system

Abstract

We provide the system. [Solution] Means for collecting message patterns related to fraudulent activity, A means for training a generative model that learns the characteristics of fraudulent activity using the aforementioned message patterns, A method for generating pattern files based on fraudulent activity characteristics obtained from a trained generative model, A means for detecting the possibility of fraud by comparing the received message with the pattern file, A means of notifying users of detected fraudulent messages, A system that includes this.

Description

Technical Field

[0004] , , , ,

[0005] , , , , ,

[0003] , ,

[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, the method 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 in response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In recent years, messages related to fraud have become sophisticated, making it difficult for ordinary users to identify them. Such fraud involves significant risks of personal information leakage and financial damage, and occurs particularly frequently through emails and messaging applications. Therefore, effective detection and prompt countermeasures against these threats are required. Conventional systems have the problem that it takes time to detect new fraud patterns and it is difficult to respond in real time.

Means for Solving the Problems

[0005] This invention is characterized by collecting message patterns related to fraudulent activity and training a generative model that learns the characteristics of fraudulent activity using these patterns. A pattern file is generated based on the characteristics obtained from this generative model, and the possibility of fraud is detected in real time by comparing received messages with this pattern file. This system protects users from fraudulent activity by notifying them of detected fraudulent messages and, in some cases, blocking the reception of messages.

[0006] "Fraudulent activity" refers to acts carried out primarily with dishonest or malicious intent, aimed at deceiving others to illegally obtain money or information.

[0007] A "message pattern" refers to a combination of characteristic structures, keywords, phrases, and sender information within a particular message.

[0008] A "generative model" is a type of AI algorithm used to learn features from large amounts of data and generate new data or patterns based on those features.

[0009] A "pattern file" is a data file created based on the characteristics of fraudulent activity obtained from a generative model, and is used to detect fraudulent messages.

[0010] "Matching" refers to the process of comparing the information in a received message with an existing pattern file to find matching parts.

[0011] "Real-time" refers to a state where a process is performed for a user or system almost instantaneously, with virtually no delay.

[0012] A "warning" refers to a notification or message sent to a user to inform them of a potential danger or problem.

[0013] "Blocking" refers to the means or measures taken to deny access to malicious messages or suspicious content. [Brief explanation of the drawing]

[0014] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, which incorporates an emotion engine. [Figure 14]It is a sequence diagram showing the processing flow of a data processing system in Application Example 2 when a sentiment engine is combined.

Embodiments for Carrying Out the Invention

[0015] 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.

[0016] First, the terms used in the following description will be explained.

[0017] In the following embodiments, a 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.

[0018] In the following embodiments, a numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0019] In the following embodiments, a 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, etc.

[0020] 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).

[0021] 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."

[0022] [First Embodiment]

[0023] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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".

[0035] This invention relates to an embodiment of an advanced detection system for automatically detecting messages related to fraudulent activity and protecting users. The specific implementation details are described below.

[0036] First, the server retrieves data on past fraudulent activities collected from around the world. This includes existing databases and samples of newly reported fraudulent messages. Based on the collected data, the server analyzes the characteristics of these messages and trains a generative model. This training learns patterns specific to fraudulent messages and generates similar new messages.

[0037] Next, the server analyzes the patterns in the generated messages to create a new pattern file. This pattern file contains characteristic keywords and phrases common to fraudulent messages and is used for comparison with the received messages.

[0038] When a user receives a new message, it is automatically sent from their device to the server. The server scans the received message in real time and compares it against a pattern file. This comparison allows the server to determine if the message is potentially fraudulent. For example, if the subject line of a message contains phrases such as "Urgent action required," it may be judged to be similar to past fraud patterns.

[0039] If the server detects a message as fraudulent, it will notify the user of the warning. The warning will include instructions that the message may be dangerous and that the user should not view the details. The server can also block specific messages to prevent the reception of malicious messages.

[0040] In this way, by identifying fraudulent activity and responding quickly, a system is created that protects users from fraud and provides safe access to content.

[0041] The following describes the processing flow.

[0042] Step 1:

[0043] The servers collect data related to past fraudulent activities from sources around the world. This data includes known phishing emails, fake investment offers, and more.

[0044] Step 2:

[0045] The server analyzes the collected data and extracts key features. This process includes data cleaning, deduplication, and identification of relevant keywords and phrases.

[0046] Step 3:

[0047] The server uses the extracted features to train a generative AI model. The AI ​​model learns the characteristics of fraudulent messages and is able to generate similar messages.

[0048] Step 4:

[0049] The server analyzes the output of the trained AI model and generates a pattern file containing the characteristics of fraudulent messages. This pattern file is used to detect fraudulent activity.

[0050] Step 5:

[0051] When a user receives a message, the device sends that message to the server for scanning.

[0052] Step 6:

[0053] The server compares incoming messages against a pattern file in real time. This comparison helps determine whether the message is potentially fraudulent.

[0054] Step 7:

[0055] If a message is identified as fraudulent, the server sends details to the device and warns the user of the danger. The warning includes cautionary information.

[0056] Step 8:

[0057] If necessary, the server will block the reception of messages suspected of being fraudulent. This process prevents such messages from reaching the user.

[0058] (Example 1)

[0059] 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."

[0060] Fraudulent messages are becoming more sophisticated year after year, making them difficult to detect with conventional technologies and increasing the risk of users becoming victims. To address this problem, an effective means is needed to analyze messages in real time, quickly and accurately detect potential fraud, and warn users.

[0061] 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.

[0062] In this invention, the server includes means for collecting information, means for training a program that learns the characteristics of fraudulent activities using the information, and means for generating pattern information based on the characteristics of fraudulent activities obtained from the trained program. This makes it possible to analyze the received data in real time, detect the possibility of fraud with high accuracy and speed, and warn the user.

[0063] "Information" refers to any data related to past fraudulent activities collected from databases and online sources.

[0064] A "program" is a system that uses information to learn the characteristics of fraudulent activities and makes inferences and predictions based on that learning.

[0065] "Training" is the process by which a program uses information to learn fraud patterns and improve its ability to identify the characteristics of fraudulent activities.

[0066] "Pattern information" refers to a dataset containing characteristic keywords, phrases, and related structures common to fraudulent activities.

[0067] "Data" refers to any messages or pieces of information that a system receives and analyzes.

[0068] "Analysis" is the process of evaluating received data and determining whether it contains characteristics of fraudulent activity.

[0069] "User" refers to an individual or organization that is protected from fraudulent activity by using the system.

[0070] This invention relates to an advanced detection system for automatically detecting messages related to fraudulent activity and protecting users. The system is implemented through server, terminal, and user components.

[0071] The server collects data related to past fraudulent activities from existing databases and online resources to gather information. This data is used to train a generative AI model for the server to learn. This process uses natural language processing techniques and machine learning algorithms, and Python libraries (e.g., scikit-learn, TENSORFLOW®) are used for data cleaning and feature engineering.

[0072] The generative AI model learns patterns specific to fraudulent activities and, after training, has the ability to generate pattern information based on the characteristics of fraudulent messages. This pattern information includes characteristic keywords and phrases common to fraudulent messages, which are used to compare with received data. The server has the capability to scan incoming messages in real time based on this pattern information.

[0073] The terminal automatically sends messages received by the user to the server. This is done through dedicated client software and is performed every time the user receives a new message. The server analyzes the received messages to check for potential fraud. If the message is identified as fraudulent, the server sends a warning to the user to protect them from further harm.

[0074] For example, if a user receives a message saying, "Please update your account information," the server will identify this as a phishing message and issue a warning immediately. Examples of prompts used in the system's AI model include:

[0075] "Please generate email subject lines that are easy to deceive."

[0076] In this way, users are protected from fraudulent activities and can safely use the content.

[0077] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0078] Step 1:

[0079] The server collects information. The input is data on fraudulent activities obtained from existing databases and online resources. The data includes examples of phishing emails and spam messages. The server collects this data in a secure manner and stores it in a database. This process provides foundational data for pattern analysis of fraudulent messages.

[0080] Step 2:

[0081] The server trains a generative AI model using the collected data. The input is the dataset collected in step 1, and the output is the trained generative AI model. The server utilizes natural language processing techniques and machine learning algorithms, preprocessing the data using Python libraries (e.g., scikit-learn, TensorFlow) to train the AI ​​model. This generates a model that has learned the features of fraudulent messages.

[0082] Step 3:

[0083] The server generates pattern information from the trained model. The input is the AI ​​model from step 2, and the output is pattern information summarizing the characteristics of fraudulent activity. The server analyzes the features learned by the AI ​​model, extracts keywords and phrases common to fraudulent messages, and creates a pattern file. This pattern information will be used for future message evaluation.

[0084] Step 4:

[0085] When a user receives a new message, the terminal sends that message to the server. The input is the message received by the user, and the output is the message data sent to the server. The terminal uses dedicated client software, and the message is automatically forwarded to the server according to the settings of the email client, etc.

[0086] Step 5:

[0087] The server analyzes received messages in real time and compares them with the pattern information generated in step 3. The input is the message sent by the user and the pattern information, and the output is the result of identifying potentially fraudulent messages. The server determines whether keywords and phrases in the message match the pattern information and identifies them if they are likely to be fraudulent.

[0088] Step 6:

[0089] If the server detects a fraudulent message, it notifies the user of the warning. The input is the fraudulent message identified in step 5, and the output is the warning message to the user. The server strengthens security by warning the user not to view details about potentially dangerous messages. Specifically, a notification is sent to the user's device, and measures are taken to prevent fraudulent activity.

[0090] (Application Example 1)

[0091] 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."

[0092] In today's information society, the damage caused by fraudulent information and fraudulent activities is increasing, making it a crucial issue to ensure user information security. However, conventional methods make it difficult to quickly and accurately detect fraudulent activities and effectively protect users. Therefore, there is a need for technology that can automatically detect fraudulent information in real time and respond quickly.

[0093] 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.

[0094] In this invention, the server includes means for collecting information patterns related to fraudulent activity, means for training a generative model that learns the characteristics of fraudulent activity, and means for generating specific pattern files. This enables immediate detection of fraudulent information and notification of warnings to users.

[0095] "Fraudulent activity" refers to unethical actions taken with the intent of deceiving others to unjustly obtain money or information.

[0096] An "information pattern" is a structure or form that exhibits common characteristics or tendencies in a particular type of information.

[0097] A "generative model" is a machine learning model that learns specific patterns or features from data and is used to predict or generate new data.

[0098] A "specific pattern file" is a data file used for detecting fraudulent activity, constructed based on the characteristics of fraudulent behavior learned by a generative model.

[0099] "Real-time" means responding to or processing an event immediately at the moment it occurs.

[0100] "Notification" is the process of informing recipients of changes or events in information.

[0101] "Restricting" means inhibiting certain actions or access to prevent the influx or use of undesirable information.

[0102] The system for implementing this invention primarily revolves around interaction between a server, a terminal, and the user. The server first collects information patterns related to fraudulent activity from around the world. This includes using existing databases and collecting samples of newly reported fraud cases. The server uses the collected information patterns to train a generative AI model and learn the characteristics of fraudulent activity. Once training is complete, the server generates specific pattern files based on the characteristics of fraudulent activity. This builds the foundational data for detecting fraudulent activity.

[0103] The terminal is responsible for forwarding messages received by the user to the server in real time. The received messages are immediately compared against a specific pattern file on the server. This comparison checks, for example, whether the message subject contains a phrase such as "urgent action required." If a pattern match is found, the server detects the message as malicious information.

[0104] Users will receive a warning via their device regarding malicious information detected by the server. The warning will include information about the potential danger of the message and a request not to view the details. The server will also take additional safety measures, such as restricting the reception of such messages, as needed.

[0105] For example, if a user receives a message from a travel booking service stating, "There is a problem with your account. Immediate action is required," the server will immediately identify this message as fraudulent information and issue an appropriate warning to the user.

[0106] An example of a prompt message would be, "I would like to learn about the common characteristics of scam messages. Please tell me about past patterns in scam emails."

[0107] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0108] Step 1:

[0109] The server collects information patterns related to fraudulent activities from a database. This includes past fraud messages and reported cases. By feeding relevant data as input and extracting information patterns, it obtains a dataset for analysis as output.

