past call data, patterns of spam calls, and other similar information.

[0327] "Means of risk assessment" refers to an evaluation process or algorithm for determining the safety of a call based on collected information.

[0328] "Speech recognition technology" is a technology that converts speech into text data, processing human voices as digital information.

[0329] "Text data" refers to information in text format converted by speech recognition, and is used for further analysis and evaluation.

[0330] To implement this invention, a server and a terminal must work together. The server automatically activates an agent when a call begins and quickly collects information from the other party via an information processing device. This collection utilizes speech recognition technology, where speech data is converted into text data in real time and transmitted to the server.

[0331] The speech recognition process utilizes the Google Speech-to-Text API, among others, and uses the smartphone's microphone as hardware. The server compares the received text data with existing data aggregation media, such as a database of past spam calls, and performs a risk assessment using an AI model. The risk assessment algorithm is built using machine learning libraries such as TensorFlow and PyTorch.

[0332] Once the evaluation results are generated, the device immediately displays the results on the information display device, i.e., the smartphone's screen. If the evaluation determines that the call is safe, the call is transferred to the user as usual. On the other hand, if the evaluation determines that the risk is high, the device displays a warning to the user, allowing them to choose a course of action.

[0333] For example, if a call is suspected of being a scam call, the system evaluates it and displays a warning message on the device such as, "This may be a spam call." In this way, the user can decide whether or not to answer the call.

[0334] An example of a prompt message to a generative AI model could be, based on information collected at the start of a call, "What is your name and purpose of this call? Please assess the risk based on the voice pattern." This would allow the server to perform a quick and accurate risk assessment.

[0335] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0336] Step 1:

[0337] The server automatically activates the agent as soon as a call begins. The input is the incoming signal, and the output confirms that the agent has been activated. This action prepares the server to begin collecting information from the other party.

[0338] Step 2:

[0339] The terminal uses a communication device to record the other party's voice and converts the voice data into text data using speech recognition technology. The input is voice data, and the output is the converted text data. This data is processed using the Google Speech-to-Text API, and the converted text data is sent to the server.

[0340] Step 3:

[0341] The server compares the received text data with a database and performs a risk assessment using an AI model. The input is text data, and the output is the result of the risk assessment. TensorFlow is used to apply the risk assessment algorithm and generate the assessment results.

[0342] Step 4:

[0343] The server sends the risk assessment results back to the terminal. The input is the risk assessment results, and the output is notification data for the user. The information is transmitted to the terminal to prepare for the next action.

[0344] Step 5:

[0345] The device notifies the user of the call's safety based on the received risk assessment results. The input is notification data, and the output is a displayed message. For low-risk calls, the call is transferred; for high-risk calls, a warning is displayed.

[0346] Step 6:

[0347] After a call ends, the user sends feedback to the server via their device. The input is user feedback, and the output is improvement data. Receiving this feedback helps improve the accuracy of the server's AI model.

[0348] Step 7:

[0349] The server uses feedback information as training data to retrain the AI ​​model, improving the accuracy of risk assessment. The input is the improved data, and the output is the updated AI model. This is then reflected in the next risk assessment.

[0350] 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.

[0351] This embodiment provides a telephone system that combines an emotion engine that recognizes the user's emotions and reflects that information in the risk assessment of the call. First, the server detects an incoming call and monitors the call content in real time using an AI agent. At this time, the terminal converts the voice data into text through speech recognition and sends it to the server.

[0352] In addition, the emotion engine detects the user's emotions through voice analysis. By analyzing the tone, tempo, and other voice characteristics, the server evaluates the user's emotional state in real time. This emotion data is used as part of the risk assessment, and if an abnormal emotional response is detected, the risk score of the call is adjusted.

[0353] As a concrete example, after the server detects an incoming call, the terminal uses AI to ask the caller, "What is your name and why are you calling?" This audio data is first fed into the emotion engine. The server analyzes the emotional nuances contained in the caller's response, and if there are significant signs of anxiety or tension, it incorporates this into the risk assessment. If the risk is deemed high, the terminal displays a warning to the user saying, "Please wait a moment. This call requires caution."

[0354] Based on this, the user decides whether to continue or end the call. After the call ends, the server records all voice and emotional data. The accumulated data is used as training material for the system and helps improve the risk assessment algorithm to ensure user safety. This makes it possible to provide users with a safer and more secure telephone environment.

[0355] The following describes the processing flow.

[0356] Step 1:

[0357] The server detects an incoming call and immediately activates the AI ​​agent. It prepares the call connection and starts a session with the other party.

[0358] Step 2:

[0359] The device utilizes voice recognition to ask the caller initial questions such as, "This is an automated response system. Please tell us your name and purpose of your call." It then converts the caller's voice data into text data and sends it to the server.

[0360] Step 3:

[0361] The device simultaneously activates an emotion engine, analyzing emotional characteristics such as tone, pitch, and voice intensity of the audio data, and generating emotion data in real time. It then sends the emotion recognition results to the server.

[0362] Step 4:

[0363] The server retrieves text data and sentiment data, which are then compared against existing databases. The sentiment data is added as a new input to risk assessment, allowing for a quantitative analysis of the user's level of calmness and anxiety.

[0364] Step 5:

[0365] The server compiles the results of the risk assessment, and if it is determined to be high risk, the terminal displays a warning message to the user saying, "This call requires caution." If it is low risk, it will inform the user that "We will transfer your call" and prepare to connect the call to the user.

[0366] Step 6:

[0367] The user reviews the warning and decides whether to continue, transfer, or end the call. Based on the warning, the user reconfirms the safety of the call.

[0368] Step 7:

[0369] After a call ends, the server saves voice and sentiment data as logs, which are then used in subsequent improvement processes. This data is used to train the system and contribute to improving the accuracy of risk assessment in future calls.

[0370] (Example 2)

[0371] 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".

[0372] In modern telephone systems, accurately assessing the risks during calls and appropriately notifying users is crucial. However, mechanisms for analyzing call content in real time, evaluating emotions, and warning users about problematic calls are not yet well-established. As a result, many users may be unable to properly handle risky calls, potentially leading to problems. In light of this situation, there is a need for a system that enables emotional assessment during calls, calculation of risk scores, and subsequent appropriate responses.

[0373] 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.

[0374] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via a communication device; means for converting the collected voice data into text using speech recognition technology and analyzing the text data for sentiment evaluation; and means for calculating a risk score based on the sentiment evaluation results and displaying a warning to the user if the risk is high. This makes it possible to accurately evaluate the risks during a call and alert the user.

[0375] An "agent" is a program that automatically starts up when a call begins, monitors the call content, and collects necessary information.

[0376] "Communication devices" refers to all equipment and devices used to send and receive audio.

[0377] "Speech recognition technology" is a technology that analyzes speech data and converts it into text.

[0378] "Text data" refers to data in character format converted using speech recognition technology.

[0379] "Emotional assessment methods" refer to processes and algorithms that analyze collected text data to identify the emotions of the caller.

[0380] A "risk score" is a numerical value or indicator that shows the potential risk of a phone call, calculated based on the results of an emotional assessment.

[0381] A "means of displaying warnings" refers to a system function that displays a message to alert the user when a high risk is detected.

[0382] A "data recording means" is a function that stores data collected during a call and uses it for subsequent analysis and system improvement.

[0383] A "generative AI model" is an artificial intelligence model that learns from collected data and is used to improve the performance of a system.

[0384] This invention relates to a system that enables real-time sentiment and risk assessment in a voice call system. In this system, the server automatically activates an agent at the start of a call and collects voice data. Specific hardware components include a microphone and communication devices. Upon receiving the voice, the terminal converts this voice data into text data using speech recognition technology (e.g., a common speech recognition API). This text data is immediately transmitted to the server.

[0385] The server analyzes the received text data using sentiment evaluation tools. This process utilizes specific APIs (e.g., general sentiment analysis APIs) to identify the caller's emotional state. Based on the resulting sentiment data, the server calculates a risk score. If a significant emotional abnormality is detected, this risk score is affected and set higher.

[0386] If the risk score is determined to be high, the device will display a warning to alert the user. Specifically, a message such as "Please wait a moment. This call requires attention." will be displayed. Based on this warning, the user will decide whether to continue or end the call.

