system

Abstract

We provide the system. [Solution] A machine learning method that recognizes patterns of fraudulent contracts by collecting and learning from past contract information and fraud cases, A data analysis means that analyzes data received during the contract process in real time using natural language processing and image recognition technologies, and evaluates the validity of the data, An anomaly detection means that compares the input information with learned fraud patterns, detects anomalies, and calculates the risk. A notification mechanism that assesses the possibility of fraud and automatically notifies the person in charge based on that possibility, A model improvement method that continuously improves the model based on feedback, A system that includes this.

Description

【Technical Field】 【0001】 The technology of the present disclosure relates to a system. 【Background Art】 【0002】 Patent Document 1 discloses a persona chatbot control method performed by at least one processor, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance as a response to the user utterance. 【Prior Art Documents】 【Patent Documents】 【0003】 【Patent Document 1】 Japanese Patent Application Laid-Open No. 2022-180282 【Summary of the Invention】 【Problems to be Solved by the Invention】 【0004】 In current communication services and financial services, economic losses due to illegal contracts are a major problem. Since the methods of illegal contracts are particularly sophisticated, it is difficult to detect them quickly and accurately by conventional manual means. In conventional systems, it is difficult to take preventive measures based on past data, and continuous model improvement for dealing with new illegal methods has not been carried out. In such a situation, there is a need to provide a system that can detect and prevent illegal contracts efficiently and with high accuracy. 【Means for Solving the Problems】 【0005】 This invention uses machine learning to recognize typical patterns of fraudulent contracts by collecting and learning from past contract data and fraud cases. This enables real-time analysis of data received during the contract process using natural language processing and image recognition technologies to evaluate the data's legitimacy. Furthermore, the input information is compared with learned fraud patterns to detect anomalies and calculate risk. The system enables rapid response by automatically notifying responsible personnel of potential fraud, and provides a system that can respond to new fraudulent methods through continuous model improvement based on feedback. 【0006】 "Past contract information" refers to detailed information about contracts concluded to date in the field of telecommunications services and financial services, and includes the personal information and contract details of subscribers. 【0007】 "Fraudulent cases" refer to specific examples of fraudulent contracts that have actually occurred, and include information about their patterns and methods. 【0008】 "Machine learning methods" refer to processing methods implemented in computers using algorithms and models to learn data and recognize patterns. 【0009】 "Data analysis means" refers to a function that analyzes contract information received in real time using natural language processing and image recognition technologies, and evaluates its validity and consistency. 【0010】 An "anomaly detection method" refers to a function that detects anomalies or potential fraud by comparing input contract information with machine learning-generated fraud patterns. 【0011】 "Notification means" refers to a system that automatically communicates the detected potential for fraud to the responsible person in charge, prompting a swift response. 【0012】 "Model improvement means" refers to a function in a system that uses feedback to continuously improve machine learning models and enhance their accuracy. 【0013】 "Natural language processing" refers to the technology that enables computers to understand and process human language, and is used in the analysis of contract information. 【0014】 "Image recognition technology" refers to the technology used to analyze image data, extract information from it, and recognize it, and is used to evaluate the legitimacy of identity verification documents. 【0015】 "Real-time" refers to the time frame in which the entered information is analyzed immediately during the contract process and the results are output. [Brief explanation of the drawing] 【0016】 [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 the data processing device and 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]It is a sequence diagram showing the processing flow of the data processing system in Embodiment 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Embodiment 2 when combined with an emotion engine. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when combined with an emotion engine. 【Modes for Carrying Out the Invention】 【0017】 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. 【0018】 First, the terms used in the following description will be explained. 【0019】 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. 【0020】 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. 【0021】 In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes. 【0022】 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). 【0023】 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." 【0024】 [First Embodiment] 【0025】 Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment. 【0026】 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. 【0027】 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). 【0028】 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. 【0029】 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. 【0030】 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. 【0031】 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. 【0032】 Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14. 【0033】 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. 【0034】 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. 【0035】 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. 【0036】 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". 【0037】 The present invention is a system for effectively detecting and preventing fraudulent contracts, and embodiments thereof are shown below. 【0038】 Data collection and learning 【0039】 First, the server collects data on past contract information and fraud cases. This data includes basic information about the contract holder, contract details, and patterns of past fraud cases. This data is then used to train a machine learning model. For example, it learns cases where the addresses of multiple contract holders suddenly change as a potential pattern of fraud. 【0040】 Real-time data analysis 【0041】 During the new contract process, the terminal receives entered subscriber information and identification documents. This data is sent to a server, where natural language processing is used to analyze the subscriber's text information. For example, it verifies that the user's name and address match other publicly available information. Image recognition technology is also used to evaluate the authenticity of the submitted identification documents, checking for tampering or inconsistencies. 【0042】 Anomaly detection and risk assessment 【0043】 Contract information analyzed in real time is compared against previously learned fraud patterns. The server detects potentially fraudulent information during this process. For example, if the entered information resembles a pattern previously associated with fraud, a risk score is calculated. A high risk score indicates that an anomaly has been detected. 【0044】 Automated response and notifications 【0045】 If a transaction is deemed highly suspicious, an automated notification is sent from the server to the responsible party. This notification includes details of the potentially fraudulent contract and a suggestion to implement additional identity verification procedures. This allows the responsible party to take the necessary action quickly. 【0046】 Continuous model improvements 【0047】 The results of each contract process and the feedback obtained from subsequent actions are accumulated. The server analyzes this feedback and uses it to improve the model. In this way, the accuracy of the system improves, making it capable of handling new fraudulent methods. 【0048】 Through this series of processes, the present invention can function as a system that reduces the risk of fraudulent contracts and enables the provision of safer services to users. 【0049】 The following describes the processing flow. 【0050】 Step 1: 【0051】 The server collects historical contract information and fraud case data obtained from communication and financial services. This data includes subscriber information, contract details, and whether or not fraud occurred. Based on this data, the server extracts features to identify fraudulent contracts and uses them to train machine learning models. 【0052】 Step 2: 【0053】 The device receives contract information and images of identification documents entered by the user when a new contract is made. This information is necessary to verify the legitimacy of the contract and is sent to the next analysis step. 【0054】 Step 3: 【0055】 The server receives contract information sent from the terminal and analyzes the subscriber's text data using natural language processing. For example, it verifies whether the user's name, address, etc., are consistent. It also uses image recognition technology to verify the integrity and validity of identity verification documents and check for scanning or tampering. 【0056】 Step 4: 【0057】 The server compares the contract information obtained through analysis with pre-trained patterns of fraudulent contracts. Using a predictive model, it evaluates how similar this is to past fraud cases and calculates a risk score. If this score is high, it is determined that the input information is likely to be fraudulent. 【0058】 Step 5: 【0059】 The server automatically notifies the responsible party if it detects an anomaly based on the risk score. This notification includes detailed information about potentially fraudulent contracts and suggestions for additional verification steps. This allows the responsible party to respond quickly and appropriately. 【0060】 Step 6: 【0061】 The server stores the feedback and detection results obtained after all contract processes are completed. This information is used to continuously improve the machine learning model and update the system to enhance its accuracy. This continuous process ensures the system remains capable of responding to new fraudulent methods. 【0062】 (Example 1) 【0063】 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." 【0064】 Traditional contract management systems often had a problem of delayed detection of fraudulent contracts, resulting in damages caused by fraudulent activities. Furthermore, it was difficult to analyze information in real time at each stage of the contract process and accurately detect anomalies. Therefore, it was difficult to respond quickly to potential fraud and comprehensive fraud prevention was challenging. 【0065】 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. 【0066】 In this invention, the server includes machine learning means that collect and learn past contract information and fraud cases using a data storage device; information analysis means that analyze information received during new contract procedures in real time using text analysis and image processing technology and evaluate its legitimacy; and anomaly identification means that compare input data with learned fraud characteristics to identify anomalies and calculate the degree of risk. This enables rapid detection of fraudulent contracts and early notification to the responsible person. 【0067】 A "data storage device" is a device designed to securely store information such as contract details and fraud cases for extended periods, and to allow for quick access when needed. 【0068】 "Machine learning methods" refer to methods that use algorithms and processes to learn patterns based on collected data and identify the characteristics of fraudulent contracts. 【0069】 "Text analysis" is the process of analyzing text data using natural language processing techniques to understand its content and structure. 【0070】 "Image processing technology" refers to the techniques used to analyze digital images and extract and recognize information within them. 【0071】 "Information analysis means" refers to methods for evaluating received contract information in real time and checking the validity and consistency of the data. 【0072】 An "anomaly identification method" is a process that compares input data with existing fraudulent characteristics to quickly detect abnormal patterns and behaviors. 【0073】 "Risk level" is an indicator that numerically represents the likelihood of fraudulent activity occurring under specific conditions, and it indicates the risk level. 【0074】 "Notification methods" refer to methods or systems for informing relevant parties about the potential for misconduct and the need for action. 【0075】 "Model tuning" refers to the process of updating a machine learning model based on feedback to improve the system's accuracy and responsiveness. 【0076】 "Procedure" means a series of operations or processes related to a contract, including the process from data entry to verification and finalization. 【0077】 This invention is an advanced system for detecting and preventing fraudulent contracts. This system combines multiple solutions to monitor, detect, and quickly address fraudulent activity in the contract process in real time. 【0078】 The server collects past contract information and fraud cases via a data storage device and uses this to train a machine learning model. Specific software used includes Tensorflow® and PyTorch. The machine learning model learns fraud patterns and responds to fraud detection in real time. 【0079】 The server also uses natural language processing (NLP) techniques for text analysis. During the new contract process, it analyzes subscriber information transmitted from the terminal and evaluates the validity of the data. Specific software used includes NLP libraries (e.g., spaCy and NLTK). 【0080】 Regarding image recognition technology, the server processes images of identity verification documents sent from the terminal and evaluates their authenticity. Specific technologies used here include OpenCV and general cloud-based image recognition services (e.g., AWS® Rekognition). This analysis checks for document tampering and inconsistencies. 【0081】 When a user applies for a new contract, the server evaluates the entered information based on existing fraud patterns. If an anomaly is detected and a risk assessment is performed, the server automatically sends a notification to the responsible person using a notification system. This notification is sent via email or the company's internal messaging service. 【0082】 As a concrete example, in a bank loan application, if the information provided by the user resembles past fraud patterns, the system detects the anomaly and immediately sends an alert to the bank representative. An example of a prompt message to the generated AI model in this case would be, "Evaluate the new loan application information and check for similarities to past fraud patterns." 【0083】 This invention reduces the risk of fraudulent activity in the contract process and makes it possible to provide users with safer and more appropriate services. 【0084】 The flow of the specific processing in Example 1 will be explained using Figure 11. 【0085】 Step 1: 【0086】 The server collects past contract information and fraud cases from a database. This database includes basic information about the contractor, detailed contract terms, and past fraud patterns. Using the collected data, the server trains a machine learning model using frameworks such as TensorFlow or PyTorch. This results in a model that has learned the characteristics of fraudulent contracts. 【0087】 Step 2: 【0088】 When a user initiates a new contract, the device inputs subscriber information and identity verification documents and sends them to the server. This input includes name, address, contact information, and scanned images of identity verification documents. The server receives this information, analyzes the text data using NLP libraries such as spaCy or NLTK, and evaluates the accuracy of the information. The resulting analyzed text information is then output. 