[0110] Step 2:

[0111] The server trains a generative AI model using the collected dataset. By supplying information patterns as input and allowing the generative AI model to learn the characteristics of fraudulent behavior, the server obtains a trained model as output.

[0112] Step 3:

[0113] The server generates specific pattern files using a trained generative AI model. It uses the trained model as input and analyzes the characteristics of fraudulent activity to create identification pattern files as output.

[0114] Step 4:

[0115] The terminal sends new messages received by the user to the server. By supplying received messages as input and performing a real-time scan on the server, the output becomes the result of message analysis.

[0116] Step 5:

[0117] The server compares received messages against a specific pattern file to determine if they are potentially fraudulent. Using message data and a pattern file as input, the server performs a data matching process, and the output is the result of the fraud detection.

[0118] Step 6:

[0119] If the server determines that the information is fraudulent, it will notify the user via the terminal. The input is the result of the fraud detection, and the server generates a warning message and sends it to the user, resulting in a warning notification as output.

[0120] Step 7:

[0121] The server, if it determines a message is fraudulent, will restrict its reception. By using the fraud detection result as input and blocking messages, it enhances security as output.

[0122] 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.

[0123] This invention relates to a system that automatically detects messages related to fraudulent activity and more effectively protects users by recognizing their emotions. Specific embodiments are described below.

[0124] First, the server collects data on past fraudulent messages and analyzes their characteristics. Based on this analysis, it trains an AI model to generate patterns related to fraudulent activity from these messages. This enhances the ability to detect potential fraud in unknown messages in real time.

[0125] When a user receives a message, the device sends it to a server for fraud pattern matching. Simultaneously, the device is equipped with an emotion engine that tracks the user's reactions. The emotion engine analyzes the user's facial expressions and tone of voice while they are reading the message to infer their emotional state.

[0126] For example, if a user receives a message prompting them to click a suspicious link and displays a wary or anxious expression, the emotion engine sends that information to the server. The server analyzes the user's emotional response to the potentially fraudulent message and generates a personalized warning message, which is then sent to the device. This warning message provides practical safety advice tailored to the user's response.

[0127] Furthermore, the server records user sentiment data and uses it to improve the fraud detection model. This allows the system to learn appropriate responses for individual users and improve its accuracy.

[0128] Thus, in addition to the rapid identification and countermeasures against fraudulent activities, the present invention can further enhance user safety by providing personalized services that take user emotions into consideration.

[0129] The following describes the processing flow.

[0130] Step 1:

[0131] The server collects historical message data related to fraudulent activity from a wide range of data sources. This data includes fraudulent emails, phishing messages, and more.

[0132] Step 2:

[0133] The server analyzes the collected data to extract common patterns in fraudulent messages. This includes analyzing characteristic phrases and sender information.

[0134] Step 3:

[0135] The server trains a generative AI model based on the extracted patterns and generates pattern files to detect fraudulent activity.

[0136] Step 4:

[0137] When a user receives a message, the device scans it and sends it to the server in real time.

[0138] Step 5:

[0139] The server compares the received message against a pattern file to assess the likelihood of fraud. Based on this assessment, it determines whether or not the message is fraudulent.

[0140] Step 6:

[0141] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine analyzes the user's facial expressions and voice when they check a message to infer their emotions.

[0142] Step 7:

[0143] If the emotion engine detects any anxiety or concerns from the user, that information is sent to the server and analyzed along with the fraud detection results.

[0144] Step 8:

[0145] The server generates and sends personalized warning messages to the terminal based on the user's emotions. These messages include specific safety advice.

[0146] Step 9:

[0147] The server records user sentiment data and uses it to improve the accuracy of the model. This continuous feedback improves the overall responsiveness of the system.

[0148] (Example 2)

[0149] 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".

[0150] In recent years, online fraud has increased, causing significant losses for individuals and organizations. Traditional methods are insufficient to effectively detect and prevent fraud in real time, and they fail to take into account user emotions and reactions. It is necessary to address this situation and provide a more secure communication environment.

[0151] 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.

[0152] In this invention, the server includes means for collecting information patterns related to fraudulent activities, means for training a generative model that learns the characteristics of fraudulent activities using the information patterns, and means for recording sentiment data and improving the detection model. This enables the detection of potential fraud in real time and individual responses that take into account the user's emotions.

[0153] "Fraud" refers to any malicious act or attempt to deceive another person and illegally obtain money or assets.

[0154] "Information patterns" refer to a set of patterns or forms that indicate characteristic data or behavioral tendencies associated with fraudulent activity.

[0155] A "generative model" refers to an artificial intelligence model used to learn specific features or patterns from data and then generate or detect those features based on new data.

[0156] An "identification file" is a data file containing characteristic patterns of fraudulent activity, providing a standard used for matching with received information.

[0157] "Emotional state" refers to the psychological or emotional reactions or attitudes that a user exhibits when acquiring information.

[0158] "Emotional data" refers to data that quantifies or digitizes information about a user's emotional state and is used for system learning and improvement.

[0159] This invention relates to a system that automatically detects messages related to fraudulent activity and provides protection while also considering the user's emotions. Specific embodiments thereof are described below.

[0160] The server first collects information patterns related to past fraudulent activities from the internet and other databases. Based on this, the server trains a generative AI model. The generative AI model uses the collected data to learn the characteristics of fraudulent messages and generates identification files to identify the likelihood of fraud in new messages.

[0161] When a user receives a message, the device immediately sends it to the server. The server compares the message against the aforementioned identification file and assesses the likelihood of fraud in real time. Simultaneously, the device is equipped with an emotion engine that analyzes the user's facial expressions and tone of voice in response to the message to infer the user's emotional state.

[0162] For example, if a user receives a message prompting them to click a suspicious link and displays a surprised or anxious expression, the emotion engine sends that data to the server. Based on this, the server analyzes the user's response to the potentially fraudulent message, generates an appropriate warning message, and sends it back to the device. This warning is customized to the user's current emotions and includes practical safety advice.

[0163] Furthermore, the server records the acquired user sentiment data and uses it to improve the generative AI model. This allows the system to provide more personalized responses over time, increasing the accuracy of user protection.

[0164] As a concrete example, a possible prompt message might be: "While online shopping, you received a message offering a high-priced item at a very low price. How should you handle this message?" This system allows users to communicate more safely and securely.

[0165] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0166] Step 1:

[0167] The server collects information patterns related to fraudulent activity. This process gathers information from multiple sources, including databases of fraudulent emails and unauthorized access history. Using a dataset of past fraudulent activities as input, the server outputs information patterns that characterize fraud.

[0168] Step 2:

[0169] The server trains a generative AI model using the acquired information patterns. The collected information patterns are used as input, and machine learning algorithms are applied to learn these patterns. The output is a trained generative AI model for identifying fraudulent activity. This model improves the ability to identify potential fraud even in new messages.

[0170] Step 3:

[0171] When a user receives a message, the terminal sends that message to the server. The input is the newly received message, which is then sent to the server. The output is the message, which is then used in the next matching step.

[0172] Step 4:

[0173] The server compares received messages against an identification file. The input is a message received from a user, which is compared to an already constructed identification file. A generative AI model is used to assess the likelihood of fraud, resulting in a judgment regarding the message's fraud risk as output.

[0174] Step 5:

[0175] The device uses an emotion engine to analyze the user's emotional state. As input, the emotion engine collects the user's facial expressions and tone of voice as they read messages. As output, the user's emotional state is estimated and sent to the server.

[0176] Step 6:

[0177] The server generates personalized warning messages based on fraud risk and emotional state. The server receives emotional data as input and combines it with fraud risk data to construct specific warning messages. The output is customized advice sent to the user.

[0178] Step 7:

[0179] The server records user sentiment data and feedback on warnings to improve the generative AI model. It receives personalized warning message responses and sentiment data as input, and uses this data to retrain and improve the model. As output, it provides a generative AI model with improved system adaptability and accuracy.

[0180] (Application Example 2)

[0181] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0182] Internet-based fraud is becoming increasingly sophisticated, rendering conventional message filtering and warning systems insufficient. Furthermore, issuing uniform warnings without considering the user's emotions or circumstances can lead to false alarms and user stress. This invention aims to solve these problems by providing a personalized warning system that not only detects fraudulent messages quickly and effectively but also takes into account the user's emotional response.

[0183] 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.

[0184] In this invention, the server includes means for collecting message patterns related to fraudulent activity; means for training a generative model that learns the characteristics of fraudulent activity using the message patterns; means for generating a pattern file based on the characteristics of fraudulent activity obtained from the trained generative model; means for detecting the possibility of fraud by comparing the received message with the pattern file; means for inferring the emotional state by analyzing the user's facial expressions and voice when receiving a fraudulent message; and means for generating and notifying the user of a personalized warning message based on the user's emotional state. This makes it possible to detect unknown fraudulent messages and to provide appropriate warnings that take the user's emotions into consideration.

[0185] A "message pattern associated with fraudulent activity" is a form of information that has a common structure and characteristics and is used to identify communication content that contains fraudulent activity.

[0186] "Training a generative model" is the process of improving a machine learning algorithm using message data related to fraudulent activity to enhance its ability to identify new fraudulent messages.

[0187] "Generating pattern files" is the process of aggregating fraud features extracted from trained models and establishing criteria for detecting specific fraudulent activities based on those features.

[0188] "Detecting potential fraud" means analyzing the received message and determining whether it contains elements that suggest fraudulent activity.

[0189] "Analyzing the user's facial expressions and voice to infer their emotional state" means analyzing the user's facial expressions and tone of voice when they receive a message to estimate their emotions and psychological state at that time.

[0190] "Generating and notifying users of personalized warning messages" means customizing the content of warnings against fraudulent messages according to the user's emotional state and informing the user of that content.

[0191] To implement the invention, it is necessary to build a system that combines fraudulent message detection with user sentiment analysis. This system mainly consists of two main components: a server and a terminal.

[0192] The server collects message patterns related to fraudulent activity and trains a generative model based on historical data. Specifically, it develops a generative AI model using TensorFlow and learns the characteristics of fraudulent activity. Furthermore, it generates pattern files using the obtained characteristics and establishes criteria for detecting potential fraud. At this time, prompt statements are utilized to efficiently train the AI ​​model.

[0193] The device sends messages received by the user to a server in real time to check for potential fraud. The device also features an emotion engine utilizing the Emotion SDK and OpenCV, which analyzes the user's facial expressions and voice to infer their emotional state. This emotional information is used to generate personalized warning messages.

[0194] For example, if a user receives a message stating, "Your account will be closed if you don't act now," and shows a worried expression, the device sends this information to the server. The server flags it as a scam message, generates a personalized alert based on the facial expression data, and sends this to the device.

[0195] Examples of specific prompt messages used to achieve this process include the following:

[0196] "Message received: 'Your bank account is scheduled to be frozen. Please click this link to confirm.' User reaction: Frowning, in a worried voice."

[0197] This structure allows the system to effectively detect unknown fraudulent messages while providing personalized security measures that take user emotions into consideration.

[0198] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0199] Step 1:

[0200] The server collects message patterns related to fraudulent activity from an external database. It takes a database of known fraudulent messages as input and outputs message data exhibiting fraudulent characteristics. This data is used to train machine learning algorithms.

[0201] Step 2:

[0202] The server trains a generative AI model using the collected message patterns. The message data with fraud features obtained in step 1 is used as input, and a trained model for fraud detection is generated as output. TensorFlow is used to train the model during this process.

[0203] Step 3:

[0204] The server generates a pattern file based on the fraud features obtained from the trained model. The input is the trained model obtained in step 2, and the output is a pattern file for detecting fraudulent activity. This enables rapid detection even for unknown messages.

[0205] Step 4:

[0206] The user's device sends received messages to the server. Messages received by the user are taken into the device as input, and these messages are uploaded to the server as output. The device uses a specific protocol for transmission.

[0207] Step 5:

[0208] The server scans received messages in real time using the aforementioned pattern file to detect potential fraud. The input is the message obtained in step 4, and the output is a judgment on whether the message is fraudulent or not. This process utilizes the inference capabilities of a trained model.