[0387] After a call ends, the server records all audio and emotional data. This stored data is then analyzed by a generative AI model and used to improve the system. This process enhances the accuracy of the risk assessment algorithm, resulting in a safer calling environment.

[0388] As a concrete example, during a call, the server plays an automated message saying, "Please tell me your name and purpose of your call." If the caller's response indicates anxiety or tension, this is reflected in the risk score. An example of a prompt to the generating AI model would be, "What emotions do you perceive from the tone and tempo of this caller's voice? Please analyze in as much detail as possible and provide information that will help in the risk assessment."

[0389] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0390] Step 1:

[0391] The server detects an incoming call. Once the call begins, the server automatically activates an agent. The agent collects voice data from the communication device. At this point, the input is the incoming call information, and the output is the activation of the agent.

[0392] Step 2:

[0393] The terminal uses a communication device to record call audio in real time. The recorded audio data is converted into text data using speech recognition technology and sent to the server. The input is audio data, and the output of speech recognition is text data. This process utilizes a speech recognition API.

[0394] Step 3:

[0395] The server analyzes the received text data. Using sentiment evaluation tools, it assesses the user's emotions from the text data. The input is text data, and the output is data with sentiment tags attached. A sentiment analysis API is used for this analysis.

[0396] Step 4:

[0397] The server calculates a risk score based on the emotion assessment results. If an abnormal emotion is detected, the risk score is set higher. The input is emotion-tagged data, and the output is the risk score. For example, if strong anger is detected, the risk score will be higher.

[0398] Step 5:

[0399] The server displays a warning to the user on the terminal based on the calculated risk score. If the risk score is particularly high, the terminal displays the message, "Please wait a moment. This call requires caution." The input is the risk score, and the output is the warning display.

[0400] Step 6:

[0401] Based on the warning from the device, the user decides whether to continue or end the call. At this point, the input is a warning message, and the output will be either to continue or end the call, depending on the user's decision.

[0402] Step 7:

[0403] The server records all call audio data and sentiment evaluation results. This data is later analyzed using a generative AI model. The data is used to improve the accuracy of the risk assessment algorithm. The input in this process is audio and sentiment evaluation data, and the output is stored data.

[0404] (Application Example 2)

[0405] 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 as the "terminal".

[0406] Telephone calls present a challenge in responding appropriately to the other party's emotional state. There is a particular need to quickly identify emotional states requiring special attention and ensure call safety. Conventional systems lack risk assessment using emotion recognition, resulting in insufficient quality and security of calls.

[0407] 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.

[0408] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via a communication device, means for comparing the collected information with an existing information set to perform a risk assessment, and means for analyzing voice data to detect emotional characteristics. This makes it possible to reinforce the risk assessment based on emotional data and decide how to handle the call.

[0409] An "agent" is a piece of software or process that automatically performs specific tasks, such as automatically starting up when a call begins and collecting information.

[0410] A "communication device" is an electronic device used to send and receive voice and text data, and capable of processing information in real time.

[0411] An "information collection" is a database or a collection of information that has been collected and organized in the past, and is used as a reference point for risk assessment.

[0412] "Risk assessment" is the process of analyzing collected information, calculating the risks associated with phone calls as numerical values ​​or indicators, and determining safety.

[0413] "Voice data" refers to voice information acquired during a phone call and serves as the basis for analyzing emotional characteristics.

[0414] "Emotional characteristics" are indicators of emotional state determined based on the tone, tempo, and other characteristics of the voice obtained from audio data.

[0415] "Call processing" refers to a series of procedures that determine actions such as continuing, ending, or displaying a warning based on risk assessment results and emotional characteristics.

[0416] The system for implementing the present invention combines a communication method and a signal processing algorithm. In this system, a server first detects the start of a call using a communication device and activates an agent. The agent processes the voice data collected from the other party in real time. The hardware used is assumed to be a smartphone, and the software utilizes a voice recognition service API to convert the voice data into text.

[0417] The server converts the audio data into text data using a speech recognition API. This text data is then analyzed using an emotion analysis library to extract emotional characteristics. Specifically, the tone of voice, tempo, and volume are among the elements analyzed. An example of an emotion analysis library used here is an emotion recognition engine.

[0418] Once voice analysis is complete, the server performs a risk assessment based on emotional characteristics. Here, a generative AI model is used to model the relationship between emotional characteristics and risk, enabling real-time assessment. If the risk is assessed as high, the device displays a warning message to the user. Based on this warning, the user can decide whether to continue or end the call.

[0419] As a concrete example, when a user is on a call with a client who is exhibiting unstable emotions in a work context, this system displays a warning on the smartphone screen saying, "This call requires caution." This warning allows the user to respond flexibly based on the other party's psychological state.

[0420] An example of a prompt message is: "Analyze the emotional state of the person on the other end of this call, and if factors that significantly increase the risk assessment are detected, display a warning to the user."

[0421] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0422] Step 1:

[0423] The server detects the start of a call and automatically activates the agent. At this time, it begins receiving voice data using the communication device. The input is the call start signal, and the output is the activation of the agent.

[0424] Step 2:

[0425] The terminal sends the received audio data to the speech recognition service API. The input is the audio data from the call, which the API analyzes and outputs as text data. Specifically, the process involves converting the audio signal into text.

[0426] Step 3:

[0427] The server sends text data to an emotion analysis library to extract emotional features. The input is transcribed audio data, and the output is emotional features analyzed based on tone, tempo, and volume. Specifically, emotions are identified by quantifying the features.

[0428] Step 4:

[0429] The server uses a generative AI model to assess risk based on emotional characteristics. The input is emotional characteristics, and the output is a risk score. If this risk score exceeds a certain level, the risk is judged to be high. The process involves analyzing emotional data and calculating risk using a learned model based on past data.

[0430] Step 5:

[0431] The device displays warning messages to the user based on a risk score. The input is the risk score, and the output is a warning message displayed on the screen. Specifically, a dialog box appears on the screen to alert the user when a warning is necessary.

[0432] Step 6:

[0433] The user receives a warning message and decides whether to continue or end the call. The input is the warning message displayed on the device, and the output is the user's decision to perform on the call. The action involves choosing whether to continue or disconnect the call.

[0434] 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.

[0435] 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.

[0436] 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.

[0437] [Third Embodiment]

[0438] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0439] 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.

[0440] 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).

[0441] 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.

[0442] 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.

[0443] 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).

[0444] 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.

[0445] 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.

[0446] 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.

[0447] 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.

[0448] 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.

[0449] 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".

[0450] This invention relates to a telephone system that automatically evaluates the risks associated with a call and takes appropriate action. In one embodiment, a server first detects an incoming call and quickly collects information from the caller using an AI agent. The terminal converts the caller's voice into text data using its voice recognition function, and the server performs a risk assessment based on that information.

[0451] During this risk assessment, the server compares the call against a database containing past cases of spam and fraud. This allows the server to evaluate the safety of the call content and determine the next course of action based on the result. Specifically, if the risk is assessed as low, the terminal will smoothly transfer the call to the user. On the other hand, if the risk is determined to be high, the terminal will display a warning to the user to alert them.

[0452] Furthermore, users can provide feedback using options offered after the call ends. The server uses this feedback to improve the AI ​​agent's performance, enabling more accurate risk assessments.

[0453] As a concrete example, when the server detects an incoming call, an agent speaks to the caller saying, "This is the telephone answering system. Please tell me your name and purpose of your call." The terminal transcribes the voice into text in real time and sends it to the server. The server assesses the risk based on this information, and if it determines that the call is secure, the terminal responds, "Please wait a moment. I will connect you to the appropriate person," and connects the call to the user. This entire process helps prevent spam calls and scams, ensuring safe communication.

[0454] The following describes the processing flow.

[0455] Step 1:

[0456] The server detects an incoming call and activates the AI ​​agent. It prepares for the call and initiates the connection with the other party.

[0457] Step 2:

[0458] The device uses voice recognition to automatically ask the person on the other end of the call initial questions such as, "Could you please tell me your name and the purpose of your call?" The collected audio is converted into text data and sent to the server.

[0459] Step 3:

[0460] The server receives the information in text format and inputs it into the risk assessment module. Here, it compares it with existing patterns of spam calls and scams stored in the database and calculates a risk score.

[0461] Step 4:

[0462] The server determines whether the risk is high or low based on the risk assessment. If it is determined to be low risk, the terminal will notify the user with "We will transfer your call" and prepare to connect the call to the user.