【0089】 Step 3: 【0090】 The server receives images of identity verification documents sent from the terminal and analyzes them using image recognition technologies such as OpenCV and AWS Rekognition. The server checks the integrity of the images and detects tampering, outputting an evaluation result of their legitimacy. This process confirms that the documents are genuine. 【0091】 Step 4: 【0092】 The server uses the results of text and image analysis to compare the input information with learned fraud patterns. By matching it against previously learned fraud patterns, it identifies anomalies and assesses the risk. The server then calculates a risk score and outputs the result. 【0093】 Step 5: 【0094】 Based on the detection results of an anomaly, the server automatically sends a notification to the responsible person if it determines that notification is necessary according to internal evaluation criteria. The notification content is conveyed via email or messaging system. This notification includes the specific nature of the anomaly and recommended additional actions. The notified information is output. 【0095】 Step 6: 【0096】 The final results and feedback from the contract process are stored in a database. Based on this feedback, the server adjusts and improves the machine learning model to address new fraudulent methods. As a result, an improved model is output. 【0097】 (Application Example 1) 【0098】 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." 【0099】 The problem that this invention aims to solve is to improve security and facilitate smooth procedures for legitimate use by detecting potential fraudulent activity in real time during online transactions and account registration, and warning users and related parties in advance. 【0100】 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. 【0101】 This invention includes a server comprising: machine learning means for recognizing patterns of fraudulent activity by collecting and learning from past transaction information and fraud cases; data analysis means for analyzing data received during the transaction process in real time using natural language processing and visual data recognition technologies and evaluating the validity of the data; and information provision means for warning of potential fraud in real time through the user's digital device. This makes it possible to quickly and automatically assess the risk of fraudulent activity and notify users and stakeholders in advance. 【0102】 "Transaction information" refers to data related to past commercial activities or negotiations, including details such as date, time, conditions, and participants. 【0103】 "Fraudulent activity cases" are actual examples of fraudulent acts that have occurred in the past, and are used as data to learn patterns of fraud and deception. 【0104】 "Machine learning" is a technology in which computers automatically learn patterns from large amounts of data to make future predictions and classifications. 【0105】 "Natural language processing" is a technology that enables computers to understand, analyze, and generate human language. 【0106】 "Visual data recognition technology" is a technology that analyzes visual data such as images and videos and extracts information from it. 【0107】 "Data analysis means" refers to technologies and methods that analyze received data in real time to evaluate its validity and detect anomalies. 【0108】 "Information provision means" refers to methods and technologies for providing users with real-time warnings and guidance regarding potential fraud. 【0109】 A "warning" is a notification that prompts users to take action by informing them in advance of potential risks or fraudulent activities. 【0110】 In the system that implements this application, the server first collects past transaction information and fraud cases, and uses this to train a machine learning model. The server leverages machine learning libraries such as TensorFlow and PyTorch to learn patterns of fraudulent activity from large amounts of data. This process creates a model that can recognize potential patterns of fraud. 【0111】 Next, when a user conducts a transaction, the device (such as a smartphone) collects the transaction information and identity verification data entered by the user. The device sends the entered data to the server, which processes the received data in real time. At this stage, text data is analyzed using natural language processing tools such as Hugging Face Transformers, and the validity of visual data is verified using image recognition libraries such as OpenCV. 【0112】 If the server analyzes transaction information and the data entered by the user resembles a learned fraud pattern, the server performs a risk assessment. If an anomaly is detected, the user is warned in real time through the terminal's information provision system. This notification allows the user to immediately review their input information. In short, the system provides a mechanism to quickly and effectively prevent fraudulent activity. 【0113】 For example, when a user creates a new account while shopping online, if the name and address information they enter matches past fraud cases identified on the server, a warning will be displayed on their smartphone, and the user will be prompted to verify the information. 【0114】 Examples of prompt statements when using a generative AI model are as follows: 【0115】 "Please enter your username, address, and identification documents to assess the risk of fraudulent contracts." 【0116】 Based on this prompt, the AI ​​model evaluates the security of contracts and transactions, helping to proactively prevent fraud risks. 【0117】 The flow of a specific process in Application Example 1 will be explained using Figure 12. 【0118】 Step 1: 【0119】 The user initiates a transaction and enters the necessary information into the terminal. This information includes name, address, payment information, and scanned images of identification documents. The terminal then prepares to send this data to the server. 【0120】 Step 2: 【0121】 The server receives input data sent from the terminal. After receiving the data, it uses natural language processing tools (such as Hugging Face Transformers) to analyze the text data and check for any information similar to malicious patterns. The processed data is then used in the next processing step. 【0122】 Step 3: 【0123】 The server uses visual data recognition technology (such as OpenCV) to analyze the scanned images of identification documents. This process verifies that the images have not been tampered with and that the documents are authentic. The legitimacy of the processed image data is evaluated, and the results are sent to the next step. 【0124】 Step 4: 【0125】 The server analyzes the data and compares it to pre-trained fraud patterns to detect anomalies. A risk score is then calculated for the input data. If the risk score is high, it is marked as an anomaly, and a warning is deemed necessary. 【0126】 Step 5: 【0127】 If an anomaly is detected, the server will send a warning message to the terminal in real time. This message will contain details of the information that needs to be checked. Upon receiving this notification, the user can review the information they entered and make corrections or reconfirmations as necessary. 【0128】 Step 6: 【0129】 After all transactions are completed, the execution results and feedback are saved to the server. This data helps identify new fraudulent patterns and areas for improvement, which in turn helps train the generative AI model. As a result, the accuracy of the system will improve in future transactions. 【0130】 In this way, users and servers can work together to provide a secure trading environment while establishing protection against fraudulent activity. 【0131】 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. 【0132】 This invention provides a comprehensive judgment capability that takes into account the user's emotional state by combining an emotion engine with a system that effectively detects and prevents fraudulent contracts. An embodiment thereof is shown below. 【0133】 Data collection and learning 【0134】 First, the server collects historical contract data and past fraud cases. This dataset includes detailed information about contractors, contract terms, and cases marked as fraudulent. Machine learning algorithms are used to learn patterns of fraudulent contracts based on this data. This allows the system to recognize existing fraud patterns and identify risks in new contracts. 【0135】 Real-time data analysis and emotion recognition 【0136】 During the contract process, the terminal not only receives contract information and identity verification documents provided by the user, but also captures the user's intent and emotions through voice or text. This information is transmitted to a server and analyzed by a natural language processing engine. Image recognition technology is also used to verify the authenticity of the submitted documents. 【0137】 Furthermore, an emotion engine analyzes the user's emotional state. This engine analyzes the tone of voice, word choice, and conversational context to recognize the user's emotions in real time. For example, if a user is clearly experiencing stress during a contract, that information could influence the risk assessment. 【0138】 Anomaly detection and notification 【0139】 Based on the analyzed information, the server compares it to previously learned fraud cases and calculates an anomaly risk score for the contract. During this process, user sentiment analysis results are also integrated, and particular attention is paid to any unusual patterns. If the risk is high, the system detects the anomaly and sends a notification to alert the responsible party. 【0140】 Improvement and response 【0141】 In situations deemed high-risk, the server automatically notifies the responsible party of the details and countermeasures. For example, it provides specific suggestions if it determines that additional information should be requested from the user or if a different identity verification procedure is necessary. 【0142】 Continuous model improvement 【0143】 The server collects all feedback from the contract process and uses it to improve the overall model, including the sentiment engine. Newly emerging fraud methods and data on user sentiment improve the model's accuracy and responsiveness. This enables the system to perform more comprehensive fraud detection and prevention, thereby enhancing the security of the service. 【0144】 In this way, by incorporating emotion recognition, the present invention functions as a system that, in addition to detecting fraudulent contracts, improves the user experience and enables the realization of more accurate services. 【0145】 The following describes the processing flow. 【0146】 Step 1: 【0147】 The server collects past contract information and fraud cases from a database, inputs them into a machine learning algorithm, and begins analysis. This builds a dataset that learns common patterns and features in fraudulent contracts. 【0148】 Step 2: 【0149】 The device receives contract information and identity verification documents provided by the user at the start of the contract process. At this time, the user's spoken voice and entered text data are also captured and used for emotion recognition. 【0150】 Step 3: 【0151】 The server processes contract information sent from the terminal in real time. A natural language processing engine analyzes the text data and formats information such as the contract holder's name and address. It also uses image recognition technology to verify whether the submitted identification documents are genuine. 【0152】 Step 4: 【0153】 The server operates an emotion engine that recognizes emotions from the user's voice or written words. It analyzes the user's tone of voice, word choice, and speech rhythm to understand the user's emotional state. This information is also incorporated into the risk assessment of the entire contract process. 【0154】 Step 5: 【0155】 The server compares the analyzed contract information and sentiment recognition results with previously learned fraud patterns. This comparison calculates a risk score to assess the likelihood that the currently ongoing contract is fraudulent. 【0156】 Step 6: 【0157】 If the server exceeds a certain risk score, it determines that there is a high probability of fraud and automatically sends a notification to the responsible party. This notification includes additional recommended actions based on the user's contract details, detected risks, and emotional state. 【0158】 Step 7: 【0159】 The server collects all feedback after the contract process is complete. This data, including new fraudulent practices and user sentiment, is used to improve the machine learning models and sentiment engine. This process enhances the overall accuracy and reliability of the system. 【0160】 (Example 2) 【0161】 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 will be referred to as the "terminal." 【0162】 In modern contract processes, the risk of fraudulent contracts is increasing, and existing detection methods based on existing technologies are insufficient to prevent them. Furthermore, there is a lack of technology that considers the emotional state of users when determining fraud risk, thus necessitating a more comprehensive contract management approach. Additionally, there is a need for technology that analyzes the emotional state of users during contract procedures and utilizes that information to improve the accuracy of services. 【0163】 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. 【0164】 This invention includes a server comprising: a machine learning means that recognizes patterns of contracts that violate rules by collecting and learning information on past contracts and fraud cases; a data analysis means that analyzes data received during the contract procedure in real time using natural language processing and image recognition technology and evaluates the validity of the information; and an emotion recognition means that analyzes the emotional state from the user's voice and text and reflects that emotional information in the contract risk assessment. This enables comprehensive anomaly detection that takes into account the user's emotional information, which was lacking in conventional fraudulent contract detection technology, thereby reducing fraud risk and improving the accuracy of the service. 【0165】 "Machine learning methods" are technologies that collect information on past contracts and cases of fraud, and use that information to recognize contract patterns that violate regulations. 【0166】 "Data analysis means" refers to technologies that analyze data received during contract procedures in real time using natural language processing and image recognition technologies to evaluate the validity of the information. 【0167】 "Emotion recognition means" refers to technology that analyzes a user's emotional state from their voice or text data and incorporates that emotional information into the risk assessment of a contract. 【0168】 An "anomaly detection method" is a technology that detects anomalies and calculates risk by comparing the analyzed information with learned fraud patterns. 【0169】 A "notification method" is a technology that assesses the possibility of a rule violation and automatically notifies the responsible person based on that possibility. 【0170】 "Model improvement methods" refer to techniques for continuously improving the overall model based on the feedback received. 【0171】 This invention provides a system for effectively detecting and preventing fraudulent contracts, enabling comprehensive judgment that takes into account the user's emotional state. By combining machine learning and sentiment analysis technologies, this system aims to reduce the risk of fraud at each stage of the contract process while improving the user experience. 【0172】 The server collects historical contract information and data on fraudulent cases, and uses this data to train machine learning models. Specifically, it uses machine learning libraries such as TensorFlow and PyTorch to train the models. The models are used to recognize fraudulent contract patterns and identify risks in new contracts. 