[0209] Step 6:

[0210] The device analyzes the user's facial expressions and voice using the Emotion SDK and OpenCV while they are viewing a message, and infers their emotional state. The input is the user's camera video and audio data, and the output is the emotion analysis result.

[0211] Step 7:

[0212] The server generates personalized warning messages based on sentiment analysis results and notifies the terminal. The input combines the fraud detection results from step 5 and the sentiment analysis results from step 6, creating a user-specific warning as output. This allows the user to receive warnings about fraud.

[0213] 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.

[0214] 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.

[0215] 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.

[0216] [Second Embodiment]

[0217] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0218] 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.

[0219] 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).

[0220] 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.

[0221] 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.

[0222] 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).

[0223] 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.

[0224] 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.

[0225] 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.

[0226] 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.

[0227] 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.

[0228] 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".

[0229] This invention relates to an embodiment of an advanced detection system for automatically detecting messages related to fraudulent activity and protecting users. The specific implementation details are described below.

[0230] First, the server retrieves data on past fraudulent activities collected from around the world. This includes existing databases and samples of newly reported fraudulent messages. Based on the collected data, the server analyzes the characteristics of these messages and trains a generative model. This training learns patterns specific to fraudulent messages and generates similar new messages.

[0231] Next, the server analyzes the patterns in the generated messages to create a new pattern file. This pattern file contains characteristic keywords and phrases common to fraudulent messages and is used for comparison with the received messages.

[0232] When a user receives a new message, it is automatically sent from their device to the server. The server scans the received message in real time and compares it against a pattern file. This comparison allows the server to determine if the message is potentially fraudulent. For example, if the subject line of a message contains phrases such as "Urgent action required," it may be judged to be similar to past fraud patterns.

[0233] If the server detects a message as fraudulent, it will notify the user of the warning. The warning will include instructions that the message may be dangerous and that the user should not view the details. The server can also block specific messages to prevent the reception of malicious messages.

[0234] In this way, by identifying fraudulent activity and responding quickly, a system is created that protects users from fraud and provides safe access to content.

[0235] The following describes the processing flow.

[0236] Step 1:

[0237] The servers collect data related to past fraudulent activities from sources around the world. This data includes known phishing emails, fake investment offers, and more.

[0238] Step 2:

[0239] The server analyzes the collected data and extracts key features. This process includes data cleaning, deduplication, and identification of relevant keywords and phrases.

[0240] Step 3:

[0241] The server uses the extracted features to train a generative AI model. The AI ​​model learns the characteristics of fraudulent messages and is able to generate similar messages.

[0242] Step 4:

[0243] The server analyzes the output of the trained AI model and generates a pattern file containing the characteristics of fraudulent messages. This pattern file is used to detect fraudulent activity.

[0244] Step 5:

[0245] When a user receives a message, the device sends that message to the server for scanning.

[0246] Step 6:

[0247] The server compares incoming messages against a pattern file in real time. This comparison helps determine whether the message is potentially fraudulent.

[0248] Step 7:

[0249] If a message is identified as fraudulent, the server sends details to the device and warns the user of the danger. The warning includes cautionary information.

[0250] Step 8:

[0251] If necessary, the server will block the reception of messages suspected of being fraudulent. This process prevents such messages from reaching the user.

[0252] (Example 1)

[0253] 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."

[0254] Fraudulent messages are becoming more sophisticated year after year, making them difficult to detect with conventional technologies and increasing the risk of users becoming victims. To address this problem, an effective means is needed to analyze messages in real time, quickly and accurately detect potential fraud, and warn users.

[0255] 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.

[0256] In this invention, the server includes means for collecting information, means for training a program that learns the characteristics of fraudulent activities using the information, and means for generating pattern information based on the characteristics of fraudulent activities obtained from the trained program. This makes it possible to analyze the received data in real time, detect the possibility of fraud with high accuracy and speed, and warn the user.

[0257] "Information" refers to any data related to past fraudulent activities collected from databases and online sources.

[0258] A "program" is a system that uses information to learn the characteristics of fraudulent activities and makes inferences and predictions based on that learning.

[0259] "Training" is the process by which a program uses information to learn fraud patterns and improve its ability to identify the characteristics of fraudulent activities.

[0260] "Pattern information" refers to a dataset containing characteristic keywords, phrases, and related structures common to fraudulent activities.

[0261] "Data" refers to any messages or pieces of information that a system receives and analyzes.

[0262] "Analysis" is the process of evaluating received data and determining whether it contains characteristics of fraudulent activity.

[0263] "User" refers to an individual or organization that is protected from fraudulent activity by using the system.

[0264] This invention relates to an advanced detection system for automatically detecting messages related to fraudulent activity and protecting users. The system is implemented through server, terminal, and user components.

[0265] The server collects data related to past fraudulent activities from existing databases and online resources to gather information. This data is used to train a generative AI model for the server to learn. This process uses natural language processing techniques and machine learning algorithms, and Python libraries (e.g., scikit-learn, TensorFlow) are used for data cleaning and feature engineering.

[0266] The generative AI model learns patterns specific to fraudulent activities and, after training, has the ability to generate pattern information based on the characteristics of fraudulent messages. This pattern information includes characteristic keywords and phrases common to fraudulent messages, which are used to compare with received data. The server has the capability to scan incoming messages in real time based on this pattern information.

[0267] The terminal automatically sends messages received by the user to the server. This is done through dedicated client software and is performed every time the user receives a new message. The server analyzes the received messages to check for potential fraud. If the message is identified as fraudulent, the server sends a warning to the user to protect them from further harm.

[0268] For example, if a user receives a message saying, "Please update your account information," the server will identify this as a phishing message and issue a warning immediately. Examples of prompts used in the system's AI model include:

[0269] "Please generate email subject lines that are easy to deceive."

[0270] In this way, users are protected from fraudulent activities and can safely use the content.

[0271] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0272] Step 1:

[0273] The server collects information. The input is data on fraudulent activities obtained from existing databases and online resources. The data includes examples of phishing emails and spam messages. The server collects this data in a secure manner and stores it in a database. This process provides foundational data for pattern analysis of fraudulent messages.

[0274] Step 2:

[0275] The server trains a generative AI model using the collected data. The input is the dataset collected in step 1, and the output is the trained generative AI model. The server utilizes natural language processing techniques and machine learning algorithms, preprocessing the data using Python libraries (e.g., scikit-learn, TensorFlow) to train the AI ​​model. This generates a model that has learned the features of fraudulent messages.

[0276] Step 3:

[0277] The server generates pattern information from the trained model. The input is the AI ​​model from step 2, and the output is pattern information summarizing the characteristics of fraudulent activity. The server analyzes the features learned by the AI ​​model, extracts keywords and phrases common to fraudulent messages, and creates a pattern file. This pattern information will be used for future message evaluation.

[0278] Step 4:

[0279] When a user receives a new message, the terminal sends that message to the server. The input is the message received by the user, and the output is the message data sent to the server. The terminal uses dedicated client software, and the message is automatically forwarded to the server according to the settings of the email client, etc.

[0280] Step 5:

[0281] The server analyzes received messages in real time and compares them with the pattern information generated in step 3. The input is the message sent by the user and the pattern information, and the output is the result of identifying potentially fraudulent messages. The server determines whether keywords and phrases in the message match the pattern information and identifies them if they are likely to be fraudulent.

[0282] Step 6:

[0283] When the server detects a fraudulent message, it notifies the user with a warning. The input is the fraudulent message identified in step 5, and the output is a warning message to the user. The server warns the user not to confirm details about dangerous messages, strengthening security. Specifically, a notification is sent to the user's terminal, and measures are taken against the illegal act.

[0284] (Application Example 1)

[0285] Next, Application Example 1 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart glasses 214 are referred to as the "terminal".

[0286] In modern information society, damages caused by illegal information and fraud are increasing, and ensuring the information security of users has become an important issue. However, it is difficult to detect fraud quickly and accurately with conventional methods and effectively protect users. Therefore, there is a need for a technology that can automatically detect illegal information in real time and respond quickly.

[0287] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0288] In this invention, the server includes means for collecting information patterns related to fraud, means for training a generation model that learns the characteristics of illegal acts, and means for generating a specific pattern file. Thereby, it becomes possible to immediately detect illegal information and notify the user with a warning.

[0289] "Fraud" is an illegal act committed for the purpose of deceiving others to obtain money or information improperly.

[0290] "Information pattern" is a structure or form that shows the characteristics and tendencies common to a specific type of information.

[0291] A "generative model" is a machine learning model that learns specific patterns or features from data and is used to predict or generate new data.

[0292] A "specific pattern file" is a data file used for detecting fraudulent activity, constructed based on the characteristics of fraudulent behavior learned by a generative model.

[0293] "Real-time" means responding to or processing an event immediately at the moment it occurs.

[0294] "Notification" is the process of informing recipients of changes or events in information.

[0295] "Restricting" means inhibiting certain actions or access to prevent the influx or use of undesirable information.

[0296] The system for implementing this invention primarily revolves around interaction between a server, a terminal, and the user. The server first collects information patterns related to fraudulent activity from around the world. This includes using existing databases and collecting samples of newly reported fraud cases. The server uses the collected information patterns to train a generative AI model and learn the characteristics of fraudulent activity. Once training is complete, the server generates specific pattern files based on the characteristics of fraudulent activity. This builds the foundational data for detecting fraudulent activity.

[0297] The terminal is responsible for forwarding messages received by the user to the server in real time. The received messages are immediately compared against a specific pattern file on the server. This comparison checks, for example, whether the message subject contains a phrase such as "urgent action required." If a pattern match is found, the server detects the message as malicious information.

[0298] Users will receive a warning via their device regarding malicious information detected by the server. The warning will include information about the potential danger of the message and a request not to view the details. The server will also take additional safety measures, such as restricting the reception of such messages, as needed.

[0299] For example, if a user receives a message from a travel booking service stating, "There is a problem with your account. Immediate action is required," the server will immediately identify this message as fraudulent information and issue an appropriate warning to the user.

[0300] An example of a prompt message would be, "I would like to learn about the common characteristics of scam messages. Please tell me about past patterns in scam emails."

[0301] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0302] Step 1:

[0303] The server collects information patterns related to fraudulent activities from a database. This includes past fraud messages and reported cases. By feeding relevant data as input and extracting information patterns, it obtains a dataset for analysis as output.

[0304] Step 2:

[0305] The server trains a generative AI model using the collected dataset. By supplying information patterns as input and allowing the generative AI model to learn the characteristics of fraudulent behavior, the server obtains a trained model as output.

[0306] Step 3:

[0307] The server uses the trained generative AI model to generate a specific pattern file. By using the trained model as input and analyzing the characteristics of fraud behavior, an identification pattern file is created as output.

[0308] Step 4:

[0309] The terminal sends the new message received by the user to the server. By supplying the received message as input and performing real-time scanning on the server, the output is the result of message analysis.

[0310] Step 5:

[0311] The server compares the received message with the specific pattern file to determine the possibility of fraud. By using the message data and the pattern file as input and performing data comparison processing, the output is the result of fraud determination.

[0312] Step 6:

[0313] If the server determines that the information is fraudulent, it notifies the user of a warning through the terminal. The input is the result of fraud determination. By generating a warning message and sending it to the user, a warning notification is generated as output.

[0314] Step 7:

[0315] If the server determines that the message is fraudulent, it performs a process to restrict the reception of that message. By using the result of fraud determination as input and implementing message blocking, security is enhanced as output.

[0316] Furthermore, an emotion engine for estimating the user's emotion may be combined. That is, the specific processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform specific processing using the user's emotion.

[0317] This invention relates to a system that automatically detects messages related to fraudulent activity and more effectively protects users by recognizing their emotions. Specific embodiments are described below.

[0318] First, the server collects data on past fraudulent messages and analyzes their characteristics. Based on this analysis, it trains an AI model to generate patterns related to fraudulent activity from these messages. This enhances the ability to detect potential fraud in unknown messages in real time.

[0319] When a user receives a message, the device sends it to a server for fraud pattern matching. Simultaneously, the device is equipped with an emotion engine that tracks the user's reactions. The emotion engine analyzes the user's facial expressions and tone of voice while they are reading the message to infer their emotional state.