[0463] Step 5:

[0464] If the device is deemed high-risk, it will display a warning to the user stating, "This call may be suspicious," and prompt the user to take action. If necessary, it will either terminate the call or offer alternative solutions.

[0465] Step 6:

[0466] Users evaluate a call by selecting feedback from options provided after the call.

[0467] Step 7:

[0468] The server collects user feedback and uses it to update the database and improve the AI ​​agent. This feedback process contributes to further improving the accuracy of risk assessments.

[0469] (Example 1)

[0470] 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."

[0471] In modern society, voice calls over communication networks are commonplace, but nuisance calls and fraudulent calls remain a problem. This presents a challenge for users in ensuring safe and reliable communication. Furthermore, existing systems often fail to effectively incorporate user feedback for improvement. Therefore, a system is needed that automatically assesses call risks and responds appropriately to these problems.

[0472] 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.

[0473] In this invention, the server includes means for detecting incoming calls via a communication network and acquiring acoustic data, means for using a terminal device to convert the acquired acoustic data into text data, and means for comparing the text data with an existing data collection and performing a risk assessment. This makes it possible to quickly assess nuisance calls and fraudulent calls and realize safe and reliable communication.

[0474] A "communication network" is the infrastructure for sending and receiving data and voice through digital or analog communication systems.

[0475] "Incoming call" refers to the arrival of a voice call or message sent via a communication network to the recipient's device.

[0476] "Audio data" refers to data that records or transmits audio signals in digital or analog format.

[0477] "Character data" refers to data in which linguistic information is represented in text format, and is usually generated by character recognition technology.

[0478] A "terminal device" is a device that a user can directly operate and is used for converting voice data and establishing communication.

[0479] A "data collection" is a database that compiles past records and case information, and is used to cross-reference it with other data.

[0480] "Risk assessment" is the process of determining whether a particular event or communication poses a risk to the user, based on the information collected.

[0481] "Feedback" refers to the opinions and evaluations provided by users, and is information used to improve and adjust the system.

[0482] This invention relates to a system for automatically evaluating the risks during a call using a communication network and for achieving secure communication. Detailed embodiments thereof are described below.

[0483] The server detects incoming calls via the communication network. A communication protocol with a function to monitor incoming signals is used. Upon detecting an incoming call, the server activates an AI agent to acquire audio data from the other party. The terminal then converts this audio data into text data using speech recognition software. A general-purpose speech recognition engine is used in this process.

[0484] The text data sent from the terminal to the server is matched by the server against a data collection. This matching process queries a large database and evaluates whether it matches known spam calls or scam cases. Based on this evaluation, the server determines the risk level of the call and decides on the necessary actions.

[0485] As a concrete example, when a server receives an incoming call, the AI ​​agent can say, "This is the answering system. Please tell me your name and purpose of your call." The terminal uses speech recognition software to quickly transcribe the received audio into text and send this information to the server. The server uses this information to perform a risk assessment, and if it determines that the call is safe, it can provide the user with a message through the terminal saying, "Please wait a moment. I will connect you to the appropriate person," and smoothly transfer the call.

[0486] Users can provide feedback through an interface provided after a call. The server aggregates this feedback and uses it to improve the accuracy of the AI ​​model. This is expected to make the system's risk assessment increasingly refined over time.

[0487] An example of a prompt would be, "Please describe the process of converting voice data to text used in the telephone answering system. Also, please describe in detail the risk assessment method using historical data." This makes it easier to understand the detailed operation and processes of the system.

[0488] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0489] Step 1:

[0490] The server detects the incoming call.

[0491] Input: Incoming signal from the communication network

[0492] Specific operation: The server continuously monitors incoming signals via the communication protocol. When an incoming signal is received, the server activates the AI ​​agent.

[0493] Output: Ready to collect audio data from the other party.

[0494] Step 2:

[0495] The server uses an AI agent to acquire acoustic data.

[0496] Input: Caller's voice based on incoming call

[0497] Specific operation: The AI ​​agent speaks to the other party saying, "This is the response system. Please tell me your name and how can I help you?" and records the voice signal.

[0498] Output: Acquired acoustic data

[0499] Step 3:

[0500] The device converts the audio data into text data.

[0501] Input: Acoustic data acquired by the AI ​​agent

[0502] Specific operation: The device uses speech recognition software to convert audio data into text data. A phonological analysis algorithm is used in this conversion process.

[0503] Output: Transcripted call content as text data

[0504] Step 4:

[0505] The server uses text data to perform a risk assessment.

[0506] Input: Transcripted call content as text data

[0507] Specific operation: The server compares the text data with a data collection and performs an evaluation by applying an algorithm that matches it against known spam calls and fraud cases.

[0508] Output: Results of the call risk assessment

[0509] Step 5:

[0510] The server decides on the next action based on the evaluation results.

[0511] Input: Results of the call risk assessment

[0512] Specific actions: If the risk is low, the server instructs the terminal to transfer the call to the user, and the terminal notifies the user with "Please wait a moment. I will connect you to the appropriate person." If the risk is high, a warning message is displayed on the terminal.

[0513] Output: Call forwarding or warning notification to the user.

[0514] Step 6:

[0515] Users provide feedback after the call.

[0516] Input: User feedback on the call experience

[0517] Specific operation: Users use the provided interface to enter feedback about their call experience. This typically includes simple questions and satisfaction ratings.

[0518] Output: Feedback information sent to the server

[0519] Step 7:

[0520] The server improves the AI ​​model based on the feedback.

[0521] Input: Feedback information received from users

[0522] Specific operation: The server analyzes the feedback and adjusts the AI ​​model's algorithm to improve the accuracy of future risk assessments.

[0523] Output: More accurate risk assessment results from the improved AI agent.

[0524] (Application Example 1)

[0525] 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."

[0526] In today's communication environment, users make a large number of calls daily, including spam and fraudulent calls. This increases the risk of users becoming involved in unintended troubles. Furthermore, the management and subsequent analysis of call information are insufficient, making it difficult to maintain a secure communication environment. To solve this problem, there is a need for technology that can quickly and accurately assess the security of calls and provide a secure communication environment for users.

[0527] 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.

[0528] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via an information processing device; means for comparing the collected information with an existing information aggregation medium and performing a risk assessment; and means for converting speech into text data using speech recognition technology and performing an additional risk assessment using the text data. This makes it possible to prevent nuisance calls and fraudulent calls, and to provide users with a safe and secure communication experience.

[0529] "Call initiation" refers to the moment when communication begins, and this is when the system automatically executes the necessary processes.

[0530] An "agent" is a virtual unit of operation responsible for information processing, and is an entity that collects and analyzes speech and text.

[0531] An "information processing device" is an electronic device that plays a role in collecting, analyzing, and evaluating information, and it forms the core of the entire system.

[0532] "Means of collecting information from the other party" refers to the processes and technologies that function to collect information that can be obtained from the person on the other end of a call.

[0533] An "information aggregation medium" is a database or recording system that stores past call data, patterns of spam calls, and other similar information.

[0534] "Means of risk assessment" refers to an evaluation process or algorithm for determining the safety of a call based on collected information.

[0535] "Speech recognition technology" is a technology that converts speech into text data, processing human voices as digital information.

[0536] "Text data" refers to information in text format converted by speech recognition, and is used for further analysis and evaluation.

[0537] To implement this invention, a server and a terminal must work together. The server automatically activates an agent when a call begins and quickly collects information from the other party via an information processing device. This collection utilizes speech recognition technology, where speech data is converted into text data in real time and transmitted to the server.

[0538] The speech recognition process utilizes the Google Speech-to-Text API, among others, and uses the smartphone's microphone as hardware. The server compares the received text data with existing data aggregation media, such as a database of past spam calls, and performs a risk assessment using an AI model. The risk assessment algorithm is built using machine learning libraries such as TensorFlow and PyTorch.

[0539] Once the evaluation results are generated, the device immediately displays the results on the information display device, i.e., the smartphone's screen. If the evaluation determines that the call is safe, the call is transferred to the user as usual. On the other hand, if the evaluation determines that the risk is high, the device displays a warning to the user, allowing them to choose a course of action.

[0540] For example, if a call is suspected of being a scam call, the system evaluates it and displays a warning message on the device such as, "This may be a spam call." In this way, the user can decide whether or not to answer the call.