【0173】 The terminal receives contract information and identity verification documents provided by the user during the contract process. The terminal digitizes the documents using optical character recognition (OCR) technology and converts the user's voice into text using a speech recognition system. This information is transmitted to the server in real time and analyzed by a natural language processing engine. 【0174】 Based on the analyzed audio and text data, the server uses an emotion engine to evaluate the user's emotional state. Specifically, it analyzes the tone of voice, word choice, and conversational context to determine what emotions the user is experiencing. The Hugging Face emotion analysis model is used for this evaluation. 【0175】 This system further compares the analyzed information against patterns of fraudulent contracts, detects anomalies, and calculates a risk score. Scikit-learn's anomaly detection algorithm is used for this purpose. If a high risk is determined, the system sends a notification to the responsible party and suggests additional countermeasures. 【0176】 A concrete example is a mortgage application. When a user attempts to sign a mortgage contract, the terminal detects stress from the user's voice. The server evaluates the detected stress level as high risk and sends a notification to the person in charge stating, "The user may be experiencing stress; please consider additional identity verification procedures." In this case, a prompt such as, "Suggest countermeasures to take when a specific emotional state (e.g., anxiety, stress) is detected during the mortgage contract review process," is used as an example input to the generating AI model. 【0177】 By implementing the invention in this way, it is possible to simultaneously achieve improved fraud detection capabilities and enhanced user experience. 【0178】 The flow of the specific processing in Example 2 will be explained using Figure 13. 【0179】 Step 1: 【0180】 The server collects historical contract information and fraud cases from the contract database. The input consists of the contractor's personal information, contract details, and records of fraud cases. The server executes SQL queries to extract this information as a dataset. The output is a fraudulent contract dataset in an analyzable format. 【0181】 Step 2: 【0182】 The server trains a machine learning model using the collected dataset. The input is the fraudulent contract dataset obtained in Step 1. Specifically, the server uses TensorFlow or PyTorch to preprocess the data, select features, and train the model. The output is a trained model that recognizes fraudulent contract patterns. 【0183】 Step 3: 【0184】 The terminal receives contract information, voice recordings, and identification documents entered by the user during the contract process. The input consists of this information provided by the user. The terminal uses OCR technology to digitize documents and applies speech recognition technology to convert speech to text. The output is data in a digital format suitable for processing on the server. 【0185】 Step 4: 【0186】 The server receives digital data transmitted from the terminal and analyzes its content using a natural language processing engine. The input is the text data processed in step 3. The server analyzes the text content and extracts information to determine the legitimacy and potential fraud of the contract. The output is a dataset containing the analysis results. 【0187】 Step 5: 【0188】 The server analyzes the user's emotional state using an emotion engine. The input is the audio and text data obtained in step 3. Specifically, it uses the Hugging Face emotion analysis model to analyze voice tone, word choice, and context. The output is data indicating the user's emotional state. 【0189】 Step 6: 【0190】 The server detects anomalies by comparing the analyzed contract data and sentiment state with a trained model and calculates a risk score. The inputs are the analyzed data from step 4 and the sentiment data from step 5. The server uses Scikit-learn to evaluate the similarity to known fraudulent contract patterns and quantifies the risk of the contract. The output is the risk score. 【0191】 Step 7: 【0192】 The server notifies the responsible person based on the risk score. The input is the risk score obtained in step 6. Notifications are sent via email or the company's internal messaging system, and the responsible person is notified if additional verification work is required based on the risk. The output is the notification message sent to the responsible person. 【0193】 Step 8: 【0194】 The server collects feedback obtained after the contract process and uses it to improve the model. Inputs are actual contract results and user feedback obtained from the entire system. The server incorporates new fraud patterns and sentiment data into the model to improve the system's adaptability. The output is the updated, trained model. 【0195】 (Application Example 2) 【0196】 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". 【0197】 In modern electronic transactions, fraudulent transactions are on the rise, posing a significant security challenge. Furthermore, conventional fraud detection systems often focus solely on transaction patterns and fail to consider the user's emotional state, resulting in inaccurate detection. Therefore, the objective of this invention is to improve the accuracy of fraud detection and realize a safer trading environment. 【0198】 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. 【0199】 In this invention, the server includes machine learning means for collecting and recognizing past transaction information and fraudulent activity cases; data analysis means for analyzing data received during transaction procedures in real time using natural language processing and image recognition technologies and evaluating the legitimacy of the data; and emotion recognition means for analyzing user voice data and evaluating emotional state. This enables highly accurate detection of fraudulent transactions and enhanced security that takes user experience into consideration. 【0200】 "Past transaction information" refers to data on previously conducted transactions that are collected for the purpose of detecting fraudulent activity. 【0201】 "Examples of fraudulent activity" refers to data that shows specific examples of fraudulent transactions or actions that have occurred in the past. 【0202】 "Machine learning methods" are algorithms and techniques that recognize patterns based on data, enabling prediction and classification. 【0203】 "Natural language processing methods" are technologies used to analyze text and speech and understand their meaning and context. 【0204】 "Image recognition technology" is a technology that analyzes image data to recognize objects and characters, and to extract features. 【0205】 "Data analysis methods" refer to techniques and systems for processing various types of data using statistics and algorithms to gain insights. 【0206】 "Emotion recognition methods" are technologies for inferring and evaluating human emotions from things like voice and facial expressions. 【0207】 An "anomaly detection method" is a system for detecting patterns or behaviors that deviate from normal and for evaluating the associated risks. 【0208】 A "notification mechanism" refers to a function or device used by the system to inform relevant parties of anomalies or important information it has detected. 【0209】 "Model improvement methods" are techniques for continuously improving machine learning models based on new data and feedback, thereby enhancing their accuracy. 【0210】 The system that implements this application is server-centric and consists of multiple means with various roles. The server first collects past transaction information and fraudulent activity cases, and uses machine learning to recognize patterns of fraudulent transactions based on this data. During the transaction process, the server analyzes the data received from the terminal in real time using natural language processing and image recognition technologies to evaluate the legitimacy of the data. 【0211】 Furthermore, the terminal acquires the user's voice data, and an emotion recognition system evaluates the user's emotional state. This allows the server to compare the input information with learned fraud patterns, detect anomalies, and calculate the risk. If the calculated risk exceeds a certain level, the server immediately and automatically notifies the responsible person through a notification system. 【0212】 The hardware used will primarily consist of devices such as smartphones and tablets, while the software will utilize Google Cloud's Natural Language API and various speech analysis libraries. 【0213】 As a concrete example, the terminal receives audio from the user during payment, the server analyzes the audio data to read the emotion, and if it determines that there is a high probability of fraud, an alert is immediately sent to the person in charge. An example of a prompt sentence to be input into the generating AI model is: "Explain how to proactively detect fraud by analyzing the emotion from the user's voice and context during smartphone payments." 【0214】 The flow of a specific process in Application Example 2 will be explained using Figure 14. 【0215】 Step 1: 【0216】 The server collects past transaction information and fraud cases from a database. Using this data as input, it learns patterns of fraudulent transactions using a machine learning algorithm and outputs the results as a model. This model becomes the new standard for evaluating transaction data. 【0217】 Step 2: 【0218】 The terminal receives data entered by the user during the transaction process. This input data includes voice data, text data, and image data. The terminal then sends this data to the server. 【0219】 Step 3: 【0220】 The server uses natural language processing techniques to analyze the meaning of the received audio and text data. The audio data is converted to text using a speech analysis library, and based on the output, an emotion recognition system evaluates the user's emotional state. This result serves as a basis for determining the user's emotions during the transaction. 【0221】 Step 4: 【0222】 The server analyzes the received image data using image recognition technology to verify the authenticity and integrity of the submitted identity verification documents. Based on the results, it determines whether the data has been forged or tampered with and outputs information to assess the risk of fraud. 【0223】 Step 5: 【0224】 The server compares pre-processed transaction data with historical database data based on processed sentiment data and image analysis results, and uses an anomaly detection algorithm to assess the likelihood of fraud. A risk score is calculated and output as a result. If the likelihood of fraud is high, the information is passed on to the next step. 【0225】 Step 6: 【0226】 Based on the calculated risk score, the server immediately notifies the responsible party of the risk using notification methods. The notification includes detailed information about the detected fraud and countermeasures. This notification facilitates a rapid response. 【0227】 The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data. 【0228】 Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization. 【0229】 In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14. 【0230】 [Second Embodiment] 【0231】 Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment. 【0232】 As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server. 【0233】 The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network). 【0234】 The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52. 【0235】 The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46. 【0236】 Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision). 【0237】 Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner. 【0238】 Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56. 【0239】 The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30. 【0240】 The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290. 【0241】 In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48. 【0242】 Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal". 【0243】 The present invention is a system for effectively detecting and preventing fraudulent contracts, and embodiments thereof are shown below. 【0244】 Data collection and learning 【0245】 First, the server collects data on past contract information and fraud cases. This data includes basic information about the contract holder, contract details, and patterns of past fraud cases. This data is then used to train a machine learning model. For example, it learns cases where the addresses of multiple contract holders suddenly change as a potential pattern of fraud. 【0246】 Real-time data analysis 【0247】 During the new contract process, the terminal receives entered subscriber information and identification documents. This data is sent to a server, where natural language processing is used to analyze the subscriber's text information. For example, it verifies that the user's name and address match other publicly available information. Image recognition technology is also used to evaluate the authenticity of the submitted identification documents, checking for tampering or inconsistencies. 【0248】 Anomaly detection and risk assessment 【0249】 Contract information analyzed in real time is compared against previously learned fraud patterns. The server detects potentially fraudulent information during this process. For example, if the entered information resembles a pattern previously associated with fraud, a risk score is calculated. A high risk score indicates that an anomaly has been detected. 【0250】 Automated response and notifications 【0251】 If a transaction is deemed highly suspicious, an automated notification is sent from the server to the responsible party. This notification includes details of the potentially fraudulent contract and a suggestion to implement additional identity verification procedures. This allows the responsible party to take the necessary action quickly. 【0252】 Continuous model improvements 【0253】 The results of each contract process and the feedback obtained from subsequent actions are accumulated. The server analyzes this feedback and uses it to improve the model. In this way, the accuracy of the system improves, making it capable of handling new fraudulent methods. 【0254】 Through this series of processes, the present invention can function as a system that reduces the risk of fraudulent contracts and enables the provision of safer services to users. 【0255】 The following describes the processing flow. 【0256】 Step 1: 【0257】 The server collects historical contract information and fraud case data obtained from communication and financial services. This data includes subscriber information, contract details, and whether or not fraud occurred. Based on this data, the server extracts features to identify fraudulent contracts and uses them to train machine learning models. 【0258】 Step 2: 【0259】 The device receives contract information and images of identification documents entered by the user when a new contract is made. This information is necessary to verify the legitimacy of the contract and is sent to the next analysis step. 【0260】 Step 3: 【0261】 The server receives contract information sent from the terminal and analyzes the subscriber's text data using natural language processing. For example, it verifies whether the user's name, address, etc., are consistent. It also uses image recognition technology to verify the integrity and validity of identity verification documents and check for scanning or tampering. 【0262】 Step 4: 【0263】 The server compares the contract information obtained through analysis with pre-trained patterns of fraudulent contracts. Using a predictive model, it evaluates how similar this is to past fraud cases and calculates a risk score. If this score is high, it is determined that the input information is likely to be fraudulent. 【0264】 Step 5: 【0265】 The server automatically notifies the responsible party if it detects an anomaly based on the risk score. This notification includes detailed information about potentially fraudulent contracts and suggestions for additional verification steps. This allows the responsible party to respond quickly and appropriately. 