[0320] For example, if a user receives a message prompting them to click a suspicious link and displays a wary or anxious expression, the emotion engine sends that information to the server. The server analyzes the user's emotional response to the potentially fraudulent message and generates a personalized warning message, which is then sent to the device. This warning message provides practical safety advice tailored to the user's response.

[0321] Furthermore, the server records user sentiment data and uses it to improve the fraud detection model. This allows the system to learn appropriate responses for individual users and improve its accuracy.

[0322] Thus, in addition to the rapid identification and countermeasures against fraudulent activities, the present invention can further enhance user safety by providing personalized services that take user emotions into consideration.

[0323] The following describes the processing flow.

[0324] Step 1:

[0325] The server collects historical message data related to fraudulent activity from a wide range of data sources. This data includes fraudulent emails, phishing messages, and more.

[0326] Step 2:

[0327] The server analyzes the collected data to extract common patterns in fraudulent messages. This includes analyzing characteristic phrases and sender information.

[0328] Step 3:

[0329] The server trains a generative AI model based on the extracted patterns and generates pattern files to detect fraudulent activity.

[0330] Step 4:

[0331] When a user receives a message, the device scans it and sends it to the server in real time.

[0332] Step 5:

[0333] The server compares the received message against a pattern file to assess the likelihood of fraud. Based on this assessment, it determines whether or not the message is fraudulent.

[0334] Step 6:

[0335] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine analyzes the user's facial expressions and voice when they check a message to infer their emotions.

[0336] Step 7:

[0337] If the emotion engine detects any anxiety or concerns from the user, that information is sent to the server and analyzed along with the fraud detection results.

[0338] Step 8:

[0339] The server generates and sends personalized warning messages to the terminal based on the user's emotions. These messages include specific safety advice.

[0340] Step 9:

[0341] The server records user sentiment data and uses it to improve the accuracy of the model. This continuous feedback improves the overall responsiveness of the system.

[0342] (Example 2)

[0343] 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".

[0344] In recent years, online fraud has increased, causing significant losses for individuals and organizations. Traditional methods are insufficient to effectively detect and prevent fraud in real time, and they fail to take into account user emotions and reactions. It is necessary to address this situation and provide a more secure communication environment.

[0345] 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.

[0346] In this invention, the server includes means for collecting information patterns related to fraudulent activities, means for training a generative model that learns the characteristics of fraudulent activities using the information patterns, and means for recording sentiment data and improving the detection model. This enables the detection of potential fraud in real time and individual responses that take into account the user's emotions.

[0347] "Fraud" refers to any malicious act or attempt to deceive another person and illegally obtain money or assets.

[0348] "Information patterns" refer to a set of patterns or forms that indicate characteristic data or behavioral tendencies associated with fraudulent activity.

[0349] A "generative model" refers to an artificial intelligence model used to learn specific features or patterns from data and then generate or detect those features based on new data.

[0350] An "identification file" is a data file containing characteristic patterns of fraudulent activity, providing a standard used for matching with received information.

[0351] "Emotional state" refers to the psychological or emotional reactions or attitudes that a user exhibits when acquiring information.

[0352] "Emotional data" refers to data that quantifies or digitizes information about a user's emotional state and is used for system learning and improvement.

[0353] This invention relates to a system that automatically detects messages related to fraudulent activity and provides protection while also considering the user's emotions. Specific embodiments thereof are described below.

[0354] The server first collects information patterns related to past fraudulent activities from the internet and other databases. Based on this, the server trains a generative AI model. The generative AI model uses the collected data to learn the characteristics of fraudulent messages and generates identification files to identify the likelihood of fraud in new messages.

[0355] When a user receives a message, the device immediately sends it to the server. The server compares the message against the aforementioned identification file and assesses the likelihood of fraud in real time. Simultaneously, the device is equipped with an emotion engine that analyzes the user's facial expressions and tone of voice in response to the message to infer the user's emotional state.

[0356] For example, if a user receives a message prompting them to click a suspicious link and displays a surprised or anxious expression, the emotion engine sends that data to the server. Based on this, the server analyzes the user's response to the potentially fraudulent message, generates an appropriate warning message, and sends it back to the device. This warning is customized to the user's current emotions and includes practical safety advice.

[0357] Furthermore, the server records the acquired user sentiment data and uses it to improve the generative AI model. This allows the system to provide more personalized responses over time, increasing the accuracy of user protection.

[0358] As a concrete example, a possible prompt message might be: "While online shopping, you received a message offering a high-priced item at a very low price. How should you handle this message?" This system allows users to communicate more safely and securely.

[0359] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0360] Step 1:

[0361] The server collects information patterns related to fraudulent activity. This process gathers information from multiple sources, including databases of fraudulent emails and unauthorized access history. Using a dataset of past fraudulent activities as input, the server outputs information patterns that characterize fraud.

[0362] Step 2:

[0363] The server trains a generative AI model using the acquired information patterns. The collected information patterns are used as input, and machine learning algorithms are applied to learn these patterns. The output is a trained generative AI model for identifying fraudulent activity. This model improves the ability to identify potential fraud even in new messages.

[0364] Step 3:

[0365] When a user receives a message, the terminal sends that message to the server. The input is the newly received message, which is then sent to the server. The output is the message, which is then used in the next matching step.

[0366] Step 4:

[0367] The server compares received messages against an identification file. The input is a message received from a user, which is compared to an already constructed identification file. A generative AI model is used to assess the likelihood of fraud, resulting in a judgment regarding the message's fraud risk as output.

[0368] Step 5:

[0369] The device uses an emotion engine to analyze the user's emotional state. As input, the emotion engine collects the user's facial expressions and tone of voice as they read messages. As output, the user's emotional state is estimated and sent to the server.

[0370] Step 6:

[0371] The server generates personalized warning messages based on fraud risk and emotional state. The server receives emotional data as input and combines it with fraud risk data to construct specific warning messages. The output is customized advice sent to the user.

[0372] Step 7:

[0373] The server records user sentiment data and feedback on warnings to improve the generative AI model. It receives personalized warning message responses and sentiment data as input, and uses this data to retrain and improve the model. As output, it provides a generative AI model with improved system adaptability and accuracy.

[0374] (Application Example 2)

[0375] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0376] Internet-based fraud is becoming increasingly sophisticated, rendering conventional message filtering and warning systems insufficient. Furthermore, issuing uniform warnings without considering the user's emotions or circumstances can lead to false alarms and user stress. This invention aims to solve these problems by providing a personalized warning system that not only detects fraudulent messages quickly and effectively but also takes into account the user's emotional response.

[0377] 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.

[0378] In this invention, the server includes means for collecting message patterns related to fraudulent activity; means for training a generative model that learns the characteristics of fraudulent activity using the message patterns; means for generating a pattern file based on the characteristics of fraudulent activity obtained from the trained generative model; means for detecting the possibility of fraud by comparing the received message with the pattern file; means for inferring the emotional state by analyzing the user's facial expressions and voice when receiving a fraudulent message; and means for generating and notifying the user of a personalized warning message based on the user's emotional state. This makes it possible to detect unknown fraudulent messages and to provide appropriate warnings that take the user's emotions into consideration.

[0379] A "message pattern associated with fraudulent activity" is a form of information that has a common structure and characteristics and is used to identify communication content that contains fraudulent activity.

[0380] "Training a generative model" is the process of improving a machine learning algorithm using message data related to fraudulent activity to enhance its ability to identify new fraudulent messages.

[0381] "Generating pattern files" is the process of aggregating fraud features extracted from trained models and establishing criteria for detecting specific fraudulent activities based on those features.

[0382] "Detecting potential fraud" means analyzing the received message and determining whether it contains elements that suggest fraudulent activity.

[0383] "Analyzing the user's facial expressions and voice to infer their emotional state" means analyzing the user's facial expressions and tone of voice when they receive a message to estimate their emotions and psychological state at that time.

[0384] "Generating and notifying users of personalized warning messages" means customizing the content of warnings against fraudulent messages according to the user's emotional state and informing the user of that content.

[0385] To implement the invention, it is necessary to build a system that combines fraudulent message detection with user sentiment analysis. This system mainly consists of two main components: a server and a terminal.

[0386] The server collects message patterns related to fraudulent activity and trains a generative model based on historical data. Specifically, it develops a generative AI model using TensorFlow and learns the characteristics of fraudulent activity. Furthermore, it generates pattern files using the obtained characteristics and establishes criteria for detecting potential fraud. At this time, prompt statements are utilized to efficiently train the AI ​​model.

[0387] The device sends messages received by the user to a server in real time to check for potential fraud. The device also features an emotion engine utilizing the Emotion SDK and OpenCV, which analyzes the user's facial expressions and voice to infer their emotional state. This emotional information is used to generate personalized warning messages.

[0388] For example, if a user receives a message stating, "Your account will be closed if you don't act now," and shows a worried expression, the device sends this information to the server. The server flags it as a scam message, generates a personalized alert based on the facial expression data, and sends this to the device.

[0389] Examples of specific prompt messages used to achieve this process include the following:

[0390] "Message received: 'Your bank account is scheduled to be frozen. Please click this link to confirm.' User reaction: Frowning, in a worried voice."

[0391] This structure allows the system to effectively detect unknown fraudulent messages while providing personalized security measures that take user emotions into consideration.

[0392] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0393] Step 1:

[0394] The server collects message patterns related to fraudulent activity from an external database. It takes a database of known fraudulent messages as input and outputs message data exhibiting fraudulent characteristics. This data is used to train machine learning algorithms.

[0395] Step 2:

[0396] The server trains a generative AI model using the collected message patterns. The message data with fraud features obtained in step 1 is used as input, and a trained model for fraud detection is generated as output. TensorFlow is used to train the model during this process.

[0397] Step 3:

[0398] The server generates a pattern file based on the fraud features obtained from the trained model. The input is the trained model obtained in step 2, and the output is a pattern file for detecting fraudulent activity. This enables rapid detection even for unknown messages.

[0399] Step 4:

[0400] The user's device sends received messages to the server. Messages received by the user are taken into the device as input, and these messages are uploaded to the server as output. The device uses a specific protocol for transmission.

[0401] Step 5:

[0402] The server scans received messages in real time using the aforementioned pattern file to detect potential fraud. The input is the message obtained in step 4, and the output is a judgment on whether the message is fraudulent or not. This process utilizes the inference capabilities of a trained model.

[0403] Step 6:

[0404] The device analyzes the user's facial expressions and voice using the Emotion SDK and OpenCV while they are viewing a message, and infers their emotional state. The input is the user's camera video and audio data, and the output is the emotion analysis result.

[0405] Step 7:

[0406] The server generates personalized warning messages based on sentiment analysis results and notifies the terminal. The input combines the fraud detection results from step 5 and the sentiment analysis results from step 6, creating a user-specific warning as output. This allows the user to receive warnings about fraud.

[0407] 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.

[0408] 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.

[0409] 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.

[0410] [Third Embodiment]

[0411] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0412] 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.

[0413] 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).

[0414] 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.

[0415] 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.

[0416] 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).

[0417] 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.

[0418] 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.

[0419] 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.

[0420] 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.

[0421] 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.

[0422] 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".

[0423] This invention relates to an embodiment of an advanced detection system for automatically detecting messages related to fraudulent activity and protecting users. The specific implementation details are described below.

[0424] First, the server retrieves data on past fraudulent activities collected from around the world. This includes existing databases and samples of newly reported fraudulent messages. Based on the collected data, the server analyzes the characteristics of these messages and trains a generative model. This training learns patterns specific to fraudulent messages and generates similar new messages.

[0425] Next, the server analyzes the patterns in the generated messages to create a new pattern file. This pattern file contains characteristic keywords and phrases common to fraudulent messages and is used for comparison with the received messages.

[0426] When a user receives a new message, it is automatically sent from their device to the server. The server scans the received message in real time and compares it against a pattern file. This comparison allows the server to determine if the message is potentially fraudulent. For example, if the subject line of a message contains phrases such as "Urgent action required," it may be judged to be similar to past fraud patterns.

[0427] If the server detects a message as fraudulent, it will notify the user of the warning. The warning will include instructions that the message may be dangerous and that the user should not view the details. The server can also block specific messages to prevent the reception of malicious messages.