[0541] An example of a prompt message to a generative AI model could be, based on information collected at the start of a call, "What is your name and purpose of this call? Please assess the risk based on the voice pattern." This would allow the server to perform a quick and accurate risk assessment.

[0542] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0543] Step 1:

[0544] The server automatically activates the agent as soon as a call begins. The input is the incoming signal, and the output confirms that the agent has been activated. This action prepares the server to begin collecting information from the other party.

[0545] Step 2:

[0546] The terminal uses a communication device to record the other party's voice and converts the voice data into text data using speech recognition technology. The input is voice data, and the output is the converted text data. This data is processed using the Google Speech-to-Text API, and the converted text data is sent to the server.

[0547] Step 3:

[0548] The server compares the received text data with a database and performs a risk assessment using an AI model. The input is text data, and the output is the result of the risk assessment. TensorFlow is used to apply the risk assessment algorithm and generate the assessment results.

[0549] Step 4:

[0550] The server sends the risk assessment results back to the terminal. The input is the risk assessment results, and the output is notification data for the user. The information is transmitted to the terminal to prepare for the next action.

[0551] Step 5:

[0552] The device notifies the user of the call's safety based on the received risk assessment results. The input is notification data, and the output is a displayed message. For low-risk calls, the call is transferred; for high-risk calls, a warning is displayed.

[0553] Step 6:

[0554] After a call ends, the user sends feedback to the server via their device. The input is user feedback, and the output is improvement data. Receiving this feedback helps improve the accuracy of the server's AI model.

[0555] Step 7:

[0556] The server uses feedback information as training data to retrain the AI ​​model, improving the accuracy of risk assessment. The input is the improved data, and the output is the updated AI model. This is then reflected in the next risk assessment.

[0557] 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.

[0558] This embodiment provides a telephone system that combines an emotion engine that recognizes the user's emotions and reflects that information in the risk assessment of the call. First, the server detects an incoming call and monitors the call content in real time using an AI agent. At this time, the terminal converts the voice data into text through speech recognition and sends it to the server.

[0559] In addition, the emotion engine detects the user's emotions through voice analysis. By analyzing the tone, tempo, and other voice characteristics, the server evaluates the user's emotional state in real time. This emotion data is used as part of the risk assessment, and if an abnormal emotional response is detected, the risk score of the call is adjusted.

[0560] As a concrete example, after the server detects an incoming call, the terminal uses AI to ask the caller, "What is your name and why are you calling?" This audio data is first fed into the emotion engine. The server analyzes the emotional nuances contained in the caller's response, and if there are significant signs of anxiety or tension, it incorporates this into the risk assessment. If the risk is deemed high, the terminal displays a warning to the user saying, "Please wait a moment. This call requires caution."

[0561] Based on this, the user decides whether to continue or end the call. After the call ends, the server records all voice and emotional data. The accumulated data is used as training material for the system and helps improve the risk assessment algorithm to ensure user safety. This makes it possible to provide users with a safer and more secure telephone environment.

[0562] The following describes the processing flow.

[0563] Step 1:

[0564] The server detects an incoming call and immediately activates the AI ​​agent. It prepares the call connection and starts a session with the other party.

[0565] Step 2:

[0566] The device utilizes voice recognition to ask the caller initial questions such as, "This is an automated response system. Please tell us your name and purpose of your call." It then converts the caller's voice data into text data and sends it to the server.

[0567] Step 3:

[0568] The device simultaneously activates an emotion engine, analyzing emotional characteristics such as tone, pitch, and voice intensity of the audio data, and generating emotion data in real time. It then sends the emotion recognition results to the server.

[0569] Step 4:

[0570] The server retrieves text data and sentiment data, which are then compared against existing databases. The sentiment data is added as a new input to risk assessment, allowing for a quantitative analysis of the user's level of calmness and anxiety.

[0571] Step 5:

[0572] The server compiles the results of the risk assessment, and if it is determined to be high risk, the terminal displays a warning message to the user saying, "This call requires caution." If it is low risk, it will inform the user that "We will transfer your call" and prepare to connect the call to the user.

[0573] Step 6:

[0574] The user reviews the warning and decides whether to continue, transfer, or end the call. Based on the warning, the user reconfirms the safety of the call.

[0575] Step 7:

[0576] After a call ends, the server saves voice and sentiment data as logs, which are then used in subsequent improvement processes. This data is used to train the system and contribute to improving the accuracy of risk assessment in future calls.

[0577] (Example 2)

[0578] 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."

[0579] In modern telephone systems, accurately assessing the risks during calls and appropriately notifying users is crucial. However, mechanisms for analyzing call content in real time, evaluating emotions, and warning users about problematic calls are not yet well-established. As a result, many users may be unable to properly handle risky calls, potentially leading to problems. In light of this situation, there is a need for a system that enables emotional assessment during calls, calculation of risk scores, and subsequent appropriate responses.

[0580] 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.

[0581] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via a communication device; means for converting the collected voice data into text using speech recognition technology and analyzing the text data for sentiment evaluation; and means for calculating a risk score based on the sentiment evaluation results and displaying a warning to the user if the risk is high. This makes it possible to accurately evaluate the risks during a call and alert the user.

[0582] An "agent" is a program that automatically starts up when a call begins, monitors the call content, and collects necessary information.

[0583] "Communication devices" refers to all equipment and devices used to send and receive audio.

[0584] "Speech recognition technology" is a technology that analyzes speech data and converts it into text.

[0585] "Text data" refers to data in character format converted using speech recognition technology.

[0586] "Emotional assessment methods" refer to processes and algorithms that analyze collected text data to identify the emotions of the caller.

[0587] A "risk score" is a numerical value or indicator that shows the potential risk of a phone call, calculated based on the results of an emotional assessment.

[0588] A "means of displaying warnings" refers to a system function that displays a message to alert the user when a high risk is detected.

[0589] A "data recording means" is a function that stores data collected during a call and uses it for subsequent analysis and system improvement.

[0590] A "generative AI model" is an artificial intelligence model that learns from collected data and is used to improve the performance of a system.

[0591] This invention relates to a system that enables real-time sentiment and risk assessment in a voice call system. In this system, the server automatically activates an agent at the start of a call and collects voice data. Specific hardware components include a microphone and communication devices. Upon receiving the voice, the terminal converts this voice data into text data using speech recognition technology (e.g., a common speech recognition API). This text data is immediately transmitted to the server.

[0592] The server analyzes the received text data using sentiment evaluation tools. This process utilizes specific APIs (e.g., general sentiment analysis APIs) to identify the caller's emotional state. Based on the resulting sentiment data, the server calculates a risk score. If a significant emotional abnormality is detected, this risk score is affected and set higher.

[0593] If the risk score is determined to be high, the device will display a warning to alert the user. Specifically, a message such as "Please wait a moment. This call requires attention." will be displayed. Based on this warning, the user will decide whether to continue or end the call.

[0594] After a call ends, the server records all audio and emotional data. This stored data is then analyzed by a generative AI model and used to improve the system. This process enhances the accuracy of the risk assessment algorithm, resulting in a safer calling environment.

[0595] As a concrete example, during a call, the server plays an automated message saying, "Please tell me your name and purpose of your call." If the caller's response indicates anxiety or tension, this is reflected in the risk score. An example of a prompt to the generating AI model would be, "What emotions do you perceive from the tone and tempo of this caller's voice? Please analyze in as much detail as possible and provide information that will help in the risk assessment."

[0596] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0597] Step 1:

[0598] The server detects an incoming call. Once the call begins, the server automatically activates an agent. The agent collects voice data from the communication device. At this point, the input is the incoming call information, and the output is the activation of the agent.

[0599] Step 2:

[0600] The terminal uses a communication device to record call audio in real time. The recorded audio data is converted into text data using speech recognition technology and sent to the server. The input is audio data, and the output of speech recognition is text data. This process utilizes a speech recognition API.

[0601] Step 3:

[0602] The server analyzes the received text data. Using sentiment evaluation tools, it assesses the user's emotions from the text data. The input is text data, and the output is data with sentiment tags attached. A sentiment analysis API is used for this analysis.

[0603] Step 4:

[0604] The server calculates a risk score based on the emotion assessment results. If an abnormal emotion is detected, the risk score is set higher. The input is emotion-tagged data, and the output is the risk score. For example, if strong anger is detected, the risk score will be higher.