【0266】 Step 6: 【0267】 The server stores the feedback and detection results obtained after all contract processes are completed. This information is used to continuously improve the machine learning model and update the system to enhance its accuracy. This continuous process ensures the system remains capable of responding to new fraudulent methods. 【0268】 (Example 1) 【0269】 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." 【0270】 Traditional contract management systems often had a problem of delayed detection of fraudulent contracts, resulting in damages caused by fraudulent activities. Furthermore, it was difficult to analyze information in real time at each stage of the contract process and accurately detect anomalies. Therefore, it was difficult to respond quickly to potential fraud and comprehensive fraud prevention was challenging. 【0271】 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. 【0272】 In this invention, the server includes machine learning means that collect and learn past contract information and fraud cases using a data storage device; information analysis means that analyze information received during new contract procedures in real time using text analysis and image processing technology and evaluate its legitimacy; and anomaly identification means that compare input data with learned fraud characteristics to identify anomalies and calculate the degree of risk. This enables rapid detection of fraudulent contracts and early notification to the responsible person. 【0273】 A "data storage device" is a device designed to securely store information such as contract details and fraud cases for extended periods, and to allow for quick access when needed. 【0274】 "Machine learning methods" refer to methods that use algorithms and processes to learn patterns based on collected data and identify the characteristics of fraudulent contracts. 【0275】 "Text analysis" is the process of analyzing text data using natural language processing techniques to understand its content and structure. 【0276】 "Image processing technology" refers to the techniques used to analyze digital images and extract and recognize information within them. 【0277】 "Information analysis means" refers to methods for evaluating received contract information in real time and checking the validity and consistency of the data. 【0278】 An "anomaly identification method" is a process that compares input data with existing fraudulent characteristics to quickly detect abnormal patterns and behaviors. 【0279】 "Risk level" is an indicator that numerically represents the likelihood of fraudulent activity occurring under specific conditions, and it indicates the risk level. 【0280】 "Notification methods" refer to methods or systems for informing relevant parties about the potential for misconduct and the need for action. 【0281】 "Model tuning" refers to the process of updating a machine learning model based on feedback to improve the system's accuracy and responsiveness. 【0282】 "Procedure" means a series of operations or processes related to a contract, including the process from data entry to verification and finalization. 【0283】 This invention is an advanced system for detecting and preventing fraudulent contracts. This system combines multiple solutions to monitor, detect, and quickly address fraudulent behavior in the contract process in real time. 【0284】 The server collects past contract information and fraud cases via a data storage device and trains a machine learning model based on this. Specific software used includes TensorFlow, PyTorch, etc. The machine learning model learns fraud patterns and responds to real-time fraud detection. 【0285】 The server also uses natural language processing technology for text analysis. During the new contract procedure, it analyzes the contractor information sent from the terminal and evaluates the validity of the data. As specific software, NLP libraries (e.g., spaCy, NLTK) are used. 【0286】 Regarding image recognition technology, the server processes the image of the identity document sent from the terminal and evaluates its validity. Specific technologies used here include OpenCV and general cloud-based image recognition services (e.g., AWS Rekognition). This analysis checks for document forgery and inconsistencies. 【0287】 When a user applies for a new contract, the server evaluates the input information based on existing fraud patterns. If an anomaly is detected and a risk assessment is performed, the server automatically sends a notification to the responsible person using notification means. This notification is sent using email or an in-house messaging service. 【0288】 As a specific example, in a loan contract at a bank, if the information provided by the user is similar to past fraud patterns, the system detects the anomaly and immediately sends an alert to the bank's responsible person. An example of the prompt text for the generative AI model at this time is "Evaluate the application information for a new loan and check for similarity with past fraud patterns." 【0289】 This invention reduces the risk of fraudulent activity in the contract process and makes it possible to provide users with safer and more appropriate services. 【0290】 The flow of the specific processing in Example 1 will be explained using Figure 11. 【0291】 Step 1: 【0292】 The server collects past contract information and fraud cases from a database. This database includes basic information about the contractor, detailed contract terms, and past fraud patterns. Using the collected data, the server trains a machine learning model using frameworks such as TensorFlow or PyTorch. This results in a model that has learned the characteristics of fraudulent contracts. 【0293】 Step 2: 【0294】 When a user initiates a new contract, the device inputs subscriber information and identity verification documents and sends them to the server. This input includes name, address, contact information, and scanned images of identity verification documents. The server receives this information, analyzes the text data using NLP libraries such as spaCy or NLTK, and evaluates the accuracy of the information. The resulting analyzed text information is then output. 【0295】 Step 3: 【0296】 The server receives images of identity verification documents sent from the terminal and analyzes them using image recognition technologies such as OpenCV and AWS Rekognition. The server checks the integrity of the images and detects tampering, outputting an evaluation result of their legitimacy. This process confirms that the documents are genuine. 【0297】 Step 4: 【0298】 The server uses the results of text and image analysis to compare the input information with learned fraud patterns. By matching it against previously learned fraud patterns, it identifies anomalies and assesses the risk. The server then calculates a risk score and outputs the result. 【0299】 Step 5: 【0300】 Based on the detection results of an anomaly, the server automatically sends a notification to the responsible person if it determines that notification is necessary according to internal evaluation criteria. The notification content is conveyed via email or messaging system. This notification includes the specific nature of the anomaly and recommended additional actions. The notified information is output. 【0301】 Step 6: 【0302】 The final results and feedback from the contract process are stored in a database. Based on this feedback, the server adjusts and improves the machine learning model to address new fraudulent methods. As a result, an improved model is output. 【0303】 (Application Example 1) 【0304】 Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal." 【0305】 The problem that this invention aims to solve is to improve security and facilitate smooth procedures for legitimate use by detecting potential fraudulent activity in real time during online transactions and account registration, and warning users and related parties in advance. 【0306】 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. 【0307】 In this invention, the server includes a machine learning means for recognizing patterns of fraudulent behavior by collecting and learning past transaction information and fraud cases, a data analysis means for analyzing in real time the data received during the transaction process using natural language processing and visual data recognition technologies to evaluate the validity of the data, and an information providing means for warning in real time of the possibility of fraud through the user's digital device. As a result, it becomes possible to quickly and automatically evaluate the risk of fraudulent behavior and notify users and interested parties in advance. 【0308】 "Transaction information" refers to data related to past commercial activities and business negotiations, including details such as date and time, conditions, and participants. 【0309】 "Fraud case" refers to an example of past fraudulent behavior and is used as data for learning patterns of fraud and deception. 【0310】 "Machine learning" is a technology in which a computer automatically learns patterns from a large amount of data and makes predictions and classifications for the future. 【0311】 "Natural language processing" is a technology in which a computer understands, analyzes, and generates human language. 【0312】 "Visual data recognition technology" is a technology for analyzing visual data such as images and videos and extracting information therefrom. 【0313】 "Data analysis means" refers to technologies and methods for analyzing the received data in real time to evaluate validity and detect anomalies. 【0314】 "Information providing means" refers to methods and technologies for providing users with warnings and guidance regarding the possibility of fraud in real time. 【0315】 "Warning" is a notice for prompting countermeasures by notifying users in advance of potential risks and the possibility of fraud. 【0316】 In the system that implements this application, the server first collects past transaction information and fraud cases, and uses this to train a machine learning model. The server leverages machine learning libraries such as TensorFlow and PyTorch to learn patterns of fraudulent activity from large amounts of data. This process creates a model that can recognize potential patterns of fraud. 【0317】 Next, when a user conducts a transaction, the device (such as a smartphone) collects the transaction information and identity verification data entered by the user. The device sends the entered data to the server, which processes the received data in real time. At this stage, text data is analyzed using natural language processing tools such as Hugging Face Transformers, and the validity of visual data is verified using image recognition libraries such as OpenCV. 【0318】 If the server analyzes transaction information and the data entered by the user resembles a learned fraud pattern, the server performs a risk assessment. If an anomaly is detected, the user is warned in real time through the terminal's information provision system. This notification allows the user to immediately review their input information. In short, the system provides a mechanism to quickly and effectively prevent fraudulent activity. 【0319】 For example, when a user creates a new account while shopping online, if the name and address information they enter matches past fraud cases identified on the server, a warning will be displayed on their smartphone, and the user will be prompted to verify the information. 【0320】 Examples of prompt statements when using a generative AI model are as follows: 【0321】 "Please enter your username, address, and identification documents to assess the risk of fraudulent contracts." 【0322】 Based on this prompt, the AI ​​model evaluates the security of contracts and transactions, helping to proactively prevent fraud risks. 【0323】 The flow of a specific process in Application Example 1 will be explained using Figure 12. 【0324】 Step 1: 【0325】 The user initiates a transaction and enters the necessary information into the terminal. This information includes name, address, payment information, and scanned images of identification documents. The terminal then prepares to send this data to the server. 【0326】 Step 2: 【0327】 The server receives input data sent from the terminal. After receiving the data, it uses natural language processing tools (such as Hugging Face Transformers) to analyze the text data and check for any information similar to malicious patterns. The processed data is then used in the next processing step. 【0328】 Step 3: 【0329】 The server uses visual data recognition technology (such as OpenCV) to analyze the scanned images of identification documents. This process verifies that the images have not been tampered with and that the documents are authentic. The legitimacy of the processed image data is evaluated, and the results are sent to the next step. 【0330】 Step 4: 【0331】 The server analyzes the data and compares it to pre-trained fraud patterns to detect anomalies. A risk score is then calculated for the input data. If the risk score is high, it is marked as an anomaly, and a warning is deemed necessary. 【0332】 Step 5: 【0333】 If an anomaly is detected, the server will send a warning message to the terminal in real time. This message will contain details of the information that needs to be checked. Upon receiving this notification, the user can review the information they entered and make corrections or reconfirmations as needed. 【0334】 Step 6: 【0335】 After all transactions are completed, the execution results and feedback are saved to the server. This data helps identify new fraudulent patterns and areas for improvement, which in turn helps train the generative AI model. As a result, the accuracy of the system will improve in future transactions. 【0336】 In this way, users and servers can work together to provide a secure trading environment while establishing protection against fraudulent activity. 【0337】 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. 【0338】 This invention provides a comprehensive judgment capability that takes into account the user's emotional state by combining an emotion engine with a system that effectively detects and prevents fraudulent contracts. An embodiment thereof is shown below. 【0339】 Data collection and learning 【0340】 First, the server collects historical contract data and past fraud cases. This dataset includes detailed information about contractors, contract terms, and cases marked as fraudulent. Machine learning algorithms are used to learn patterns of fraudulent contracts based on this data. This allows the system to recognize existing fraud patterns and identify risks in new contracts. 【0341】 Real-time data analysis and emotion recognition 【0342】 During the contract process, the terminal not only receives contract information and identity verification documents provided by the user, but also captures the user's intent and emotions through voice or text. This information is transmitted to a server and analyzed by a natural language processing engine. Image recognition technology is also used to verify the authenticity of the submitted documents. 【0343】 Furthermore, an emotion engine analyzes the user's emotional state. This engine analyzes the tone of voice, word choice, and conversational context to recognize the user's emotions in real time. For example, if a user is clearly experiencing stress during a contract, that information could influence the risk assessment. 【0344】 Anomaly detection and notification 【0345】 Based on the analyzed information, the server compares it to previously learned fraud cases and calculates an anomaly risk score for the contract. During this process, user sentiment analysis results are also integrated, and particular attention is paid to any unusual patterns. If the risk is high, the system detects the anomaly and sends a notification to alert the responsible party. 【0346】 Improvement and response 【0347】 In situations deemed high-risk, the server automatically notifies the responsible party of the details and countermeasures. For example, it provides specific suggestions if it determines that additional information should be requested from the user or if a different identity verification procedure is necessary. 