[0428] In this way, by identifying fraudulent activity and responding quickly, a system is created that protects users from fraud and provides safe access to content.

[0429] The following describes the processing flow.

[0430] Step 1:

[0431] The servers collect data related to past fraudulent activities from sources around the world. This data includes known phishing emails, fake investment offers, and more.

[0432] Step 2:

[0433] The server analyzes the collected data and extracts key features. This process includes data cleaning, deduplication, and identification of relevant keywords and phrases.

[0434] Step 3:

[0435] The server uses the extracted features to train a generative AI model. The AI ​​model learns the characteristics of fraudulent messages and is able to generate similar messages.

[0436] Step 4:

[0437] The server analyzes the output of the trained AI model and generates a pattern file containing the characteristics of fraudulent messages. This pattern file is used to detect fraudulent activity.

[0438] Step 5:

[0439] When a user receives a message, the device sends that message to the server for scanning.

[0440] Step 6:

[0441] The server compares incoming messages against a pattern file in real time. This comparison helps determine whether the message is potentially fraudulent.

[0442] Step 7:

[0443] If a message is identified as fraudulent, the server sends details to the device and warns the user of the danger. The warning includes cautionary information.

[0444] Step 8:

[0445] If necessary, the server will block the reception of messages suspected of being fraudulent. This process prevents such messages from reaching the user.

[0446] (Example 1)

[0447] 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."

[0448] Fraudulent messages are becoming more sophisticated year after year, making them difficult to detect with conventional technologies and increasing the risk of users becoming victims. To address this problem, an effective means is needed to analyze messages in real time, quickly and accurately detect potential fraud, and warn users.

[0449] 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.

[0450] In this invention, the server includes means for collecting information, means for training a program that learns the characteristics of fraudulent activities using the information, and means for generating pattern information based on the characteristics of fraudulent activities obtained from the trained program. This makes it possible to analyze the received data in real time, detect the possibility of fraud with high accuracy and speed, and warn the user.

[0451] "Information" refers to any data related to past fraudulent activities collected from databases and online sources.

[0452] A "program" is a system that uses information to learn the characteristics of fraudulent activities and makes inferences and predictions based on that learning.

[0453] "Training" is the process by which a program uses information to learn fraud patterns and improve its ability to identify the characteristics of fraudulent activities.

[0454] "Pattern information" refers to a dataset containing characteristic keywords, phrases, and related structures common to fraudulent activities.

[0455] "Data" refers to any messages or pieces of information that a system receives and analyzes.

[0456] "Analysis" is the process of evaluating received data and determining whether it contains characteristics of fraudulent activity.

[0457] "User" refers to an individual or organization that is protected from fraudulent activity by using the system.

[0458] This invention relates to an advanced detection system for automatically detecting messages related to fraudulent activity and protecting users. The system is implemented through server, terminal, and user components.

[0459] The server collects data related to past fraudulent activities from existing databases and online resources to gather information. This data is used to train a generative AI model for the server to learn. This process uses natural language processing techniques and machine learning algorithms, and Python libraries (e.g., scikit-learn, TensorFlow) are used for data cleaning and feature engineering.

[0460] The generative AI model learns patterns specific to fraudulent activities and, after training, has the ability to generate pattern information based on the characteristics of fraudulent messages. This pattern information includes characteristic keywords and phrases common to fraudulent messages, which are used to compare with received data. The server has the capability to scan incoming messages in real time based on this pattern information.

[0461] The terminal automatically sends messages received by the user to the server. This is done through dedicated client software and is performed every time the user receives a new message. The server analyzes the received messages to check for potential fraud. If the message is identified as fraudulent, the server sends a warning to the user to protect them from further harm.

[0462] For example, if a user receives a message saying, "Please update your account information," the server will identify this as a phishing message and issue a warning immediately. Examples of prompts used in the system's AI model include:

[0463] "Please generate email subject lines that are easy to deceive."

[0464] In this way, users are protected from fraudulent activities and can safely use the content.

[0465] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0466] Step 1:

[0467] The server collects information. The input is data on fraudulent activities obtained from existing databases and online resources. The data includes examples of phishing emails and spam messages. The server collects this data in a secure manner and stores it in a database. This process provides foundational data for pattern analysis of fraudulent messages.

[0468] Step 2:

[0469] The server trains a generative AI model using the collected data. The input is the dataset collected in step 1, and the output is the trained generative AI model. The server utilizes natural language processing techniques and machine learning algorithms, preprocessing the data using Python libraries (e.g., scikit-learn, TensorFlow) to train the AI ​​model. This generates a model that has learned the features of fraudulent messages.

[0470] Step 3:

[0471] The server generates pattern information from the trained model. The input is the AI ​​model from step 2, and the output is pattern information summarizing the characteristics of fraudulent activity. The server analyzes the features learned by the AI ​​model, extracts keywords and phrases common to fraudulent messages, and creates a pattern file. This pattern information will be used for future message evaluation.

[0472] Step 4:

[0473] When a user receives a new message, the terminal sends that message to the server. The input is the message received by the user, and the output is the message data sent to the server. The terminal uses dedicated client software, and the message is automatically forwarded to the server according to the settings of the email client, etc.

[0474] Step 5:

[0475] The server analyzes received messages in real time and compares them with the pattern information generated in step 3. The input is the message sent by the user and the pattern information, and the output is the result of identifying potentially fraudulent messages. The server determines whether keywords and phrases in the message match the pattern information and identifies them if they are likely to be fraudulent.

[0476] Step 6:

[0477] If the server detects a fraudulent message, it notifies the user of the warning. The input is the fraudulent message identified in step 5, and the output is the warning message to the user. The server strengthens security by warning the user not to view details about potentially dangerous messages. Specifically, a notification is sent to the user's device, and measures are taken to prevent fraudulent activity.

[0478] (Application Example 1)

[0479] 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."

[0480] In today's information society, the damage caused by fraudulent information and fraudulent activities is increasing, making it a crucial issue to ensure user information security. However, conventional methods make it difficult to quickly and accurately detect fraudulent activities and effectively protect users. Therefore, there is a need for technology that can automatically detect fraudulent information in real time and respond quickly.

[0481] 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.

[0482] In this invention, the server includes means for collecting information patterns related to fraudulent activity, means for training a generative model that learns the characteristics of fraudulent activity, and means for generating specific pattern files. This enables immediate detection of fraudulent information and notification of warnings to users.

[0483] "Fraudulent activity" refers to unethical actions taken with the intent of deceiving others to unjustly obtain money or information.

[0484] An "information pattern" is a structure or form that exhibits common characteristics or tendencies in a particular type of information.

[0485] A "generative model" is a machine learning model that learns specific patterns or features from data and is used to predict or generate new data.

[0486] A "specific pattern file" is a data file used for detecting fraudulent activity, constructed based on the characteristics of fraudulent behavior learned by a generative model.

[0487] "Real-time" means responding to or processing an event immediately at the moment it occurs.

[0488] "Notification" is the process of informing recipients of changes or events in information.

[0489] "Restricting" means inhibiting certain actions or access to prevent the influx or use of undesirable information.

[0490] The system for implementing this invention primarily revolves around interaction between a server, a terminal, and the user. The server first collects information patterns related to fraudulent activity from around the world. This includes using existing databases and collecting samples of newly reported fraud cases. The server uses the collected information patterns to train a generative AI model and learn the characteristics of fraudulent activity. Once training is complete, the server generates specific pattern files based on the characteristics of fraudulent activity. This builds the foundational data for detecting fraudulent activity.

[0491] The terminal is responsible for forwarding messages received by the user to the server in real time. The received messages are immediately compared against a specific pattern file on the server. This comparison checks, for example, whether the message subject contains a phrase such as "urgent action required." If a pattern match is found, the server detects the message as malicious information.

[0492] Users will receive a warning via their device regarding malicious information detected by the server. The warning will include information about the potential danger of the message and a request not to view the details. The server will also take additional safety measures, such as restricting the reception of such messages, as needed.

[0493] For example, if a user receives a message from a travel booking service stating, "There is a problem with your account. Immediate action is required," the server will immediately identify this message as fraudulent information and issue an appropriate warning to the user.

[0494] An example of a prompt message would be, "I would like to learn about the common characteristics of scam messages. Please tell me about past patterns in scam emails."

[0495] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0496] Step 1:

[0497] The server collects information patterns related to fraudulent activities from a database. This includes past fraud messages and reported cases. By feeding relevant data as input and extracting information patterns, it obtains a dataset for analysis as output.

[0498] Step 2:

[0499] The server trains a generative AI model using the collected dataset. By supplying information patterns as input and allowing the generative AI model to learn the characteristics of fraudulent behavior, the server obtains a trained model as output.

[0500] Step 3:

[0501] The server generates specific pattern files using a trained generative AI model. It uses the trained model as input and analyzes the characteristics of fraudulent activity to create identification pattern files as output.

[0502] Step 4:

[0503] The terminal sends new messages received by the user to the server. By supplying received messages as input and performing a real-time scan on the server, the output becomes the result of message analysis.

[0504] Step 5:

[0505] The server compares received messages against a specific pattern file to determine if they are potentially fraudulent. Using message data and a pattern file as input, the server performs a data matching process, and the output is the result of the fraud detection.

[0506] Step 6:

[0507] If the server determines that the information is fraudulent, it will notify the user via the terminal. The input is the result of the fraud detection, and the server generates a warning message and sends it to the user, resulting in a warning notification as output.

[0508] Step 7:

[0509] The server, if it determines a message is fraudulent, will restrict its reception. By using the fraud detection result as input and blocking messages, it enhances security as output.

[0510] 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.

[0511] This invention relates to a system that automatically detects messages related to fraudulent activity and more effectively protects users by recognizing their emotions. Specific embodiments are described below.

[0512] First, the server collects data on past fraudulent messages and analyzes their characteristics. Based on this analysis, it trains an AI model to generate patterns related to fraudulent activity from these messages. This enhances the ability to detect potential fraud in unknown messages in real time.

[0513] When a user receives a message, the device sends it to a server for fraud pattern matching. Simultaneously, the device is equipped with an emotion engine that tracks the user's reactions. The emotion engine analyzes the user's facial expressions and tone of voice while they are reading the message to infer their emotional state.

[0514] For example, if a user receives a message prompting them to click a suspicious link and displays a wary or anxious expression, the emotion engine sends that information to the server. The server analyzes the user's emotional response to the potentially fraudulent message and generates a personalized warning message, which is then sent to the device. This warning message provides practical safety advice tailored to the user's response.

[0515] Furthermore, the server records user sentiment data and uses it to improve the fraud detection model. This allows the system to learn appropriate responses for individual users and improve its accuracy.

[0516] Thus, in addition to the rapid identification and countermeasures against fraudulent activities, the present invention can further enhance user safety by providing personalized services that take user emotions into consideration.

[0517] The following describes the processing flow.

[0518] Step 1:

[0519] The server collects historical message data related to fraudulent activity from a wide range of data sources. This data includes fraudulent emails, phishing messages, and more.

[0520] Step 2:

[0521] The server analyzes the collected data to extract common patterns in fraudulent messages. This includes analyzing characteristic phrases and sender information.

[0522] Step 3:

[0523] The server trains a generative AI model based on the extracted patterns and generates pattern files to detect fraudulent activity.

[0524] Step 4:

[0525] When a user receives a message, the device scans it and sends it to the server in real time.

[0526] Step 5:

[0527] The server compares the received message against a pattern file to assess the likelihood of fraud. Based on this assessment, it determines whether or not the message is fraudulent.

[0528] Step 6:

[0529] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine analyzes the user's facial expressions and voice when they check a message to infer their emotions.

[0530] Step 7:

[0531] If the emotion engine detects any anxiety or concerns from the user, that information is sent to the server and analyzed along with the fraud detection results.

[0532] Step 8:

[0533] The server generates and sends personalized warning messages to the terminal based on the user's emotions. These messages include specific safety advice.

[0534] Step 9:

[0535] The server records user sentiment data and uses it to improve the accuracy of the model. This continuous feedback improves the overall responsiveness of the system.

[0536] (Example 2)

[0537] 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."