[0605] Step 5:

[0606] The server displays a warning to the user on the terminal based on the calculated risk score. If the risk score is particularly high, the terminal displays the message, "Please wait a moment. This call requires caution." The input is the risk score, and the output is the warning display.

[0607] Step 6:

[0608] Based on the warning from the device, the user decides whether to continue or end the call. At this point, the input is a warning message, and the output will be either to continue or end the call, depending on the user's decision.

[0609] Step 7:

[0610] The server records all call audio data and sentiment evaluation results. This data is later analyzed using a generative AI model. The data is used to improve the accuracy of the risk assessment algorithm. The input in this process is audio and sentiment evaluation data, and the output is stored data.

[0611] (Application Example 2)

[0612] 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."

[0613] Telephone calls present a challenge in responding appropriately to the other party's emotional state. There is a particular need to quickly identify emotional states requiring special attention and ensure call safety. Conventional systems lack risk assessment using emotion recognition, resulting in insufficient quality and security of calls.

[0614] 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.

[0615] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via a communication device, means for comparing the collected information with an existing information set to perform a risk assessment, and means for analyzing voice data to detect emotional characteristics. This makes it possible to reinforce the risk assessment based on emotional data and decide how to handle the call.

[0616] An "agent" is a piece of software or process that automatically performs specific tasks, such as automatically starting up when a call begins and collecting information.

[0617] A "communication device" is an electronic device used to send and receive voice and text data, and capable of processing information in real time.

[0618] An "information collection" is a database or a collection of information that has been collected and organized in the past, and is used as a reference point for risk assessment.

[0619] "Risk assessment" is the process of analyzing collected information, calculating the risks associated with phone calls as numerical values ​​or indicators, and determining safety.

[0620] "Voice data" refers to voice information acquired during a phone call and serves as the basis for analyzing emotional characteristics.

[0621] "Emotional characteristics" are indicators of emotional state determined based on the tone, tempo, and other characteristics of the voice obtained from audio data.

[0622] "Call processing" refers to a series of procedures that determine actions such as continuing, ending, or displaying a warning based on risk assessment results and emotional characteristics.

[0623] The system for implementing the present invention combines a communication method and a signal processing algorithm. In this system, a server first detects the start of a call using a communication device and activates an agent. The agent processes the voice data collected from the other party in real time. The hardware used is assumed to be a smartphone, and the software utilizes a voice recognition service API to convert the voice data into text.

[0624] The server converts the audio data into text data using a speech recognition API. This text data is then analyzed using an emotion analysis library to extract emotional characteristics. Specifically, the tone of voice, tempo, and volume are among the elements analyzed. An example of an emotion analysis library used here is an emotion recognition engine.

[0625] Once voice analysis is complete, the server performs a risk assessment based on emotional characteristics. Here, a generative AI model is used to model the relationship between emotional characteristics and risk, enabling real-time assessment. If the risk is assessed as high, the device displays a warning message to the user. Based on this warning, the user can decide whether to continue or end the call.

[0626] As a concrete example, when a user is on a call with a client who is exhibiting unstable emotions in a work context, this system displays a warning on the smartphone screen saying, "This call requires caution." This warning allows the user to respond flexibly based on the other party's psychological state.

[0627] An example of a prompt message is: "Analyze the emotional state of the person on the other end of this call, and if factors that significantly increase the risk assessment are detected, display a warning to the user."

[0628] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0629] Step 1:

[0630] The server detects the start of a call and automatically activates the agent. At this time, it begins receiving voice data using the communication device. The input is the call start signal, and the output is the activation of the agent.

[0631] Step 2:

[0632] The terminal sends the received audio data to the speech recognition service API. The input is the audio data from the call, which the API analyzes and outputs as text data. Specifically, the process involves converting the audio signal into text.

[0633] Step 3:

[0634] The server sends text data to an emotion analysis library to extract emotional features. The input is transcribed audio data, and the output is emotional features analyzed based on tone, tempo, and volume. Specifically, emotions are identified by quantifying the features.

[0635] Step 4:

[0636] The server uses a generative AI model to assess risk based on emotional characteristics. The input is emotional characteristics, and the output is a risk score. If this risk score exceeds a certain level, the risk is judged to be high. The process involves analyzing emotional data and calculating risk using a learned model based on past data.

[0637] Step 5:

[0638] The device displays warning messages to the user based on a risk score. The input is the risk score, and the output is a warning message displayed on the screen. Specifically, a dialog box appears on the screen to alert the user when a warning is necessary.

[0639] Step 6:

[0640] The user receives a warning message and decides whether to continue or end the call. The input is the warning message displayed on the device, and the output is the user's decision to perform on the call. The action involves choosing whether to continue or disconnect the call.

[0641] 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.

[0642] 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.

[0643] 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.

[0644] [Fourth Embodiment]

[0645] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0646] 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.

[0647] 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).

[0648] 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.

[0649] 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.

[0650] 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).

[0651] 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.

[0652] 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.

[0653] 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.

[0654] 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.

[0655] 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.

[0656] 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.

[0657] 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".

[0658] This invention relates to a telephone system that automatically evaluates the risks associated with a call and takes appropriate action. In one embodiment, a server first detects an incoming call and quickly collects information from the caller using an AI agent. The terminal converts the caller's voice into text data using its voice recognition function, and the server performs a risk assessment based on that information.

[0659] During this risk assessment, the server compares the call against a database containing past cases of spam and fraud. This allows the server to evaluate the safety of the call content and determine the next course of action based on the result. Specifically, if the risk is assessed as low, the terminal will smoothly transfer the call to the user. On the other hand, if the risk is determined to be high, the terminal will display a warning to the user to alert them.

[0660] Furthermore, users can provide feedback using options offered after the call ends. The server uses this feedback to improve the AI ​​agent's performance, enabling more accurate risk assessments.

[0661] As a concrete example, when the server detects an incoming call, an agent speaks to the caller saying, "This is the telephone answering system. Please tell me your name and purpose of your call." The terminal transcribes the voice into text in real time and sends it to the server. The server assesses the risk based on this information, and if it determines that the call is secure, the terminal responds, "Please wait a moment. I will connect you to the appropriate person," and connects the call to the user. This entire process helps prevent spam calls and scams, ensuring safe communication.

[0662] The following describes the processing flow.

[0663] Step 1:

[0664] The server detects an incoming call and activates the AI ​​agent. It prepares for the call and initiates the connection with the other party.

[0665] Step 2:

[0666] The device uses voice recognition to automatically ask the person on the other end of the call initial questions such as, "Could you please tell me your name and the purpose of your call?" The collected audio is converted into text data and sent to the server.

[0667] Step 3:

[0668] The server receives the information in text format and inputs it into the risk assessment module. Here, it compares it with existing patterns of spam calls and scams stored in the database and calculates a risk score.

[0669] Step 4:

[0670] The server determines whether the risk is high or low based on the risk assessment. If it is determined to be low risk, the terminal will notify the user with "We will transfer your call" and prepare to connect the call to the user.

[0671] Step 5:

[0672] If the device is deemed high-risk, it will display a warning to the user stating, "This call may be suspicious," and prompt the user to take action. If necessary, it will either terminate the call or offer alternative solutions.

[0673] Step 6:

[0674] Users evaluate a call by selecting feedback from options provided after the call.

[0675] Step 7:

[0676] The server collects user feedback and uses it to update the database and improve the AI ​​agent. This feedback process contributes to further improving the accuracy of risk assessments.

[0677] (Example 1)

[0678] 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".

[0679] In modern society, voice calls over communication networks are commonplace, but nuisance calls and fraudulent calls remain a problem. This presents a challenge for users in ensuring safe and reliable communication. Furthermore, existing systems often fail to effectively incorporate user feedback for improvement. Therefore, a system is needed that automatically assesses call risks and responds appropriately to these problems.

[0680] 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.

[0681] In this invention, the server includes means for detecting incoming calls via a communication network and acquiring acoustic data, means for using a terminal device to convert the acquired acoustic data into text data, and means for comparing the text data with an existing data collection and performing a risk assessment. This makes it possible to quickly assess nuisance calls and fraudulent calls and realize safe and reliable communication.

[0682] A "communication network" is the infrastructure for sending and receiving data and voice through digital or analog communication systems.