【0348】 Continuous model improvement 【0349】 The server collects all feedback from the contract process and uses it to improve the overall model, including the sentiment engine. Newly emerging fraud methods and data on user sentiment improve the model's accuracy and responsiveness. This enables the system to perform more comprehensive fraud detection and prevention, thereby enhancing the security of the service. 【0350】 In this way, by incorporating emotion recognition, the present invention functions as a system that, in addition to detecting fraudulent contracts, improves the user experience and enables the realization of more accurate services. 【0351】 The following describes the processing flow. 【0352】 Step 1: 【0353】 The server collects past contract information and fraud cases from a database, inputs them into a machine learning algorithm, and begins analysis. This builds a dataset that learns common patterns and features in fraudulent contracts. 【0354】 Step 2: 【0355】 The device receives contract information and identity verification documents provided by the user at the start of the contract process. At this time, the user's spoken voice and entered text data are also captured and used for emotion recognition. 【0356】 Step 3: 【0357】 The server processes contract information sent from the terminal in real time. A natural language processing engine analyzes the text data and formats information such as the contract holder's name and address. It also uses image recognition technology to verify whether the submitted identification documents are genuine. 【0358】 Step 4: 【0359】 The server operates an emotion engine that recognizes emotions from the user's voice or written words. It analyzes the user's tone of voice, word choice, and speech rhythm to understand the user's emotional state. This information is also incorporated into the risk assessment of the entire contract process. 【0360】 Step 5: 【0361】 The server compares the analyzed contract information and sentiment recognition results with fraud patterns learned in the past. This comparison calculates a risk score to assess the likelihood that the currently ongoing contract is fraudulent. 【0362】 Step 6: 【0363】 If the server exceeds a certain risk score, it determines that there is a high probability of fraud and automatically sends a notification to the responsible party. This notification includes additional recommended actions based on the user's contract details, detected risks, and emotional state. 【0364】 Step 7: 【0365】 The server collects all feedback after the contract process is complete. This data, including new fraudulent practices and user sentiment, is used to improve the machine learning models and sentiment engine. This process enhances the overall accuracy and reliability of the system. 【0366】 (Example 2) 【0367】 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". 【0368】 In modern contract processes, the risk of fraudulent contracts is increasing, and existing detection methods based on existing technologies are insufficient to prevent them. Furthermore, there is a lack of technology that considers the emotional state of users when determining fraud risk, thus necessitating a more comprehensive contract management approach. Additionally, there is a need for technology that analyzes the emotional state of users during contract procedures and utilizes that information to improve the accuracy of services. 【0369】 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. 【0370】 This invention includes a server comprising: a machine learning means that recognizes patterns of contracts that violate rules by collecting and learning information on past contracts and fraud cases; a data analysis means that analyzes data received during the contract procedure in real time using natural language processing and image recognition technology and evaluates the validity of the information; and an emotion recognition means that analyzes the emotional state from the user's voice and text and reflects that emotional information in the contract risk assessment. This enables comprehensive anomaly detection that takes into account the user's emotional information, which was lacking in conventional fraudulent contract detection technology, thereby reducing fraud risk and improving the accuracy of the service. 【0371】 "Machine learning methods" are technologies that collect information on past contracts and cases of fraud, and use that information to recognize contract patterns that violate regulations. 【0372】 "Data analysis means" refers to a technology that uses natural language processing and image recognition technologies to analyze data received during contract procedures in real time and evaluate the validity of the information. 【0373】 "Emotion recognition means" refers to technology that analyzes a user's emotional state from their voice or text data and incorporates that emotional information into the risk assessment of a contract. 【0374】 An "anomaly detection method" is a technology that detects anomalies and calculates risk by comparing the analyzed information with learned fraud patterns. 【0375】 A "notification method" is a technology that assesses the possibility of a rule violation and automatically notifies the responsible person based on that possibility. 【0376】 "Model improvement methods" refer to techniques for continuously improving the overall model based on the feedback received. 【0377】 This invention provides a system for effectively detecting and preventing fraudulent contracts, enabling comprehensive judgment that takes into account the user's emotional state. By combining machine learning and sentiment analysis technologies, this system aims to reduce the risk of fraud at each stage of the contract process while improving the user experience. 【0378】 The server collects historical contract information and data on fraudulent cases, and uses this data to train machine learning models. Specifically, it uses machine learning libraries such as TensorFlow and PyTorch to train the models. The models are used to recognize fraudulent contract patterns and identify risks in new contracts. 【0379】 The terminal receives contract information and identity verification documents provided by the user during the contract process. The terminal digitizes the documents using optical character recognition (OCR) technology and converts the user's voice into text using a speech recognition system. This information is transmitted to the server in real time and analyzed by a natural language processing engine. 【0380】 Based on the analyzed audio and text data, the server uses an emotion engine to evaluate the user's emotional state. Specifically, it analyzes the tone of voice, word choice, and conversational context to determine what emotions the user is experiencing. The Hugging Face emotion analysis model is used for this evaluation. 【0381】 This system further compares the analyzed information against patterns of fraudulent contracts, detects anomalies, and calculates a risk score. Scikit-learn's anomaly detection algorithm is used for this purpose. If a high risk is determined, the system sends a notification to the responsible party and suggests additional countermeasures. 【0382】 A concrete example is a mortgage application. When a user attempts to sign a mortgage contract, the terminal detects stress from the user's voice. The server evaluates the detected stress level as high risk and sends a notification to the person in charge stating, "The user may be experiencing stress; please consider additional identity verification procedures." In this case, a prompt such as, "Suggest countermeasures to take when a specific emotional state (e.g., anxiety, stress) is detected during the mortgage contract review process," is used as an example input to the generating AI model. 【0383】 By implementing the invention in this way, it is possible to simultaneously achieve improved fraud detection capabilities and enhanced user experience. 【0384】 The flow of the specific processing in Example 2 will be explained using Figure 13. 【0385】 Step 1: 【0386】 The server collects historical contract information and fraud cases from the contract database. The input consists of the contractor's personal information, contract details, and records of fraud cases. The server executes SQL queries to extract this information as a dataset. The output is a fraudulent contract dataset in an analyzable format. 【0387】 Step 2: 【0388】 The server trains a machine learning model using the collected dataset. The input is the fraudulent contract dataset obtained in Step 1. Specifically, the server uses TensorFlow or PyTorch to preprocess the data, select features, and train the model. The output is a trained model that recognizes fraudulent contract patterns. 【0389】 Step 3: 【0390】 The terminal receives contract information, voice recordings, and identification documents entered by the user during the contract process. The input consists of this information provided by the user. The terminal uses OCR technology to digitize documents and applies speech recognition technology to convert speech to text. The output is data in a digital format suitable for processing on the server. 【0391】 Step 4: 【0392】 The server receives digital data transmitted from the terminal and analyzes its content using a natural language processing engine. The input is the text data processed in step 3. The server analyzes the text content and extracts information to determine the legitimacy and potential fraud of the contract. The output is a dataset containing the analysis results. 【0393】 Step 5: 【0394】 The server analyzes the user's emotional state using an emotion engine. The input is the audio and text data obtained in step 3. Specifically, it uses the Hugging Face emotion analysis model to analyze voice tone, word choice, and context. The output is data indicating the user's emotional state. 【0395】 Step 6: 【0396】 The server detects anomalies by comparing the analyzed contract data and sentiment state with a trained model and calculates a risk score. The inputs are the analyzed data from step 4 and the sentiment data from step 5. The server uses Scikit-learn to evaluate the similarity to known fraudulent contract patterns and quantifies the risk of the contract. The output is the risk score. 【0397】 Step 7: 【0398】 The server notifies the responsible person based on the risk score. The input is the risk score obtained in step 6. Notifications are sent via email or the company's internal messaging system, and the responsible person is notified if additional verification work is required based on the risk. The output is the notification message sent to the responsible person. 【0399】 Step 8: 【0400】 The server collects feedback obtained after the contract process and uses it to improve the model. Inputs are actual contract results and user feedback obtained from the entire system. The server incorporates new fraud patterns and sentiment data into the model to improve the system's adaptability. The output is the updated, trained model. 【0401】 (Application Example 2) 【0402】 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." 【0403】 In modern electronic transactions, fraudulent transactions are on the rise, posing a significant security challenge. Furthermore, conventional fraud detection systems often focus solely on transaction patterns and fail to consider the user's emotional state, resulting in inaccurate detection. Therefore, the objective of this invention is to improve the accuracy of fraud detection and realize a safer trading environment. 【0404】 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. 【0405】 In this invention, the server includes machine learning means for collecting and recognizing past transaction information and fraudulent activity cases; data analysis means for analyzing data received during transaction procedures in real time using natural language processing and image recognition technologies and evaluating the legitimacy of the data; and emotion recognition means for analyzing user voice data and evaluating emotional state. This enables highly accurate detection of fraudulent transactions and enhanced security that takes user experience into consideration. 【0406】 "Past transaction information" refers to data on previously conducted transactions that are collected for the purpose of detecting fraudulent activity. 【0407】 "Examples of fraudulent activity" refers to data that shows specific examples of fraudulent transactions or actions that have occurred in the past. 【0408】 "Machine learning methods" are algorithms and techniques that recognize patterns based on data, enabling prediction and classification. 【0409】 "Natural language processing methods" are technologies used to analyze text and speech and understand their meaning and context. 【0410】 "Image recognition technology" is a technology that analyzes image data to recognize objects and characters, and to extract features. 【0411】 "Data analysis methods" refer to techniques and systems for processing various types of data using statistics and algorithms to gain insights. 【0412】 "Emotion recognition methods" are technologies for inferring and evaluating human emotions from things like voice and facial expressions. 【0413】 An "anomaly detection method" is a system for detecting patterns or behaviors that deviate from normal and for evaluating the associated risks. 【0414】 A "notification mechanism" refers to a function or device used by the system to inform relevant parties of anomalies or important information it has detected. 【0415】 "Model improvement methods" are techniques for continuously improving machine learning models based on new data and feedback, thereby enhancing their accuracy. 【0416】 The system that implements this application is server-centric and consists of multiple means with various roles. The server first collects past transaction information and fraudulent activity cases, and uses machine learning to recognize patterns of fraudulent transactions based on this data. During the transaction process, the server analyzes the data received from the terminal in real time using natural language processing and image recognition technologies to evaluate the legitimacy of the data. 【0417】 Furthermore, the terminal acquires the user's voice data, and an emotion recognition system evaluates the user's emotional state. This allows the server to compare the input information with learned fraud patterns, detect anomalies, and calculate the risk. If the calculated risk exceeds a certain level, the server immediately and automatically notifies the responsible person through a notification system. 【0418】 The hardware used will primarily consist of devices such as smartphones and tablets, while the software will utilize Google Cloud's Natural Language API and various speech analysis libraries. 【0419】 As a concrete example, the terminal receives audio from the user during payment, the server analyzes the audio data to read the emotion, and if it determines that there is a high probability of fraud, an alert is immediately sent to the person in charge. An example of a prompt sentence to be input into the generating AI model is: "Explain how to proactively detect fraud by analyzing the emotion from the user's voice and context during smartphone payments." 【0420】 The flow of a specific process in Application Example 2 will be explained using Figure 14. 【0421】 Step 1: 【0422】 The server collects past transaction information and fraud cases from a database. Using this data as input, it learns patterns of fraudulent transactions using a machine learning algorithm and outputs the results as a model. This model becomes the new standard for evaluating transaction data. 【0423】 Step 2: 【0424】 The terminal receives data entered by the user during the transaction process. This input data includes voice data, text data, and image data. The terminal then sends this data to the server. 【0425】 Step 3: 【0426】 The server uses natural language processing techniques to analyze the meaning of the received audio and text data. The audio data is converted to text using a speech analysis library, and based on the output, an emotion recognition system evaluates the user's emotional state. This result serves as a basis for determining the user's emotions during the transaction. 