[0538] In recent years, online fraud has increased, causing significant losses for individuals and organizations. Traditional methods are insufficient to effectively detect and prevent fraud in real time, and they fail to take into account user emotions and reactions. It is necessary to address this situation and provide a more secure communication environment.

[0539] 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.

[0540] In this invention, the server includes means for collecting information patterns related to fraudulent activities, means for training a generative model that learns the characteristics of fraudulent activities using the information patterns, and means for recording sentiment data and improving the detection model. This enables the detection of potential fraud in real time and individual responses that take into account the user's emotions.

[0541] "Fraud" refers to any malicious act or attempt to deceive another person and illegally obtain money or assets.

[0542] "Information patterns" refer to a set of patterns or forms that indicate characteristic data or behavioral tendencies associated with fraudulent activity.

[0543] A "generative model" refers to an artificial intelligence model used to learn specific features or patterns from data and then generate or detect those features based on new data.

[0544] An "identification file" is a data file containing characteristic patterns of fraudulent activity, providing a standard used for matching with received information.

[0545] "Emotional state" refers to the psychological or emotional reactions or attitudes that a user exhibits when acquiring information.

[0546] "Emotional data" refers to data that quantifies or digitizes information about a user's emotional state and is used for system learning and improvement.

[0547] This invention relates to a system that automatically detects messages related to fraudulent activity and provides protection while also considering the user's emotions. Specific embodiments thereof are described below.

[0548] The server first collects information patterns related to past fraudulent activities from the internet and other databases. Based on this, the server trains a generative AI model. The generative AI model uses the collected data to learn the characteristics of fraudulent messages and generates identification files to identify the likelihood of fraud in new messages.

[0549] When a user receives a message, the device immediately sends it to the server. The server compares the message against the aforementioned identification file and assesses the likelihood of fraud in real time. Simultaneously, the device is equipped with an emotion engine that analyzes the user's facial expressions and tone of voice in response to the message to infer the user's emotional state.

[0550] For example, if a user receives a message prompting them to click a suspicious link and displays a surprised or anxious expression, the emotion engine sends that data to the server. Based on this, the server analyzes the user's response to the potentially fraudulent message, generates an appropriate warning message, and sends it back to the device. This warning is customized to the user's current emotions and includes practical safety advice.

[0551] Furthermore, the server records the acquired user sentiment data and uses it to improve the generative AI model. This allows the system to provide more personalized responses over time, increasing the accuracy of user protection.

[0552] As a concrete example, a possible prompt message might be: "While online shopping, you received a message offering a high-priced item at a very low price. How should you handle this message?" This system allows users to communicate more safely and securely.

[0553] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0554] Step 1:

[0555] The server collects information patterns related to fraudulent activity. This process gathers information from multiple sources, including databases of fraudulent emails and unauthorized access history. Using a dataset of past fraudulent activities as input, the server outputs information patterns that characterize fraud.

[0556] Step 2:

[0557] The server trains a generative AI model using the acquired information patterns. The collected information patterns are used as input, and machine learning algorithms are applied to learn these patterns. The output is a trained generative AI model for identifying fraudulent activity. This model improves the ability to identify potential fraud even in new messages.

[0558] Step 3:

[0559] When a user receives a message, the terminal sends that message to the server. The input is the newly received message, which is then sent to the server. The output is the message, which is then used in the next matching step.

[0560] Step 4:

[0561] The server compares received messages against an identification file. The input is a message received from a user, which is compared to an already constructed identification file. A generative AI model is used to assess the likelihood of fraud, resulting in a judgment regarding the message's fraud risk as output.

[0562] Step 5:

[0563] The device uses an emotion engine to analyze the user's emotional state. As input, the emotion engine collects the user's facial expressions and tone of voice as they read messages. As output, the user's emotional state is estimated and sent to the server.

[0564] Step 6:

[0565] The server generates personalized warning messages based on fraud risk and emotional state. The server receives emotional data as input and combines it with fraud risk data to construct specific warning messages. The output is customized advice sent to the user.

[0566] Step 7:

[0567] The server records user sentiment data and feedback on warnings to improve the generative AI model. It receives personalized warning message responses and sentiment data as input, and uses this data to retrain and improve the model. As output, it provides a generative AI model with improved system adaptability and accuracy.

[0568] (Application Example 2)

[0569] 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."

[0570] Internet-based fraud is becoming increasingly sophisticated, rendering conventional message filtering and warning systems insufficient. Furthermore, issuing uniform warnings without considering the user's emotions or circumstances can lead to false alarms and user stress. This invention aims to solve these problems by providing a personalized warning system that not only detects fraudulent messages quickly and effectively but also takes into account the user's emotional response.

[0571] 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.

[0572] In this invention, the server includes means for collecting message patterns related to fraudulent activity; means for training a generative model that learns the characteristics of fraudulent activity using the message patterns; means for generating a pattern file based on the characteristics of fraudulent activity obtained from the trained generative model; means for detecting the possibility of fraud by comparing the received message with the pattern file; means for inferring the emotional state by analyzing the user's facial expressions and voice when receiving a fraudulent message; and means for generating and notifying the user of a personalized warning message based on the user's emotional state. This makes it possible to detect unknown fraudulent messages and to provide appropriate warnings that take the user's emotions into consideration.

[0573] A "message pattern associated with fraudulent activity" is a form of information that has a common structure and characteristics and is used to identify communication content that contains fraudulent activity.

[0574] "Training a generative model" is the process of improving a machine learning algorithm using message data related to fraudulent activity to enhance its ability to identify new fraudulent messages.

[0575] "Generating pattern files" is the process of aggregating fraud features extracted from trained models and establishing criteria for detecting specific fraudulent activities based on those features.

[0576] "Detecting potential fraud" means analyzing the received message and determining whether it contains elements that suggest fraudulent activity.

[0577] "Analyzing the user's facial expressions and voice to infer their emotional state" means analyzing the user's facial expressions and tone of voice when they receive a message to estimate their emotions and psychological state at that time.

[0578] "Generating and notifying users of personalized warning messages" means customizing the content of warnings against fraudulent messages according to the user's emotional state and informing the user of that content.

[0579] To implement the invention, it is necessary to build a system that combines fraudulent message detection with user sentiment analysis. This system mainly consists of two main components: a server and a terminal.

[0580] The server collects message patterns related to fraudulent activity and trains a generative model based on historical data. Specifically, it develops a generative AI model using TensorFlow and learns the characteristics of fraudulent activity. Furthermore, it generates pattern files using the obtained characteristics and establishes criteria for detecting potential fraud. At this time, prompt statements are utilized to efficiently train the AI ​​model.

[0581] The device sends messages received by the user to a server in real time to check for potential fraud. The device also features an emotion engine utilizing the Emotion SDK and OpenCV, which analyzes the user's facial expressions and voice to infer their emotional state. This emotional information is used to generate personalized warning messages.

[0582] For example, if a user receives a message stating, "Your account will be closed if you don't act now," and shows a worried expression, the device sends this information to the server. The server flags it as a scam message, generates a personalized alert based on the facial expression data, and sends this to the device.

[0583] Examples of specific prompt messages used to achieve this process include the following:

[0584] "Message received: 'Your bank account is scheduled to be frozen. Please click this link to confirm.' User reaction: Frowning, in a worried voice."

[0585] This structure allows the system to effectively detect unknown fraudulent messages while providing personalized security measures that take user emotions into consideration.

[0586] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0587] Step 1:

[0588] The server collects message patterns related to fraudulent activity from an external database. It takes a database of known fraudulent messages as input and outputs message data exhibiting fraudulent characteristics. This data is used to train machine learning algorithms.

[0589] Step 2:

[0590] The server trains a generative AI model using the collected message patterns. The message data with fraud features obtained in step 1 is used as input, and a trained model for fraud detection is generated as output. TensorFlow is used to train the model during this process.

[0591] Step 3:

[0592] The server generates a pattern file based on the fraud features obtained from the trained model. The input is the trained model obtained in step 2, and the output is a pattern file for detecting fraudulent activity. This enables rapid detection even for unknown messages.

[0593] Step 4:

[0594] The user's device sends received messages to the server. Messages received by the user are taken into the device as input, and these messages are uploaded to the server as output. The device uses a specific protocol for transmission.

[0595] Step 5:

[0596] The server scans received messages in real time using the aforementioned pattern file to detect potential fraud. The input is the message obtained in step 4, and the output is a judgment on whether the message is fraudulent or not. This process utilizes the inference capabilities of a trained model.

[0597] Step 6:

[0598] The device analyzes the user's facial expressions and voice using the Emotion SDK and OpenCV while they are viewing a message, and infers their emotional state. The input is the user's camera video and audio data, and the output is the emotion analysis result.

[0599] Step 7:

[0600] The server generates personalized warning messages based on sentiment analysis results and notifies the terminal. The input combines the fraud detection results from step 5 and the sentiment analysis results from step 6, creating a user-specific warning as output. This allows the user to receive warnings about fraud.

[0601] 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.

[0602] 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.

[0603] 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.

[0604] [Fourth Embodiment]

[0605] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0606] 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.

[0607] 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).

[0608] 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.

[0609] 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.

[0610] 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).

[0611] 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.

[0612] 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.

[0613] 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.

[0614] 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.

[0615] 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.

[0616] 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.

[0617] 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".

[0618] This invention relates to an embodiment of an advanced detection system for automatically detecting messages related to fraudulent activity and protecting users. The specific implementation details are described below.

[0619] First, the server retrieves data on past fraudulent activities collected from around the world. This includes existing databases and samples of newly reported fraudulent messages. Based on the collected data, the server analyzes the characteristics of these messages and trains a generative model. This training learns patterns specific to fraudulent messages and generates similar new messages.

[0620] Next, the server analyzes the patterns in the generated messages to create a new pattern file. This pattern file contains characteristic keywords and phrases common to fraudulent messages and is used for comparison with the received messages.

[0621] When a user receives a new message, it is automatically sent from their device to the server. The server scans the received message in real time and compares it against a pattern file. This comparison allows the server to determine if the message is potentially fraudulent. For example, if the subject line of a message contains phrases such as "Urgent action required," it may be judged to be similar to past fraud patterns.

[0622] If the server detects a message as fraudulent, it will notify the user of the warning. The warning will include instructions that the message may be dangerous and that the user should not view the details. The server can also block specific messages to prevent the reception of malicious messages.

[0623] In this way, by identifying fraudulent activity and responding quickly, a system is created that protects users from fraud and provides safe access to content.

[0624] The following describes the processing flow.

[0625] Step 1:

[0626] The servers collect data related to past fraudulent activities from sources around the world. This data includes known phishing emails, fake investment offers, and more.

[0627] Step 2:

[0628] The server analyzes the collected data and extracts key features. This process includes data cleaning, deduplication, and identification of relevant keywords and phrases.

[0629] Step 3:

[0630] The server uses the extracted features to train a generative AI model. The AI ​​model learns the characteristics of fraudulent messages and is able to generate similar messages.

[0631] Step 4:

[0632] The server analyzes the output of the trained AI model and generates a pattern file containing the characteristics of fraudulent messages. This pattern file is used to detect fraudulent activity.

[0633] Step 5:

[0634] When a user receives a message, the device sends that message to the server for scanning.

[0635] Step 6:

[0636] The server compares incoming messages against a pattern file in real time. This comparison helps determine whether the message is potentially fraudulent.

[0637] Step 7:

[0638] If a message is identified as fraudulent, the server sends details to the device and warns the user of the danger. The warning includes cautionary information.

[0639] Step 8:

[0640] If necessary, the server will block the reception of messages suspected of being fraudulent. This process prevents such messages from reaching the user.

[0641] (Example 1)

[0642] 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".

[0643] Fraudulent messages are becoming more sophisticated year after year, making them difficult to detect with conventional technologies and increasing the risk of users becoming victims. To address this problem, an effective means is needed to analyze messages in real time, quickly and accurately detect potential fraud, and warn users.

[0644] 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.

[0645] In this invention, the server includes means for collecting information, means for training a program that learns the characteristics of fraudulent activities using the information, and means for generating pattern information based on the characteristics of fraudulent activities obtained from the trained program. This makes it possible to analyze the received data in real time, detect the possibility of fraud with high accuracy and speed, and warn the user.