[0683] "Incoming call" refers to the arrival of a voice call or message sent via a communication network to the recipient's device.

[0684] "Audio data" refers to data that records or transmits audio signals in digital or analog format.

[0685] "Character data" refers to data in which linguistic information is represented in text format, and is usually generated by character recognition technology.

[0686] A "terminal device" is a device that a user can directly operate and is used for converting voice data and establishing communication.

[0687] A "data collection" is a database that compiles past records and case information, and is used to cross-reference it with other data.

[0688] "Risk assessment" is the process of determining whether a particular event or communication poses a risk to the user, based on the information collected.

[0689] "Feedback" refers to the opinions and evaluations provided by users, and is information used to improve and adjust the system.

[0690] This invention relates to a system for automatically evaluating the risks during a call using a communication network and for achieving secure communication. Detailed embodiments thereof are described below.

[0691] The server detects incoming calls via the communication network. A communication protocol with a function to monitor incoming signals is used. Upon detecting an incoming call, the server activates an AI agent to acquire audio data from the other party. The terminal then converts this audio data into text data using speech recognition software. A general-purpose speech recognition engine is used in this process.

[0692] The text data sent from the terminal to the server is matched by the server against a data collection. This matching process queries a large database and evaluates whether it matches known spam calls or scam cases. Based on this evaluation, the server determines the risk level of the call and decides on the necessary actions.

[0693] As a concrete example, when a server receives an incoming call, the AI ​​agent can say, "This is the answering system. Please tell me your name and purpose of your call." The terminal uses speech recognition software to quickly transcribe the received audio into text and send this information to the server. The server uses this information to perform a risk assessment, and if it determines that the call is safe, it can provide the user with a message through the terminal saying, "Please wait a moment. I will connect you to the appropriate person," and smoothly transfer the call.

[0694] Users can provide feedback through an interface provided after a call. The server aggregates this feedback and uses it to improve the accuracy of the AI ​​model. This is expected to make the system's risk assessment increasingly refined over time.

[0695] An example of a prompt would be, "Please describe the process of converting voice data to text used in the telephone answering system. Also, please describe in detail the risk assessment method using historical data." This makes it easier to understand the detailed operation and processes of the system.

[0696] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0697] Step 1:

[0698] The server detects the incoming call.

[0699] Input: Incoming signal from the communication network

[0700] Specific operation: The server continuously monitors incoming signals via the communication protocol. When an incoming signal is received, the server activates the AI ​​agent.

[0701] Output: Ready to collect audio data from the other party.

[0702] Step 2:

[0703] The server uses an AI agent to acquire acoustic data.

[0704] Input: Caller's voice based on incoming call

[0705] Specific operation: The AI ​​agent speaks to the other party saying, "This is the response system. Please tell me your name and how can I help you?" and records the voice signal.

[0706] Output: Acquired acoustic data

[0707] Step 3:

[0708] The device converts the audio data into text data.

[0709] Input: Acoustic data acquired by the AI ​​agent

[0710] Specific operation: The device uses speech recognition software to convert audio data into text data. A phonological analysis algorithm is used in this conversion process.

[0711] Output: Transcripted call content as text data

[0712] Step 4:

[0713] The server uses text data to perform a risk assessment.

[0714] Input: Transcripted call content as text data

[0715] Specific operation: The server compares the text data with a data collection and performs an evaluation by applying an algorithm that matches it against known spam calls and fraud cases.

[0716] Output: Results of the call risk assessment

[0717] Step 5:

[0718] The server decides on the next action based on the evaluation results.

[0719] Input: Results of the call risk assessment

[0720] Specific actions: If the risk is low, the server instructs the terminal to transfer the call to the user, and the terminal notifies the user with "Please wait a moment. I will connect you to the appropriate person." If the risk is high, a warning message is displayed on the terminal.

[0721] Output: Call forwarding or warning notification to the user.

[0722] Step 6:

[0723] Users provide feedback after the call.

[0724] Input: User feedback on the call experience

[0725] Specific operation: Users use the provided interface to enter feedback about their call experience. This typically includes simple questions and satisfaction ratings.

[0726] Output: Feedback information sent to the server

[0727] Step 7:

[0728] The server improves the AI ​​model based on the feedback.

[0729] Input: Feedback information received from users

[0730] Specific operation: The server analyzes the feedback and adjusts the AI ​​model's algorithm to improve the accuracy of future risk assessments.

[0731] Output: More accurate risk assessment results from the improved AI agent.

[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 today's communication environment, users make a large number of calls daily, including spam and fraudulent calls. This increases the risk of users becoming involved in unintended troubles. Furthermore, the management and subsequent analysis of call information are insufficient, making it difficult to maintain a secure communication environment. To solve this problem, there is a need for technology that can quickly and accurately assess the security of calls and provide a secure communication environment for users.

[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 automatically activating an agent at the start of a call and collecting information from the other party via an information processing device; means for comparing the collected information with an existing information aggregation medium and performing a risk assessment; and means for converting speech into text data using speech recognition technology and performing an additional risk assessment using the text data. This makes it possible to prevent nuisance calls and fraudulent calls, and to provide users with a safe and secure communication experience.

[0737] "Call initiation" refers to the moment when communication begins, and this is when the system automatically executes the necessary processes.

[0738] An "agent" is a virtual unit of operation responsible for information processing, and is an entity that collects and analyzes speech and text.

[0739] An "information processing device" is an electronic device that plays a role in collecting, analyzing, and evaluating information, and it forms the core of the entire system.

[0740] "Means of collecting information from the other party" refers to the processes and technologies that function to collect information that can be obtained from the person on the other end of a call.

[0741] An "information aggregation medium" is a database or recording system that stores past call data, patterns of spam calls, and other similar information.

[0742] "Means of risk assessment" refers to an evaluation process or algorithm for determining the safety of a call based on collected information.

[0743] "Speech recognition technology" is a technology that converts speech into text data, processing human voices as digital information.

[0744] "Text data" refers to information in text format converted by speech recognition, and is used for further analysis and evaluation.

[0745] To implement this invention, a server and a terminal must work together. The server automatically activates an agent when a call begins and quickly collects information from the other party via an information processing device. This collection utilizes speech recognition technology, where speech data is converted into text data in real time and transmitted to the server.

[0746] The speech recognition process utilizes the Google Speech-to-Text API, among others, and uses the smartphone's microphone as hardware. The server compares the received text data with existing data aggregation media, such as a database of past spam calls, and performs a risk assessment using an AI model. The risk assessment algorithm is built using machine learning libraries such as TensorFlow and PyTorch.

[0747] Once the evaluation results are generated, the device immediately displays the results on the information display device, i.e., the smartphone's screen. If the evaluation determines that the call is safe, the call is transferred to the user as usual. On the other hand, if the evaluation determines that the risk is high, the device displays a warning to the user, allowing them to choose a course of action.

[0748] For example, if a call is suspected of being a scam call, the system evaluates it and displays a warning message on the device such as, "This may be a spam call." In this way, the user can decide whether or not to answer the call.

[0749] An example of a prompt message to a generative AI model could be, based on information collected at the start of a call, "What is your name and purpose of this call? Please assess the risk based on the voice pattern." This would allow the server to perform a quick and accurate risk assessment.

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

[0751] Step 1:

[0752] The server automatically activates the agent as soon as a call begins. The input is the incoming signal, and the output confirms that the agent has been activated. This action prepares the server to begin collecting information from the other party.

[0753] Step 2:

[0754] The terminal uses a communication device to record the other party's voice and converts the voice data into text data using speech recognition technology. The input is voice data, and the output is the converted text data. This data is processed using the Google Speech-to-Text API, and the converted text data is sent to the server.

[0755] Step 3:

[0756] The server compares the received text data with a database and performs a risk assessment using an AI model. The input is text data, and the output is the result of the risk assessment. TensorFlow is used to apply the risk assessment algorithm and generate the assessment results.

[0757] Step 4:

[0758] The server sends the risk assessment results back to the terminal. The input is the risk assessment results, and the output is notification data for the user. The information is transmitted to the terminal to prepare for the next action.

[0759] Step 5:

[0760] The device notifies the user of the call's safety based on the received risk assessment results. The input is notification data, and the output is a displayed message. For low-risk calls, the call is transferred; for high-risk calls, a warning is displayed.