【0427】 Step 4: 【0428】 The server analyzes the received image data using image recognition technology to verify the authenticity and integrity of the submitted identity verification documents. Based on the results, it determines whether the data has been forged or tampered with and outputs information to assess the risk of fraud. 【0429】 Step 5: 【0430】 The server compares pre-processed transaction data with historical database data based on processed sentiment data and image analysis results, and uses an anomaly detection algorithm to assess the likelihood of fraud. A risk score is calculated and output as a result. If the likelihood of fraud is high, the information is passed on to the next step. 【0431】 Step 6: 【0432】 Based on the calculated risk score, the server immediately notifies the responsible party of the risk using notification methods. The notification includes detailed information about the detected fraud and countermeasures. This notification facilitates a rapid response. 【0433】 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. 【0434】 Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). An 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. 【0435】 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. 【0436】 [Third Embodiment] 【0437】 Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment. 【0438】 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. 【0439】 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). 【0440】 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. 【0441】 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. 【0442】 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). 【0443】 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. 【0444】 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. 【0445】 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. 【0446】 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. 【0447】 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. 【0448】 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". 【0449】 The present invention is a system for effectively detecting and preventing fraudulent contracts, and embodiments thereof are shown below. 【0450】 Data collection and learning 【0451】 First, the server collects data on past contract information and fraud cases. This data includes basic information about the contract holder, contract details, and patterns of past fraud cases. This data is then used to train a machine learning model. For example, it learns cases where the addresses of multiple contract holders suddenly change as a potential pattern of fraud. 【0452】 Real-time data analysis 【0453】 During the new contract process, the terminal receives entered subscriber information and identification documents. This data is sent to a server, where natural language processing is used to analyze the subscriber's text information. For example, it verifies that the user's name and address match other publicly available information. Image recognition technology is also used to evaluate the authenticity of the submitted identification documents, checking for tampering or inconsistencies. 【0454】 Anomaly detection and risk assessment 【0455】 Contract information analyzed in real time is compared against previously learned fraud patterns. The server detects potentially fraudulent information during this process. For example, if the entered information resembles a pattern previously associated with fraud, a risk score is calculated. A high risk score indicates that an anomaly has been detected. 【0456】 Automated response and notifications 【0457】 If a transaction is deemed highly suspicious, an automated notification is sent from the server to the responsible party. This notification includes details of the potentially fraudulent contract and a suggestion to implement additional identity verification procedures. This allows the responsible party to take the necessary action quickly. 【0458】 Continuous model improvements 【0459】 The results of each contract process and the feedback obtained from subsequent actions are accumulated. The server analyzes this feedback and uses it to improve the model. In this way, the accuracy of the system improves, making it capable of handling new fraudulent methods. 【0460】 Through this series of processes, the present invention can function as a system that reduces the risk of fraudulent contracts and enables the provision of safer services to users. 【0461】 The following describes the processing flow. 【0462】 Step 1: 【0463】 The server collects historical contract information and fraud case data obtained from communication and financial services. This data includes subscriber information, contract details, and whether or not fraud occurred. Based on this data, the server extracts features to identify fraudulent contracts and uses them to train machine learning models. 【0464】 Step 2: 【0465】 The device receives contract information and images of identification documents entered by the user when a new contract is made. This information is necessary to verify the legitimacy of the contract and is sent to the next analysis step. 【0466】 Step 3: 【0467】 The server receives contract information sent from the terminal and analyzes the subscriber's text data using natural language processing. For example, it verifies whether the user's name, address, etc., are consistent. It also uses image recognition technology to verify the integrity and validity of identity verification documents and check for scanning or tampering. 【0468】 Step 4: 【0469】 The server compares the contract information obtained through analysis with pre-trained patterns of fraudulent contracts. Using a predictive model, it evaluates how similar this is to past fraud cases and calculates a risk score. If this score is high, it is determined that the input information is likely to be fraudulent. 【0470】 Step 5: 【0471】 The server automatically notifies the responsible party if it detects an anomaly based on the risk score. This notification includes detailed information about potentially fraudulent contracts and suggestions for additional verification steps. This allows the responsible party to respond quickly and appropriately. 【0472】 Step 6: 【0473】 The server stores the feedback and detection results obtained after all contract processes are completed. This information is used to continuously improve the machine learning model and update the system to enhance its accuracy. This continuous process ensures the system remains capable of responding to new fraudulent methods. 【0474】 (Example 1) 【0475】 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." 【0476】 Traditional contract management systems often had a problem of delayed detection of fraudulent contracts, resulting in damages caused by fraudulent activities. Furthermore, it was difficult to analyze information in real time at each stage of the contract process and accurately detect anomalies. Therefore, it was difficult to respond quickly to potential fraud and comprehensive fraud prevention was challenging. 【0477】 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. 【0478】 In this invention, the server includes machine learning means that collect and learn past contract information and fraud cases using a data storage device; information analysis means that analyze information received during new contract procedures in real time using text analysis and image processing technology and evaluate its legitimacy; and anomaly identification means that compare input data with learned fraud characteristics to identify anomalies and calculate the degree of risk. This enables rapid detection of fraudulent contracts and early notification to the responsible person. 【0479】 A "data storage device" is a device designed to securely store information such as contract details and fraud cases for extended periods, and to allow for quick access when needed. 【0480】 "Machine learning methods" refer to methods that use algorithms and processes to learn patterns based on collected data and identify the characteristics of fraudulent contracts. 【0481】 "Text analysis" is the process of analyzing text data using natural language processing techniques to understand its content and structure. 【0482】 "Image processing technology" refers to the techniques used to analyze digital images and extract and recognize information within them. 【0483】 "Information analysis means" refers to methods for evaluating received contract information in real time and checking the validity and consistency of the data. 【0484】 An "anomaly identification method" is a process that compares input data with existing fraudulent characteristics to quickly detect abnormal patterns and behaviors. 【0485】 "Risk level" is an indicator that numerically represents the likelihood of fraudulent activity occurring under specific conditions, and it indicates the risk level. 【0486】 "Notification methods" refer to methods or systems for informing relevant parties about the potential for misconduct and the need for action. 【0487】 "Model tuning" refers to the process of updating a machine learning model based on feedback to improve the system's accuracy and responsiveness. 【0488】 "Procedure" means a series of operations or processes related to a contract, including the process from data entry to verification and finalization. 【0489】 This invention is an advanced system for detecting and preventing fraudulent contracts. This system combines multiple solutions to monitor, detect, and quickly address fraudulent activity in the contract process in real time. 【0490】 The server collects historical contract information and fraud cases via a data storage device and uses this to train a machine learning model. Specific software used includes TensorFlow and PyTorch. The machine learning model learns patterns of fraud and responds to fraud detection in real time. 【0491】 The server also uses natural language processing (NLP) techniques for text analysis. During the new contract process, it analyzes subscriber information transmitted from the terminal and evaluates the validity of the data. Specific NLP libraries (e.g., spaCy and NLTK) are used as the software. 【0492】 Regarding image recognition technology, the server processes images of identity verification documents sent from the terminal and evaluates their authenticity. Specific technologies used here include OpenCV and general cloud-based image recognition services (e.g., AWS Rekognition). This analysis checks for document tampering and inconsistencies. 【0493】 When a user applies for a new contract, the server evaluates the entered information based on existing fraud patterns. If an anomaly is detected and a risk assessment is performed, the server automatically sends a notification to the responsible person using a notification system. This notification is sent via email or the company's internal messaging service. 【0494】 As a concrete example, in a bank loan application, if the information provided by the user resembles past fraud patterns, the system detects the anomaly and immediately sends an alert to the bank representative. An example of a prompt message to the generated AI model in this case would be, "Evaluate the new loan application information and check for similarities to past fraud patterns." 【0495】 This invention reduces the risk of fraudulent activity in the contract process and makes it possible to provide users with safer and more appropriate services. 【0496】 The flow of the specific processing in Example 1 will be explained using Figure 11. 【0497】 Step 1: 【0498】 The server collects past contract information and fraud cases from a database. This database includes basic information about the contractor, detailed contract terms, and past fraud patterns. Using the collected data, the server trains a machine learning model using frameworks such as TensorFlow or PyTorch. This results in a model that has learned the characteristics of fraudulent contracts. 【0499】 Step 2: 【0500】 When a user initiates a new contract, the device inputs subscriber information and identity verification documents and sends them to the server. This input includes name, address, contact information, and scanned images of identity verification documents. The server receives this information, analyzes the text data using NLP libraries such as spaCy or NLTK, and evaluates the accuracy of the information. The resulting analyzed text information is then output. 【0501】 Step 3: 【0502】 The server receives images of identity verification documents sent from the terminal and analyzes them using image recognition technologies such as OpenCV and AWS Rekognition. The server checks the integrity of the images and detects tampering, outputting an evaluation result of their legitimacy. This process confirms that the documents are genuine. 【0503】 Step 4: 【0504】 The server uses the results of text and image analysis to compare the input information with learned fraud patterns. By matching it against previously learned fraud patterns, it identifies anomalies and assesses the risk. The server then calculates a risk score and outputs the result. 【0505】 Step 5: 【0506】 Based on the detection results of an anomaly, the server automatically sends a notification to the responsible person if it determines that notification is necessary according to internal evaluation criteria. The notification content is conveyed via email or messaging system. This notification includes the specific nature of the anomaly and recommended additional actions. The notified information is output. 【0507】 Step 6: 【0508】 The final results and feedback from the contract process are stored in a database. Based on this feedback, the server adjusts and improves the machine learning model to address new fraudulent methods. As a result, an improved model is output. 【0509】 (Application Example 1) 【0510】 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." 【0511】 The problem that this invention aims to solve is to improve security and facilitate smooth procedures for legitimate use by detecting potential fraudulent activity in real time during online transactions and account registration, and warning users and related parties in advance. 【0512】 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. 【0513】 This invention includes a server comprising: machine learning means for recognizing patterns of fraudulent activity by collecting and learning from past transaction information and fraud cases; data analysis means for analyzing data received during the transaction process in real time using natural language processing and visual data recognition technologies and evaluating the validity of the data; and information provision means for warning of potential fraud in real time through the user's digital device. This makes it possible to quickly and automatically assess the risk of fraudulent activity and notify users and stakeholders in advance. 【0514】 "Transaction information" refers to data related to past commercial activities or negotiations, including details such as date, time, conditions, and participants. 【0515】 "Fraudulent activity cases" are actual examples of fraudulent acts that have occurred in the past, and are used as data to learn patterns of fraud and deception. 【0516】 "Machine learning" is a technology in which computers automatically learn patterns from large amounts of data to make future predictions and classifications. 【0517】 "Natural language processing" is a technology that enables computers to understand, analyze, and generate human language. 【0518】 "Visual data recognition technology" is a technology that analyzes visual data such as images and videos and extracts information from it. 【0519】 "Data analysis means" refers to technologies and methods that analyze received data in real time to evaluate its validity and detect anomalies. 【0520】 "Information provision means" refers to methods and technologies for providing users with real-time warnings and guidance regarding potential fraud. 【0521】 A "warning" is a notification that prompts users to take action by informing them in advance of potential risks or fraudulent activities. 【0522】 In the system that implements this application, the server first collects past transaction information and fraud cases, and uses this to train a machine learning model. The server leverages machine learning libraries such as TensorFlow and PyTorch to learn patterns of fraudulent activity from large amounts of data. This process creates a model that can recognize potential patterns of fraud. 