[0646] "Information" refers to any data related to past fraudulent activities collected from databases and online sources.

[0647] A "program" is a system that uses information to learn the characteristics of fraudulent activities and makes inferences and predictions based on that learning.

[0648] "Training" is the process by which a program uses information to learn fraud patterns and improve its ability to identify the characteristics of fraudulent activities.

[0649] "Pattern information" refers to a dataset containing characteristic keywords, phrases, and related structures common to fraudulent activities.

[0650] "Data" refers to any messages or pieces of information that a system receives and analyzes.

[0651] "Analysis" is the process of evaluating received data and determining whether it contains characteristics of fraudulent activity.

[0652] "User" refers to an individual or organization that is protected from fraudulent activity by using the system.

[0653] This invention relates to an advanced detection system for automatically detecting messages related to fraudulent activity and protecting users. The system is implemented through server, terminal, and user components.

[0654] The server collects data related to past fraudulent activities from existing databases and online resources to gather information. This data is used to train a generative AI model for the server to learn. This process uses natural language processing techniques and machine learning algorithms, and Python libraries (e.g., scikit-learn, TensorFlow) are used for data cleaning and feature engineering.

[0655] The generative AI model learns patterns specific to fraudulent activities and, after training, has the ability to generate pattern information based on the characteristics of fraudulent messages. This pattern information includes characteristic keywords and phrases common to fraudulent messages, which are used to compare with received data. The server has the capability to scan incoming messages in real time based on this pattern information.

[0656] The terminal automatically sends messages received by the user to the server. This is done through dedicated client software and is performed every time the user receives a new message. The server analyzes the received messages to check for potential fraud. If the message is identified as fraudulent, the server sends a warning to the user to protect them from further harm.

[0657] For example, if a user receives a message saying, "Please update your account information," the server will identify this as a phishing message and issue a warning immediately. Examples of prompts used in the system's AI model include:

[0658] "Please generate email subject lines that are easy to deceive."

[0659] In this way, users are protected from fraudulent activities and can safely use the content.

[0660] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0661] Step 1:

[0662] The server collects information. The input is data on fraudulent activities obtained from existing databases and online resources. The data includes examples of phishing emails and spam messages. The server collects this data in a secure manner and stores it in a database. This process provides foundational data for pattern analysis of fraudulent messages.

[0663] Step 2:

[0664] The server trains a generative AI model using the collected data. The input is the dataset collected in step 1, and the output is the trained generative AI model. The server utilizes natural language processing techniques and machine learning algorithms, preprocessing the data using Python libraries (e.g., scikit-learn, TensorFlow) to train the AI ​​model. This generates a model that has learned the features of fraudulent messages.

[0665] Step 3:

[0666] The server generates pattern information from the trained model. The input is the AI ​​model from step 2, and the output is pattern information summarizing the characteristics of fraudulent activity. The server analyzes the features learned by the AI ​​model, extracts keywords and phrases common to fraudulent messages, and creates a pattern file. This pattern information will be used for future message evaluation.

[0667] Step 4:

[0668] When a user receives a new message, the terminal sends that message to the server. The input is the message received by the user, and the output is the message data sent to the server. The terminal uses dedicated client software, and the message is automatically forwarded to the server according to the settings of the email client, etc.

[0669] Step 5:

[0670] The server analyzes received messages in real time and compares them with the pattern information generated in step 3. The input is the message sent by the user and the pattern information, and the output is the result of identifying potentially fraudulent messages. The server determines whether keywords and phrases in the message match the pattern information and identifies them if they are likely to be fraudulent.

[0671] Step 6:

[0672] If the server detects a fraudulent message, it notifies the user of the warning. The input is the fraudulent message identified in step 5, and the output is the warning message to the user. The server strengthens security by warning the user not to view details about potentially dangerous messages. Specifically, a notification is sent to the user's device, and measures are taken to prevent fraudulent activity.

[0673] (Application Example 1)

[0674] 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".

[0675] In today's information society, the damage caused by fraudulent information and fraudulent activities is increasing, making it a crucial issue to ensure user information security. However, conventional methods make it difficult to quickly and accurately detect fraudulent activities and effectively protect users. Therefore, there is a need for technology that can automatically detect fraudulent information in real time and respond quickly.

[0676] 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.

[0677] In this invention, the server includes means for collecting information patterns related to fraudulent activity, means for training a generative model that learns the characteristics of fraudulent activity, and means for generating specific pattern files. This enables immediate detection of fraudulent information and notification of warnings to users.

[0678] "Fraudulent activity" refers to unethical actions taken with the intent of deceiving others to unjustly obtain money or information.

[0679] An "information pattern" is a structure or form that exhibits common characteristics or tendencies in a particular type of information.

[0680] A "generative model" is a machine learning model that learns specific patterns or features from data and is used to predict or generate new data.

[0681] A "specific pattern file" is a data file used for detecting fraudulent activity, constructed based on the characteristics of fraudulent behavior learned by a generative model.

[0682] "Real-time" means responding to or processing an event immediately at the moment it occurs.

[0683] "Notification" is the process of informing recipients of changes or events in information.

[0684] "Restricting" means inhibiting certain actions or access to prevent the influx or use of undesirable information.

[0685] The system for implementing this invention primarily revolves around interaction between a server, a terminal, and the user. The server first collects information patterns related to fraudulent activity from around the world. This includes using existing databases and collecting samples of newly reported fraud cases. The server uses the collected information patterns to train a generative AI model and learn the characteristics of fraudulent activity. Once training is complete, the server generates specific pattern files based on the characteristics of fraudulent activity. This builds the foundational data for detecting fraudulent activity.

[0686] The terminal is responsible for forwarding messages received by the user to the server in real time. The received messages are immediately compared against a specific pattern file on the server. This comparison checks, for example, whether the message subject contains a phrase such as "urgent action required." If a pattern match is found, the server detects the message as malicious information.

[0687] Users will receive a warning via their device regarding malicious information detected by the server. The warning will include information about the potential danger of the message and a request not to view the details. The server will also take additional safety measures, such as restricting the reception of such messages, as needed.

[0688] For example, if a user receives a message from a travel booking service stating, "There is a problem with your account. Immediate action is required," the server will immediately identify this message as fraudulent information and issue an appropriate warning to the user.

[0689] An example of a prompt message would be, "I would like to learn about the common characteristics of scam messages. Please tell me about past patterns in scam emails."

[0690] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0691] Step 1:

[0692] The server collects information patterns related to fraudulent activities from a database. This includes past fraud messages and reported cases. By feeding relevant data as input and extracting information patterns, it obtains a dataset for analysis as output.

[0693] Step 2:

[0694] The server trains a generative AI model using the collected dataset. By supplying information patterns as input and allowing the generative AI model to learn the characteristics of fraudulent behavior, the server obtains a trained model as output.

[0695] Step 3:

[0696] The server generates specific pattern files using a trained generative AI model. It uses the trained model as input and analyzes the characteristics of fraudulent activity to create identification pattern files as output.

[0697] Step 4:

[0698] The terminal sends new messages received by the user to the server. By supplying received messages as input and performing a real-time scan on the server, the output becomes the result of message analysis.

[0699] Step 5:

[0700] The server compares received messages against a specific pattern file to determine if they are potentially fraudulent. Using message data and a pattern file as input, the server performs a data matching process, and the output is the result of the fraud detection.

[0701] Step 6:

[0702] If the server determines that the information is fraudulent, it will notify the user via the terminal. The input is the result of the fraud detection, and the server generates a warning message and sends it to the user, resulting in a warning notification as output.

[0703] Step 7:

[0704] The server, if it determines a message is fraudulent, will restrict its reception. By using the fraud detection result as input and blocking messages, it enhances security as output.

[0705] 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.

[0706] This invention relates to a system that automatically detects messages related to fraudulent activity and more effectively protects users by recognizing their emotions. Specific embodiments are described below.

[0707] First, the server collects data on past fraudulent messages and analyzes their characteristics. Based on this analysis, it trains an AI model to generate patterns related to fraudulent activity from these messages. This enhances the ability to detect potential fraud in unknown messages in real time.

[0708] When a user receives a message, the device sends it to a server for fraud pattern matching. Simultaneously, the device is equipped with an emotion engine that tracks the user's reactions. The emotion engine analyzes the user's facial expressions and tone of voice while they are reading the message to infer their emotional state.

[0709] For example, if a user receives a message prompting them to click a suspicious link and displays a wary or anxious expression, the emotion engine sends that information to the server. The server analyzes the user's emotional response to the potentially fraudulent message and generates a personalized warning message, which is then sent to the device. This warning message provides practical safety advice tailored to the user's response.

[0710] Furthermore, the server records user sentiment data and uses it to improve the fraud detection model. This allows the system to learn appropriate responses for individual users and improve its accuracy.

[0711] Thus, in addition to the rapid identification and countermeasures against fraudulent activities, the present invention can further enhance user safety by providing personalized services that take user emotions into consideration.

[0712] The following describes the processing flow.

[0713] Step 1:

[0714] The server collects historical message data related to fraudulent activity from a wide range of data sources. This data includes fraudulent emails, phishing messages, and more.

[0715] Step 2:

[0716] The server analyzes the collected data to extract common patterns in fraudulent messages. This includes analyzing characteristic phrases and sender information.

[0717] Step 3:

[0718] The server trains a generative AI model based on the extracted patterns and generates pattern files to detect fraudulent activity.

[0719] Step 4:

[0720] When a user receives a message, the device scans it and sends it to the server in real time.

[0721] Step 5:

[0722] The server compares the received message against a pattern file to assess the likelihood of fraud. Based on this assessment, it determines whether or not the message is fraudulent.

[0723] Step 6:

[0724] The device is equipped with an emotion engine that recognizes the user's emotions. The emotion engine analyzes the user's facial expressions and voice when they check a message to infer their emotions.

[0725] Step 7:

[0726] If the emotion engine detects any anxiety or concerns from the user, that information is sent to the server and analyzed along with the fraud detection results.

[0727] Step 8:

[0728] The server generates and sends personalized warning messages to the terminal based on the user's emotions. These messages include specific safety advice.

[0729] Step 9:

[0730] The server records user sentiment data and uses it to improve the accuracy of the model. This continuous feedback improves the overall responsiveness of the system.

[0731] (Example 2)

[0732] 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".

[0733] In recent years, online fraud has increased, causing significant losses for individuals and organizations. Traditional methods are insufficient to effectively detect and prevent fraud in real time, and they fail to take into account user emotions and reactions. It is necessary to address this situation and provide a more secure communication environment.

[0734] 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.

[0735] In this invention, the server includes means for collecting information patterns related to fraudulent activities, means for training a generative model that learns the characteristics of fraudulent activities using the information patterns, and means for recording sentiment data and improving the detection model. This enables the detection of potential fraud in real time and individual responses that take into account the user's emotions.

[0736] "Fraud" refers to any malicious act or attempt to deceive another person and illegally obtain money or assets.

[0737] "Information patterns" refer to a set of patterns or forms that indicate characteristic data or behavioral tendencies associated with fraudulent activity.

[0738] A "generative model" refers to an artificial intelligence model used to learn specific features or patterns from data and then generate or detect those features based on new data.

[0739] An "identification file" is a data file containing characteristic patterns of fraudulent activity, providing a standard used for matching with received information.

[0740] "Emotional state" refers to the psychological or emotional reactions or attitudes that a user exhibits when acquiring information.

[0741] "Emotional data" refers to data that quantifies or digitizes information about a user's emotional state and is used for system learning and improvement.

[0742] This invention relates to a system that automatically detects messages related to fraudulent activity and provides protection while also considering the user's emotions. Specific embodiments thereof are described below.

[0743] The server first collects information patterns related to past fraudulent activities from the internet and other databases. Based on this, the server trains a generative AI model. The generative AI model uses the collected data to learn the characteristics of fraudulent messages and generates identification files to identify the likelihood of fraud in new messages.

[0744] When a user receives a message, the device immediately sends it to the server. The server compares the message against the aforementioned identification file and assesses the likelihood of fraud in real time. Simultaneously, the device is equipped with an emotion engine that analyzes the user's facial expressions and tone of voice in response to the message to infer the user's emotional state.