[0761] Step 6:

[0762] After a call ends, the user sends feedback to the server via their device. The input is user feedback, and the output is improvement data. Receiving this feedback helps improve the accuracy of the server's AI model.

[0763] Step 7:

[0764] The server uses feedback information as training data to retrain the AI ​​model, improving the accuracy of risk assessment. The input is the improved data, and the output is the updated AI model. This is then reflected in the next risk assessment.

[0765] 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.

[0766] This embodiment provides a telephone system that combines an emotion engine that recognizes the user's emotions and reflects that information in the risk assessment of the call. First, the server detects an incoming call and monitors the call content in real time using an AI agent. At this time, the terminal converts the voice data into text through speech recognition and sends it to the server.

[0767] In addition, the emotion engine detects the user's emotions through voice analysis. By analyzing the tone, tempo, and other voice characteristics, the server evaluates the user's emotional state in real time. This emotion data is used as part of the risk assessment, and if an abnormal emotional response is detected, the risk score of the call is adjusted.

[0768] As a concrete example, after the server detects an incoming call, the terminal uses AI to ask the caller, "What is your name and why are you calling?" This audio data is first fed into the emotion engine. The server analyzes the emotional nuances contained in the caller's response, and if there are significant signs of anxiety or tension, it incorporates this into the risk assessment. If the risk is deemed high, the terminal displays a warning to the user saying, "Please wait a moment. This call requires caution."

[0769] Based on this, the user decides whether to continue or end the call. After the call ends, the server records all voice and emotional data. The accumulated data is used as training material for the system and helps improve the risk assessment algorithm to ensure user safety. This makes it possible to provide users with a safer and more secure telephone environment.

[0770] The following describes the processing flow.

[0771] Step 1:

[0772] The server detects an incoming call and immediately activates the AI ​​agent. It prepares the call connection and starts a session with the other party.

[0773] Step 2:

[0774] The device utilizes voice recognition to ask the caller initial questions such as, "This is an automated response system. Please tell us your name and purpose of your call." It then converts the caller's voice data into text data and sends it to the server.

[0775] Step 3:

[0776] The device simultaneously activates an emotion engine, analyzing emotional characteristics such as tone, pitch, and voice intensity of the audio data, and generating emotion data in real time. It then sends the emotion recognition results to the server.

[0777] Step 4:

[0778] The server retrieves text data and sentiment data, which are then compared against existing databases. The sentiment data is added as a new input to risk assessment, allowing for a quantitative analysis of the user's level of calmness and anxiety.

[0779] Step 5:

[0780] The server compiles the results of the risk assessment, and if it is determined to be high risk, the terminal displays a warning message to the user saying, "This call requires caution." If it is low risk, it will inform the user that "We will transfer your call" and prepare to connect the call to the user.

[0781] Step 6:

[0782] The user reviews the warning and decides whether to continue, transfer, or end the call. Based on the warning, the user reconfirms the safety of the call.

[0783] Step 7:

[0784] After a call ends, the server saves voice and sentiment data as logs, which are then used in subsequent improvement processes. This data is used to train the system and contribute to improving the accuracy of risk assessment in future calls.

[0785] (Example 2)

[0786] 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".

[0787] In modern telephone systems, accurately assessing the risks during calls and appropriately notifying users is crucial. However, mechanisms for analyzing call content in real time, evaluating emotions, and warning users about problematic calls are not yet well-established. As a result, many users may be unable to properly handle risky calls, potentially leading to problems. In light of this situation, there is a need for a system that enables emotional assessment during calls, calculation of risk scores, and subsequent appropriate responses.

[0788] 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.

[0789] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via a communication device; means for converting the collected voice data into text using speech recognition technology and analyzing the text data for sentiment evaluation; and means for calculating a risk score based on the sentiment evaluation results and displaying a warning to the user if the risk is high. This makes it possible to accurately evaluate the risks during a call and alert the user.

[0790] An "agent" is a program that automatically starts up when a call begins, monitors the call content, and collects necessary information.

[0791] "Communication devices" refers to all equipment and devices used to send and receive audio.

[0792] "Speech recognition technology" is a technology that analyzes speech data and converts it into text.

[0793] "Text data" refers to data in character format converted using speech recognition technology.

[0794] "Emotional assessment methods" refer to processes and algorithms that analyze collected text data to identify the emotions of the caller.

[0795] A "risk score" is a numerical value or indicator that shows the potential risk of a phone call, calculated based on the results of an emotional assessment.

[0796] A "means of displaying warnings" refers to a system function that displays a message to alert the user when a high risk is detected.

[0797] A "data recording means" is a function that stores data collected during a call and uses it for subsequent analysis and system improvement.

[0798] A "generative AI model" is an artificial intelligence model that learns from collected data and is used to improve the performance of a system.

[0799] This invention relates to a system that enables real-time sentiment and risk assessment in a voice call system. In this system, the server automatically activates an agent at the start of a call and collects voice data. Specific hardware components include a microphone and communication devices. Upon receiving the voice, the terminal converts this voice data into text data using speech recognition technology (e.g., a common speech recognition API). This text data is immediately transmitted to the server.

[0800] The server analyzes the received text data using sentiment evaluation tools. This process utilizes specific APIs (e.g., general sentiment analysis APIs) to identify the caller's emotional state. Based on the resulting sentiment data, the server calculates a risk score. If a significant emotional abnormality is detected, this risk score is affected and set higher.

[0801] If the risk score is determined to be high, the device will display a warning to alert the user. Specifically, a message such as "Please wait a moment. This call requires attention." will be displayed. Based on this warning, the user will decide whether to continue or end the call.

[0802] After a call ends, the server records all audio and emotional data. This stored data is then analyzed by a generative AI model and used to improve the system. This process enhances the accuracy of the risk assessment algorithm, resulting in a safer calling environment.

[0803] As a concrete example, during a call, the server plays an automated message saying, "Please tell me your name and purpose of your call." If the caller's response indicates anxiety or tension, this is reflected in the risk score. An example of a prompt to the generating AI model would be, "What emotions do you perceive from the tone and tempo of this caller's voice? Please analyze in as much detail as possible and provide information that will help in the risk assessment."

[0804] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0805] Step 1:

[0806] The server detects an incoming call. Once the call begins, the server automatically activates an agent. The agent collects voice data from the communication device. At this point, the input is the incoming call information, and the output is the activation of the agent.

[0807] Step 2:

[0808] The terminal uses a communication device to record call audio in real time. The recorded audio data is converted into text data using speech recognition technology and sent to the server. The input is audio data, and the output of speech recognition is text data. This process utilizes a speech recognition API.

[0809] Step 3:

[0810] The server analyzes the received text data. Using sentiment evaluation tools, it assesses the user's emotions from the text data. The input is text data, and the output is data with sentiment tags attached. A sentiment analysis API is used for this analysis.

[0811] Step 4:

[0812] The server calculates a risk score based on the emotion assessment results. If an abnormal emotion is detected, the risk score is set higher. The input is emotion-tagged data, and the output is the risk score. For example, if strong anger is detected, the risk score will be higher.

[0813] Step 5:

[0814] The server displays a warning to the user on the terminal based on the calculated risk score. If the risk score is particularly high, the terminal displays the message, "Please wait a moment. This call requires caution." The input is the risk score, and the output is the warning display.

[0815] Step 6:

[0816] Based on the warning from the device, the user decides whether to continue or end the call. At this point, the input is a warning message, and the output will be either to continue or end the call, depending on the user's decision.

[0817] Step 7:

[0818] The server records all call audio data and sentiment evaluation results. This data is later analyzed using a generative AI model. The data is used to improve the accuracy of the risk assessment algorithm. The input in this process is audio and sentiment evaluation data, and the output is stored data.

[0819] (Application Example 2)

[0820] 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".

[0821] Telephone calls present a challenge in responding appropriately to the other party's emotional state. There is a particular need to quickly identify emotional states requiring special attention and ensure call safety. Conventional systems lack risk assessment using emotion recognition, resulting in insufficient quality and security of calls.

[0822] 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.

[0823] In this invention, the server includes means for automatically activating an agent at the start of a call and collecting information from the other party via a communication device, means for comparing the collected information with an existing information set to perform a risk assessment, and means for analyzing voice data to detect emotional characteristics. This makes it possible to reinforce the risk assessment based on emotional data and decide how to handle the call.