【0523】 Next, when a user conducts a transaction, the device (such as a smartphone) collects the transaction information and identity verification data entered by the user. The device sends the entered data to the server, which processes the received data in real time. At this stage, text data is analyzed using natural language processing tools such as Hugging Face Transformers, and the validity of visual data is verified using image recognition libraries such as OpenCV. 【0524】 If the server analyzes transaction information and the data entered by the user resembles a learned fraud pattern, the server performs a risk assessment. If an anomaly is detected, the user is warned in real time through the terminal's information provision system. This notification allows the user to immediately review their input information. In short, the system provides a mechanism to quickly and effectively prevent fraudulent activity. 【0525】 For example, when a user creates a new account while shopping online, if the name and address information they enter matches past fraud cases identified on the server, a warning will be displayed on their smartphone, and the user will be prompted to verify the information. 【0526】 Examples of prompt statements when using a generative AI model are as follows: 【0527】 "Please enter your username, address, and identification documents to assess the risk of fraudulent contracts." 【0528】 Based on this prompt, the AI ​​model evaluates the security of contracts and transactions, helping to proactively prevent fraud risks. 【0529】 The flow of a specific process in Application Example 1 will be explained using Figure 12. 【0530】 Step 1: 【0531】 The user initiates a transaction and enters the necessary information into the terminal. This information includes name, address, payment information, and scanned images of identification documents. The terminal then prepares to send this data to the server. 【0532】 Step 2: 【0533】 The server receives input data sent from the terminal. After receiving the data, it uses natural language processing tools (such as Hugging Face Transformers) to analyze the text data and check for any information similar to malicious patterns. The processed data is then used in the next processing step. 【0534】 Step 3: 【0535】 The server uses visual data recognition technology (such as OpenCV) to analyze the scanned images of identification documents. This process verifies that the images have not been tampered with and that the documents are authentic. The legitimacy of the processed image data is evaluated, and the results are sent to the next step. 【0536】 Step 4: 【0537】 The server analyzes the data and compares it to pre-trained fraud patterns to detect anomalies. A risk score is then calculated for the input data. If the risk score is high, it is marked as an anomaly, and a warning is deemed necessary. 【0538】 Step 5: 【0539】 If an anomaly is detected, the server will send a warning message to the terminal in real time. This message will contain details of the information that needs to be checked. Upon receiving this notification, the user can review the information they entered and make corrections or reconfirmations as needed. 【0540】 Step 6: 【0541】 After all transactions are completed, the execution results and feedback are saved to the server. This data helps identify new fraudulent patterns and areas for improvement, which in turn helps train the generative AI model. As a result, the accuracy of the system will improve in future transactions. 【0542】 In this way, users and servers can work together to provide a secure trading environment while establishing protection against fraudulent activity. 【0543】 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. 【0544】 This invention provides a comprehensive judgment capability that takes into account the user's emotional state by combining an emotion engine with a system that effectively detects and prevents fraudulent contracts. An embodiment thereof is shown below. 【0545】 Data collection and learning 【0546】 First, the server collects historical contract data and past fraud cases. This dataset includes detailed information about contractors, contract terms, and cases marked as fraudulent. Machine learning algorithms are used to learn patterns of fraudulent contracts based on this data. This allows the system to recognize existing fraud patterns and identify risks in new contracts. 【0547】 Real-time data analysis and emotion recognition 【0548】 During the contract process, the terminal not only receives contract information and identity verification documents provided by the user, but also captures the user's intent and emotions through voice or text. This information is transmitted to a server and analyzed by a natural language processing engine. Image recognition technology is also used to verify the authenticity of the submitted documents. 【0549】 Furthermore, an emotion engine analyzes the user's emotional state. This engine analyzes the tone of voice, word choice, and conversational context to recognize the user's emotions in real time. For example, if a user is clearly experiencing stress during a contract, that information could influence the risk assessment. 【0550】 Anomaly detection and notification 【0551】 Based on the analyzed information, the server compares it to previously learned fraud cases and calculates an anomaly risk score for the contract. During this process, user sentiment analysis results are also integrated, and particular attention is paid to any unusual patterns. If the risk is high, the system detects the anomaly and sends a notification to alert the responsible party. 【0552】 Improvement and response 【0553】 In situations deemed high-risk, the server automatically notifies the responsible party of the details and countermeasures. For example, it provides specific suggestions if it determines that additional information should be requested from the user or if a different identity verification procedure is necessary. 【0554】 Continuous model improvement 【0555】 The server collects all feedback from the contract process and uses it to improve the overall model, including the sentiment engine. Newly emerging fraud methods and data on user sentiment improve the model's accuracy and responsiveness. This enables the system to perform more comprehensive fraud detection and prevention, thereby enhancing the security of the service. 【0556】 In this way, by incorporating emotion recognition, the present invention functions as a system that, in addition to detecting fraudulent contracts, improves the user experience and enables the realization of more accurate services. 【0557】 The following describes the processing flow. 【0558】 Step 1: 【0559】 The server collects past contract information and fraud cases from a database, inputs them into a machine learning algorithm, and begins analysis. This builds a dataset that learns common patterns and features in fraudulent contracts. 【0560】 Step 2: 【0561】 The device receives contract information and identity verification documents provided by the user at the start of the contract process. At this time, the user's spoken voice and entered text data are also captured and used for emotion recognition. 【0562】 Step 3: 【0563】 The server processes contract information sent from the terminal in real time. A natural language processing engine analyzes the text data and formats information such as the contract holder's name and address. It also uses image recognition technology to verify whether the submitted identification documents are genuine. 【0564】 Step 4: 【0565】 The server operates an emotion engine that recognizes emotions from the user's voice or written words. It analyzes the user's tone of voice, word choice, and speech rhythm to understand the user's emotional state. This information is also incorporated into the risk assessment of the entire contract process. 【0566】 Step 5: 【0567】 The server compares the analyzed contract information and sentiment recognition results with fraud patterns learned in the past. This comparison calculates a risk score to assess the likelihood that the currently ongoing contract is fraudulent. 【0568】 Step 6: 【0569】 If the server exceeds a certain risk score, it determines that there is a high probability of fraud and automatically sends a notification to the responsible party. This notification includes additional recommended actions based on the user's contract details, detected risks, and emotional state. 【0570】 Step 7: 【0571】 The server collects all feedback after the contract process is complete. This data, including new fraudulent practices and user sentiment, is used to improve the machine learning models and sentiment engine. This process enhances the overall accuracy and reliability of the system. 【0572】 (Example 2) 【0573】 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." 【0574】 In modern contract processes, the risk of fraudulent contracts is increasing, and existing detection methods based on existing technologies are insufficient to prevent them. Furthermore, there is a lack of technology that considers the emotional state of users when determining fraud risk, thus necessitating a more comprehensive contract management approach. Additionally, there is a need for technology that analyzes the emotional state of users during contract procedures and utilizes that information to improve the accuracy of services. 【0575】 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. 【0576】 This invention includes a server comprising: a machine learning means that recognizes patterns of contracts that violate rules by collecting and learning information on past contracts and fraud cases; a data analysis means that analyzes data received during the contract procedure in real time using natural language processing and image recognition technology and evaluates the validity of the information; and an emotion recognition means that analyzes the emotional state from the user's voice and text and reflects that emotional information in the contract risk assessment. This enables comprehensive anomaly detection that takes into account the user's emotional information, which was lacking in conventional fraudulent contract detection technology, thereby reducing fraud risk and improving the accuracy of the service. 【0577】 "Machine learning methods" are technologies that collect information on past contracts and cases of fraud, and use that information to recognize contract patterns that violate regulations. 【0578】 "Data analysis means" refers to a technology that uses natural language processing and image recognition technologies to analyze data received during contract procedures in real time and evaluate the validity of the information. 【0579】 "Emotion recognition means" refers to technology that analyzes a user's emotional state from their voice or text data and incorporates that emotional information into the risk assessment of a contract. 【0580】 An "anomaly detection method" is a technology that detects anomalies and calculates risk by comparing the analyzed information with learned fraud patterns. 【0581】 A "notification method" is a technology that assesses the possibility of a rule violation and automatically notifies the responsible person based on that possibility. 【0582】 "Model improvement methods" refer to techniques for continuously improving the overall model based on the feedback received. 【0583】 This invention provides a system for effectively detecting and preventing fraudulent contracts, enabling comprehensive judgment that takes into account the user's emotional state. By combining machine learning and sentiment analysis technologies, this system aims to reduce the risk of fraud at each stage of the contract process while improving the user experience. 【0584】 The server collects historical contract information and data on fraudulent cases, and uses this data to train machine learning models. Specifically, it uses machine learning libraries such as TensorFlow and PyTorch to train the models. The models are used to recognize fraudulent contract patterns and identify risks in new contracts. 【0585】 The terminal receives contract information and identity verification documents provided by the user during the contract process. The terminal digitizes the documents using optical character recognition (OCR) technology and converts the user's voice into text using a speech recognition system. This information is transmitted to the server in real time and analyzed by a natural language processing engine. 【0586】 Based on the analyzed audio and text data, the server uses an emotion engine to evaluate the user's emotional state. Specifically, it analyzes the tone of voice, word choice, and conversational context to determine what emotions the user is experiencing. The Hugging Face emotion analysis model is used for this evaluation. 【0587】 This system further compares the analyzed information against patterns of fraudulent contracts, detects anomalies, and calculates a risk score. Scikit-learn's anomaly detection algorithm is used for this purpose. If a high risk is determined, the system sends a notification to the responsible party and suggests additional countermeasures. 【0588】 A concrete example is a mortgage application. When a user attempts to sign a mortgage contract, the terminal detects stress from the user's voice. The server evaluates the detected stress level as high risk and sends a notification to the person in charge stating, "The user may be experiencing stress; please consider additional identity verification procedures." In this case, a prompt such as, "Suggest countermeasures to take when a specific emotional state (e.g., anxiety, stress) is detected during the mortgage contract review process," is used as an example input to the generating AI model. 【0589】 By implementing the invention in this way, it is possible to simultaneously achieve improved fraud detection capabilities and enhanced user experience. 【0590】 The flow of the specific processing in Example 2 will be explained using Figure 13. 【0591】 Step 1: 【0592】 The server collects historical contract information and fraud cases from the contract database. The input consists of the contractor's personal information, contract details, and records of fraud cases. The server executes SQL queries to extract this information as a dataset. The output is a fraudulent contract dataset in an analyzable format. 【0593】 Step 2: 【0594】 The server trains a machine learning model using the collected dataset. The input is the fraudulent contract dataset obtained in Step 1. Specifically, the server uses TensorFlow or PyTorch to preprocess the data, select features, and train the model. The output is a trained model that recognizes fraudulent contract patterns. 【0595】 Step 3: 【0596】 The terminal receives contract information, voice recordings, and identification documents entered by the user during the contract process. The input consists of this information provided by the user. The terminal uses OCR technology to digitize documents and applies speech recognition technology to convert speech to text. The output is data in a digital format suitable for processing on the server. 【0597】 Step 4: 【0598】 The server receives digital data transmitted from the terminal and analyzes its content using a natural language processing engine. The input is the text data processed in step 3. The server analyzes the text content and extracts information to determine the legitimacy and potential fraud of the contract. The output is a dataset containing the analysis results. 【0599】 Step 5: 【0600】 The server analyzes the user's emotional state using an emotion engine. The input is the audio and text data obtained in step 3. Specifically, it uses the Hugging Face emotion analysis model to analyze voice tone, word choice, and context. The output is data indicating the user's emotional state. 【0601】 Step 6: 【0602】 The server detects anomalies by comparing the analyzed contract data and sentiment state with a trained model and calculates a risk score. The inputs are the analyzed data from step 4 and the sentiment data from step 5. The server uses Scikit-learn to evaluate the similarity to known fraudulent contract patterns and quantifies the risk of the contract. The output is the risk score. 