[0745] For example, if a user receives a message prompting them to click a suspicious link and displays a surprised or anxious expression, the emotion engine sends that data to the server. Based on this, the server analyzes the user's response to the potentially fraudulent message, generates an appropriate warning message, and sends it back to the device. This warning is customized to the user's current emotions and includes practical safety advice.

[0746] Furthermore, the server records the acquired user sentiment data and uses it to improve the generative AI model. This allows the system to provide more personalized responses over time, increasing the accuracy of user protection.

[0747] As a concrete example, a possible prompt message might be: "While online shopping, you received a message offering a high-priced item at a very low price. How should you handle this message?" This system allows users to communicate more safely and securely.

[0748] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0749] Step 1:

[0750] The server collects information patterns related to fraudulent activity. This process gathers information from multiple sources, including databases of fraudulent emails and unauthorized access history. Using a dataset of past fraudulent activities as input, the server outputs information patterns that characterize fraud.

[0751] Step 2:

[0752] The server trains a generative AI model using the acquired information patterns. The collected information patterns are used as input, and machine learning algorithms are applied to learn these patterns. The output is a trained generative AI model for identifying fraudulent activity. This model improves the ability to identify potential fraud even in new messages.

[0753] Step 3:

[0754] When a user receives a message, the terminal sends that message to the server. The input is the newly received message, which is then sent to the server. The output is the message, which is then used in the next matching step.

[0755] Step 4:

[0756] The server compares received messages against an identification file. The input is a message received from a user, which is compared to an already constructed identification file. A generative AI model is used to assess the likelihood of fraud, resulting in a judgment regarding the message's fraud risk as output.

[0757] Step 5:

[0758] The device uses an emotion engine to analyze the user's emotional state. As input, the emotion engine collects the user's facial expressions and tone of voice as they read messages. As output, the user's emotional state is estimated and sent to the server.

[0759] Step 6:

[0760] The server generates personalized warning messages based on fraud risk and emotional state. The server receives emotional data as input and combines it with fraud risk data to construct specific warning messages. The output is customized advice sent to the user.

[0761] Step 7:

[0762] The server records user sentiment data and feedback on warnings to improve the generative AI model. It receives personalized warning message responses and sentiment data as input, and uses this data to retrain and improve the model. As output, it provides a generative AI model with improved system adaptability and accuracy.

[0763] (Application Example 2)

[0764] 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".

[0765] Internet-based fraud is becoming increasingly sophisticated, rendering conventional message filtering and warning systems insufficient. Furthermore, issuing uniform warnings without considering the user's emotions or circumstances can lead to false alarms and user stress. This invention aims to solve these problems by providing a personalized warning system that not only detects fraudulent messages quickly and effectively but also takes into account the user's emotional response.

[0766] 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.

[0767] In this invention, the server includes means for collecting message patterns related to fraudulent activity; means for training a generative model that learns the characteristics of fraudulent activity using the message patterns; means for generating a pattern file based on the characteristics of fraudulent activity obtained from the trained generative model; means for detecting the possibility of fraud by comparing the received message with the pattern file; means for inferring the emotional state by analyzing the user's facial expressions and voice when receiving a fraudulent message; and means for generating and notifying the user of a personalized warning message based on the user's emotional state. This makes it possible to detect unknown fraudulent messages and to provide appropriate warnings that take the user's emotions into consideration.

[0768] A "message pattern associated with fraudulent activity" is a form of information that has a common structure and characteristics and is used to identify communication content that contains fraudulent activity.

[0769] "Training a generative model" is the process of improving a machine learning algorithm using message data related to fraudulent activity to enhance its ability to identify new fraudulent messages.

[0770] "Generating pattern files" is the process of aggregating fraud features extracted from trained models and establishing criteria for detecting specific fraudulent activities based on those features.

[0771] "Detecting potential fraud" means analyzing the received message and determining whether it contains elements that suggest fraudulent activity.

[0772] "Analyzing the user's facial expressions and voice to infer their emotional state" means analyzing the user's facial expressions and tone of voice when they receive a message to estimate their emotions and psychological state at that time.

[0773] "Generating and notifying users of personalized warning messages" means customizing the content of warnings against fraudulent messages according to the user's emotional state and informing the user of that content.

[0774] To implement the invention, it is necessary to build a system that combines fraudulent message detection with user sentiment analysis. This system mainly consists of two main components: a server and a terminal.

[0775] The server collects message patterns related to fraudulent activity and trains a generative model based on historical data. Specifically, it develops a generative AI model using TensorFlow and learns the characteristics of fraudulent activity. Furthermore, it generates pattern files using the obtained characteristics and establishes criteria for detecting potential fraud. At this time, prompt statements are utilized to efficiently train the AI ​​model.

[0776] The device sends messages received by the user to a server in real time to check for potential fraud. The device also features an emotion engine utilizing the Emotion SDK and OpenCV, which analyzes the user's facial expressions and voice to infer their emotional state. This emotional information is used to generate personalized warning messages.

[0777] For example, if a user receives a message stating, "Your account will be closed if you don't act now," and shows a worried expression, the device sends this information to the server. The server flags it as a scam message, generates a personalized alert based on the facial expression data, and sends this to the device.

[0778] Examples of specific prompt messages used to achieve this process include the following:

[0779] "Message received: 'Your bank account is scheduled to be frozen. Please click this link to confirm.' User reaction: Frowning, in a worried voice."

[0780] This structure allows the system to effectively detect unknown fraudulent messages while providing personalized security measures that take user emotions into consideration.

[0781] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0782] Step 1:

[0783] The server collects message patterns related to fraudulent activity from an external database. It takes a database of known fraudulent messages as input and outputs message data exhibiting fraudulent characteristics. This data is used to train machine learning algorithms.

[0784] Step 2:

[0785] The server trains a generative AI model using the collected message patterns. The message data with fraud features obtained in step 1 is used as input, and a trained model for fraud detection is generated as output. TensorFlow is used to train the model during this process.

[0786] Step 3:

[0787] The server generates a pattern file based on the fraud features obtained from the trained model. The input is the trained model obtained in step 2, and the output is a pattern file for detecting fraudulent activity. This enables rapid detection even for unknown messages.

[0788] Step 4:

[0789] The user's device sends received messages to the server. Messages received by the user are taken into the device as input, and these messages are uploaded to the server as output. The device uses a specific protocol for transmission.

[0790] Step 5:

[0791] The server scans received messages in real time using the aforementioned pattern file to detect potential fraud. The input is the message obtained in step 4, and the output is a judgment on whether the message is fraudulent or not. This process utilizes the inference capabilities of a trained model.

[0792] Step 6:

[0793] The device analyzes the user's facial expressions and voice using the Emotion SDK and OpenCV while they are viewing a message, and infers their emotional state. The input is the user's camera video and audio data, and the output is the emotion analysis result.

[0794] Step 7:

[0795] The server generates personalized warning messages based on sentiment analysis results and notifies the terminal. The input combines the fraud detection results from step 5 and the sentiment analysis results from step 6, creating a user-specific warning as output. This allows the user to receive warnings about fraud.

[0796] 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.

[0797] 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.

[0798] 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.

[0799] 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.

[0800] 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.

[0801] 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.

[0802] 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.

[0803] 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.

[0804] 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."

[0805] 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.

[0806] 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.

[0807] 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.

[0808] 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.

[0809] 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.

[0810] 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.

[0811] 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.

[0812] 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.

[0813] 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.

[0814] 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.

[0815] 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.

[0816] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[0817] The following is further disclosed regarding the embodiments described above.

[0818] (Claim 1)

[0819] Means for collecting message patterns related to fraudulent activity,

[0820] A means for training a generative model that learns the characteristics of fraudulent activity using the aforementioned message patterns,

[0821] A method for generating pattern files based on fraudulent activity characteristics obtained from a trained generative model,

[0822] A means for detecting the possibility of fraud by comparing the received message with the pattern file,

[0823] A means of notifying users of detected fraudulent messages,

[0824] A system that includes this.

[0825] (Claim 2)

[0826] The system according to claim 1, further comprising means for scanning the received messages in real time.

[0827] (Claim 3)

[0828] The system according to claim 1, further comprising means for blocking the receipt of the message in question when the possibility of the aforementioned fraudulent activity is detected.

[0829] "Example 1"

[0830] (Claim 1)

[0831] Means of collecting information,

[0832] A means for training a program that learns the characteristics of fraudulent activities using the aforementioned information,

[0833] A means of generating pattern information based on the characteristics of fraudulent activities obtained from a trained program,

[0834] A means for detecting the possibility of fraud by comparing the received data with the aforementioned pattern information,

[0835] A means of notifying users of detected fraud data,

[0836] A system that includes this.

[0837] (Claim 2)

[0838] The system according to claim 1, further comprising means for analyzing the received data in real time.

[0839] (Claim 3)

[0840] The system according to claim 1, further comprising means for blocking the reception of relevant data when the possibility of the aforementioned fraudulent activity is detected.

[0841] "Application Example 1"

[0842] (Claim 1)

[0843] Means for collecting information patterns related to fraudulent activities,

[0844] A means for training a generative model that learns the characteristics of fraudulent activity using the aforementioned information patterns,

[0845] A means of generating specific pattern files based on the characteristics of fraudulent activity obtained from a trained generative model,

[0846] A means for detecting the possibility of fraud by comparing the received information with the aforementioned specific pattern file,

[0847] A means of notifying users of detected fraudulent information,

[0848] A means of restricting the reception of information based on detected malicious information,

[0849] A system that includes this.

[0850] (Claim 2)

[0851] The system according to claim 1, further comprising means for immediately analyzing the received information.

[0852] (Claim 3)

[0853] The system according to claim 1, further comprising means for blocking the reception of relevant information when the possibility of the aforementioned fraudulent activity is detected.

[0854] "Example 2 of combining an emotion engine"

[0855] (Claim 1)

[0856] Means for collecting information patterns related to fraudulent activities,

[0857] A means for training a generative model that learns the characteristics of fraudulent activity using the aforementioned information patterns,

[0858] A means of generating an identification file based on fraudulent activity characteristics obtained from a trained generative model,

[0859] A means for detecting the possibility of fraud by comparing the received information with the aforementioned identification file,

[0860] A means of analyzing the user's emotional state in response to received information,

[0861] A means of notifying the user of detected fraud information with a warning tailored to their emotional state,

[0862] A means of recording emotion data and improving detection models,

[0863] A system that includes this.

[0864] (Claim 2)

[0865] The system according to claim 1, further comprising means for immediately analyzing the received information.

[0866] (Claim 3)

[0867] The system according to claim 1, further comprising means for blocking the reception of relevant information when the possibility of the aforementioned fraudulent activity is detected.

[0868] "Application example 2 when combining with an emotional engine"

[0869] (Claim 1)

[0870] Means for collecting message patterns related to fraudulent activity,

[0871] A means for training a generative model that learns the characteristics of fraudulent activity using the aforementioned message patterns,

[0872] A method for generating pattern files based on fraudulent activity characteristics obtained from a trained generative model,

[0873] A means for detecting the possibility of fraud by comparing the received message with the pattern file,

[0874] A method for analyzing a user's facial expressions and voice when receiving a fraudulent message to infer their emotional state,

[0875] A means for generating and notifying users of personalized warning messages based on their emotional state,

[0876] A system that includes this.

[0877] (Claim 2)

[0878] The system according to claim 1, which scans the received messages in real time and performs sentiment analysis of the user.

[0879] (Claim 3)

[0880] The system according to claim 1, which, upon detecting the possibility of the aforementioned fraudulent activity, blocks the receipt of the relevant message and provides an additional warning based on sentiment analysis. [Explanation of symbols]

[0881] 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. Means for collecting message patterns related to fraudulent activity, A means for training a generative model that learns the characteristics of fraudulent activity using the aforementioned message patterns, A method for generating pattern files based on the characteristics of fraudulent activities obtained from a trained generative model, A means for detecting the possibility of fraud by comparing the received message with the pattern file, A means of notifying users of detected fraudulent messages, A system that includes this.

2. The system according to claim 1, further comprising means for scanning the received messages in real time.

3. The system according to claim 1, further comprising means for blocking the receipt of the message in question when the possibility of the aforementioned fraudulent activity is detected.

Images