[0824] An "agent" is a piece of software or process that automatically performs specific tasks, such as automatically starting up when a call begins and collecting information.

[0825] A "communication device" is an electronic device used to send and receive voice and text data, and capable of processing information in real time.

[0826] An "information collection" is a database or a collection of information that has been collected and organized in the past, and is used as a reference point for risk assessment.

[0827] "Risk assessment" is the process of analyzing collected information, calculating the risks associated with phone calls as numerical values ​​or indicators, and determining safety.

[0828] "Voice data" refers to voice information acquired during a phone call and serves as the basis for analyzing emotional characteristics.

[0829] "Emotional characteristics" are indicators of emotional state determined based on the tone, tempo, and other characteristics of the voice obtained from audio data.

[0830] "Call processing" refers to a series of procedures that determine actions such as continuing, ending, or displaying a warning based on risk assessment results and emotional characteristics.

[0831] The system for implementing the present invention combines a communication method and a signal processing algorithm. In this system, a server first detects the start of a call using a communication device and activates an agent. The agent processes the voice data collected from the other party in real time. The hardware used is assumed to be a smartphone, and the software utilizes a voice recognition service API to convert the voice data into text.

[0832] The server converts the audio data into text data using a speech recognition API. This text data is then analyzed using an emotion analysis library to extract emotional characteristics. Specifically, the tone of voice, tempo, and volume are among the elements analyzed. An example of an emotion analysis library used here is an emotion recognition engine.

[0833] Once voice analysis is complete, the server performs a risk assessment based on emotional characteristics. Here, a generative AI model is used to model the relationship between emotional characteristics and risk, enabling real-time assessment. If the risk is assessed as high, the device displays a warning message to the user. Based on this warning, the user can decide whether to continue or end the call.

[0834] As a concrete example, when a user is on a call with a client who is exhibiting unstable emotions in a work context, this system displays a warning on the smartphone screen saying, "This call requires caution." This warning allows the user to respond flexibly based on the other party's psychological state.

[0835] An example of a prompt message is: "Analyze the emotional state of the person on the other end of this call, and if factors that significantly increase the risk assessment are detected, display a warning to the user."

[0836] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0837] Step 1:

[0838] The server detects the start of a call and automatically activates the agent. At this time, it begins receiving voice data using the communication device. The input is the call start signal, and the output is the activation of the agent.

[0839] Step 2:

[0840] The terminal sends the received audio data to the speech recognition service API. The input is the audio data from the call, which the API analyzes and outputs as text data. Specifically, the process involves converting the audio signal into text.

[0841] Step 3:

[0842] The server sends text data to an emotion analysis library to extract emotional features. The input is transcribed audio data, and the output is emotional features analyzed based on tone, tempo, and volume. Specifically, emotions are identified by quantifying the features.

[0843] Step 4:

[0844] The server uses a generative AI model to assess risk based on emotional characteristics. The input is emotional characteristics, and the output is a risk score. If this risk score exceeds a certain level, the risk is judged to be high. The process involves analyzing emotional data and calculating risk using a learned model based on past data.

[0845] Step 5:

[0846] The device displays warning messages to the user based on a risk score. The input is the risk score, and the output is a warning message displayed on the screen. Specifically, a dialog box appears on the screen to alert the user when a warning is necessary.

[0847] Step 6:

[0848] The user receives a warning message and decides whether to continue or end the call. The input is the warning message displayed on the device, and the output is the user's decision to perform on the call. The action involves choosing whether to continue or disconnect the call.

[0849] 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.

[0850] 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.

[0851] 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.

[0852] 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.

[0853] 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.

[0854] 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.

[0855] 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.

[0856] 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.

[0857] 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."

[0858] 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.

[0859] 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.

[0860] 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.

[0861] 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.

[0862] 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.

[0863] 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.

[0864] 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.

[0865] 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.

[0866] 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.

[0867] 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.

[0868] 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.

[0869] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[0870] The following is further disclosed regarding the embodiments described above.

[0871] (Claim 1)

[0872] A means of automatically activating an agent when a call starts and collecting information from the other party via the communication device,

[0873] A means of performing risk assessment by comparing the collected information with existing databases,

[0874] A means to determine how to process a call based on the risk assessment results, and to transfer the call to the user or display a warning,

[0875] A system that includes this.

[0876] (Claim 2)

[0877] The system according to claim 1, comprising means for requesting additional authentication information based on collected information and enhancing the verification process.

[0878] (Claim 3)

[0879] The system according to claim 1, comprising data recording means for storing data collected during a call and using it for subsequent analysis and improvement.

[0880] "Example 1"

[0881] (Claim 1)

[0882] A means for detecting incoming calls via a communication network and acquiring acoustic data,

[0883] A means of using a terminal device that converts acquired audio data into text data,

[0884] A means of comparing text data with existing data collections and performing a risk assessment,

[0885] Based on the risk assessment results, the communication procedure will be determined, and a means will be established to either hand over the communication to the user or display a warning.

[0886] A means of collecting feedback from users and improving the evaluation model,

[0887] A communication system that includes this.

[0888] (Claim 2)

[0889] The communication system according to claim 1, comprising means for requesting additional verification information based on acquired information and thereby strengthening the authentication procedure.

[0890] (Claim 3)

[0891] The communication system according to claim 1, comprising information recording means for storing data acquired during a call and using it for subsequent analysis and improvement.

[0892] "Application Example 1"

[0893] (Claim 1)

[0894] A means of automatically activating an agent at the start of a call and collecting information from the other party via an information processing device,

[0895] A means of conducting a risk assessment by comparing the collected information with existing information aggregation media,

[0896] A means of determining how to process a call based on the risk assessment results, and either transferring the call to the user or displaying a warning,

[0897] A means of converting speech into text data using speech recognition technology and performing additional risk assessment using that text data,

[0898] A means for immediately notifying the evaluation results on an information display device,

[0899] A means of collecting feedback from users after a call ends and using that feedback to improve the system,

[0900] A system that includes this.

[0901] (Claim 2)

[0902] The system according to claim 1, comprising means for requesting additional verification information based on the collected information and for enhancing the verification process.

[0903] (Claim 3)

[0904] The system according to claim 1, comprising information recording means for storing data collected during a call and using it for subsequent analysis and improvement.

[0905] "Example 2 of combining an emotion engine"

[0906] (Claim 1)

[0907] A means of automatically activating an agent when a call starts and collecting information from the other party via the communication device,

[0908] A means for evaluating emotions that converts collected audio data into text using speech recognition technology and analyzes the text data,

[0909] A means of calculating a risk score based on the sentiment evaluation results and displaying a warning to the user if the risk is high,

[0910] A means for storing sentiment evaluation data and voice data of phone calls and using them for subsequent analysis and system improvement,

[0911] A system that includes this.

[0912] (Claim 2)

[0913] The system according to claim 1, comprising means for requesting additional authentication information based on collected voice data and enhancing the verification process.

[0914] (Claim 3)

[0915] The system according to claim 1, comprising means for analyzing stored data using a generative AI model and contributing to the improvement of a risk assessment algorithm.

[0916] "Application example 2 when combining with an emotional engine"

[0917] (Claim 1)

[0918] A means of automatically activating an agent when a call starts and collecting information from the other party via a communication device,

[0919] A means of performing risk assessment by comparing collected information with existing information sets,

[0920] A means of analyzing audio data to detect emotional characteristics,

[0921] A means of reinforcing risk assessment based on emotional characteristics, deciding how to handle a call based on the risk assessment results, and transferring the call to the user or displaying a warning,

[0922] A system that includes this.

[0923] (Claim 2)

[0924] The system according to claim 1, comprising means for requesting additional approval information based on the collected information and for enhancing the verification process.

[0925] (Claim 3)

[0926] The system according to claim 1, comprising information recording means for storing information collected during a call and using it for subsequent analysis and improvement. [Explanation of symbols]

[0927] 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 automatically activating an agent when a call starts and collecting information from the other party via the communication device, A means of performing risk assessment by comparing the collected information with existing databases, A means to determine how to process a call based on the risk assessment results, and to transfer the call to the user or display a warning, A system that includes this.

2. The system according to claim 1, comprising means for requesting additional authentication information based on collected information and enhancing the verification process.

3. The system according to claim 1, further comprising data recording means for storing data collected during a call and using it for subsequent analysis and improvement.

Images