【0603】 Step 7: 【0604】 The server notifies the responsible person based on the risk score. The input is the risk score obtained in step 6. Notifications are sent via email or the company's internal messaging system, and the responsible person is notified if additional verification work is required based on the risk. The output is the notification message sent to the responsible person. 【0605】 Step 8: 【0606】 The server collects feedback obtained after the contract process and uses it to improve the model. Inputs are actual contract results and user feedback obtained from the entire system. The server incorporates new fraud patterns and sentiment data into the model to improve the system's adaptability. The output is the updated, trained model. 【0607】 (Application Example 2) 【0608】 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." 【0609】 In modern electronic transactions, fraudulent transactions are on the rise, posing a significant security challenge. Furthermore, conventional fraud detection systems often focus solely on transaction patterns and fail to consider the user's emotional state, resulting in inaccurate detection. Therefore, the objective of this invention is to improve the accuracy of fraud detection and realize a safer trading environment. 【0610】 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. 【0611】 In this invention, the server includes machine learning means for collecting and recognizing past transaction information and fraudulent activity cases; data analysis means for analyzing data received during transaction procedures in real time using natural language processing and image recognition technologies and evaluating the legitimacy of the data; and emotion recognition means for analyzing user voice data and evaluating emotional state. This enables highly accurate detection of fraudulent transactions and enhanced security that takes user experience into consideration. 【0612】 "Past transaction information" refers to data on previously conducted transactions that are collected for the purpose of detecting fraudulent activity. 【0613】 "Examples of fraudulent activity" refers to data that shows specific examples of fraudulent transactions or actions that have occurred in the past. 【0614】 "Machine learning methods" are algorithms and techniques that recognize patterns based on data, enabling prediction and classification. 【0615】 "Natural language processing methods" are technologies used to analyze text and speech and understand their meaning and context. 【0616】 "Image recognition technology" is a technology that analyzes image data to recognize objects and characters, and to extract features. 【0617】 "Data analysis methods" refer to techniques and systems for processing various types of data using statistics and algorithms to gain insights. 【0618】 "Emotion recognition methods" are technologies for inferring and evaluating human emotions from things like voice and facial expressions. 【0619】 An "anomaly detection method" is a system for detecting patterns or behaviors that deviate from normal and for evaluating the associated risks. 【0620】 A "notification mechanism" refers to a function or device used by the system to inform relevant parties of anomalies or important information it has detected. 【0621】 "Model improvement methods" are techniques for continuously improving machine learning models based on new data and feedback, thereby increasing their accuracy. 【0622】 The system that implements this application is server-centric and consists of multiple means with various roles. The server first collects past transaction information and fraudulent activity cases, and uses machine learning to recognize patterns of fraudulent transactions based on this data. During the transaction process, the server analyzes the data received from the terminal in real time using natural language processing and image recognition technologies to evaluate the legitimacy of the data. 【0623】 Furthermore, the terminal acquires the user's voice data, and an emotion recognition system evaluates the user's emotional state. This allows the server to compare the input information with learned fraud patterns, detect anomalies, and calculate the risk. If the calculated risk exceeds a certain level, the server immediately and automatically notifies the responsible person through a notification system. 【0624】 The hardware used will primarily consist of devices such as smartphones and tablets, while the software will utilize Google Cloud's Natural Language API and various speech analysis libraries. 【0625】 As a concrete example, the terminal receives audio from the user during payment, the server analyzes the audio data to read the emotion, and if it determines that there is a high probability of fraud, an alert is immediately sent to the person in charge. An example of a prompt sentence to be input into the generating AI model is: "Explain how to proactively detect fraud by analyzing the emotion from the user's voice and context during smartphone payments." 【0626】 The flow of a specific process in Application Example 2 will be explained using Figure 14. 【0627】 Step 1: 【0628】 The server collects past transaction information and fraud cases from a database. Using this data as input, it learns patterns of fraudulent transactions using a machine learning algorithm and outputs the results as a model. This model becomes the new standard for evaluating transaction data. 【0629】 Step 2: 【0630】 The terminal receives data entered by the user during the transaction process. This input data includes voice data, text data, and image data. The terminal then sends this data to the server. 【0631】 Step 3: 【0632】 The server uses natural language processing techniques to analyze the meaning of the received audio and text data. The audio data is converted to text using a speech analysis library, and based on the output, an emotion recognition system evaluates the user's emotional state. This result serves as a basis for determining the user's emotions during the transaction. 【0633】 Step 4: 【0634】 The server analyzes the received image data using image recognition technology to verify the authenticity and integrity of the submitted identity verification documents. Based on the results, it determines whether the data has been forged or tampered with and outputs information to assess the risk of fraud. 【0635】 Step 5: 【0636】 The server compares pre-processed transaction data with historical databases based on processed sentiment data and image analysis results, and uses an anomaly detection algorithm to assess the likelihood of fraud. A risk score is calculated and output as a result. If the likelihood of fraud is high, that information is passed on to the next step. 【0637】 Step 6: 【0638】 Based on the calculated risk score, the server immediately notifies the responsible party of the risk using notification methods. The notification includes detailed information about the detected fraud and countermeasures. This notification facilitates a rapid response. 【0639】 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. 【0640】 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. 【0641】 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. 【0642】 [Fourth Embodiment] 【0643】 Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment. 【0644】 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. 【0645】 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). 【0646】 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. 【0647】 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. 【0648】 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). 【0649】 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. 【0650】 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. 【0651】 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. 【0652】 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. 【0653】 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. 【0654】 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. 【0655】 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". 【0656】 The present invention is a system for effectively detecting and preventing fraudulent contracts, and embodiments thereof are shown below. 【0657】 Data collection and learning 【0658】 First, the server collects data on past contract information and fraud cases. This data includes basic information about the contract holder, contract details, and patterns of past fraud cases. This data is then used to train a machine learning model. For example, it learns cases where the addresses of multiple contract holders suddenly change as a potential pattern of fraud. 【0659】 Real-time data analysis 【0660】 During the new contract process, the terminal receives entered subscriber information and identification documents. This data is sent to a server, where natural language processing is used to analyze the subscriber's text information. For example, it verifies that the user's name and address match other publicly available information. Image recognition technology is also used to evaluate the authenticity of the submitted identification documents, checking for tampering or inconsistencies. 【0661】 Anomaly detection and risk assessment 【0662】 Contract information analyzed in real time is compared against previously learned fraud patterns. The server detects potentially fraudulent information during this process. For example, if the entered information resembles a pattern previously associated with fraud, a risk score is calculated. A high risk score indicates that an anomaly has been detected. 【0663】 Automated response and notifications 【0664】 If a transaction is deemed highly suspicious, an automated notification is sent from the server to the responsible party. This notification includes details of the potentially fraudulent contract and a suggestion to implement additional identity verification procedures. This allows the responsible party to take the necessary action quickly. 【0665】 Continuous model improvements 【0666】 The results of each contract process and the feedback obtained from subsequent actions are accumulated. The server analyzes this feedback and uses it to improve the model. In this way, the accuracy of the system improves, making it capable of handling new fraudulent methods. 【0667】 Through this series of processes, the present invention can function as a system that reduces the risk of fraudulent contracts and enables the provision of safer services to users. 【0668】 The following describes the processing flow. 【0669】 Step 1: 【0670】 The server collects historical contract information and fraud case data obtained from communication and financial services. This data includes subscriber information, contract details, and whether or not fraud occurred. Based on this data, the server extracts features to identify fraudulent contracts and uses them to train machine learning models. 【0671】 Step 2: 【0672】 The device receives contract information and images of identification documents entered by the user when a new contract is made. This information is necessary to verify the legitimacy of the contract and is sent to the next analysis step. 【0673】 Step 3: 【0674】 The server receives contract information sent from the terminal and analyzes the subscriber's text data using natural language processing. For example, it verifies whether the user's name, address, etc., are consistent. It also uses image recognition technology to verify the integrity and validity of identity verification documents and check for scanning or tampering. 【0675】 Step 4: 【0676】 The server compares the contract information obtained through analysis with pre-trained patterns of fraudulent contracts. Using a predictive model, it evaluates how similar this is to past fraud cases and calculates a risk score. If this score is high, it is determined that the input information is likely to be fraudulent. 【0677】 Step 5: 【0678】 The server automatically notifies the responsible party if it detects an anomaly based on the risk score. This notification includes detailed information about potentially fraudulent contracts and suggestions for additional verification steps. This allows the responsible party to respond quickly and appropriately. 【0679】 Step 6: 【0680】 The server stores the feedback and detection results obtained after all contract processes are completed. This information is used to continuously improve the machine learning model and update the system to enhance its accuracy. This continuous process ensures the system remains capable of responding to new fraudulent methods. 【0681】 (Example 1) 【0682】 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". 【0683】 Traditional contract management systems often had a problem of delayed detection of fraudulent contracts, resulting in damages caused by fraudulent activities. Furthermore, it was difficult to analyze information in real time at each stage of the contract process and accurately detect anomalies. Therefore, it was difficult to respond quickly to potential fraud and comprehensive fraud prevention was challenging. 【0684】 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. 【0685】 In this invention, the server includes machine learning means that collect and learn past contract information and fraud cases using a data storage device; information analysis means that analyze information received during new contract procedures in real time using text analysis and image processing technology and evaluate its legitimacy; and anomaly identification means that compare input data with learned fraud characteristics to identify anomalies and calculate the degree of risk. This enables rapid detection of fraudulent contracts and early notification to the responsible person. 【0686】 A "data storage device" is a device designed to securely store information such as contract details and fraud cases for extended periods, and to allow for quick access when needed. 【0687】 "Machine learning methods" refer to methods that use algorithms and processes to learn patterns based on collected data and identify the characteristics of fraudulent contracts. 【0688】 "Text analysis" is the process of analyzing text data using natural language processing techniques to understand its content and structure. 【0689】 "Image processing technology" refers to the techniques used to analyze digital images and extract and recognize information within them. 【0690】 "Information analysis means" refers to methods for evaluating received contract information in real time and checking the validity and consistency of the data. 【0691】 An "anomaly identification method" is a process that compares input data with existing fraudulent characteristics to quickly detect abnormal patterns and behaviors. 【0692】 "Risk level" is an indicator that numerically represents the likelihood of fraudulent activity occurring under specific conditions, and it indicates the risk level. 【0693】 "Notification methods" refer to methods or systems for informing relevant parties about the potential for misconduct and the need for action. 【0694】 "Model tuning" refers to the process of updating a machine learning model based on feedback to improve the system's accuracy and responsiveness. 【0695】 "Procedure" means a series of operations or processes related to a contract, including the process from data entry to verification and finalization. 【0696】 This invention is an advanced system for detecting and preventing fraudulent contracts. This system combines multiple solutions to monitor, detect, and quickly address fraudulent activity in the contract process in real time. 【0697】 The server collects historical contract information and fraud cases via a data storage device and uses this to train a machine learning model. Specific software used includes TensorFlow and PyTorch. The machine learning model learns patterns of fraud and responds to fraud detection in real time. 【0698】 The server also uses natural language processing (NLP) techniques for text analysis. During the new contract process, it analyzes subscriber information transmitted from the terminal and evaluates the validity of the data. Specific NLP libraries (e.g., spaCy and NLTK) are used as the software. 【0699】 Regarding image recognition technology, the server processes images of identity verification documents sent from the terminal and evaluates their authenticity. Specific technologies used here include OpenCV and general cloud-based image recognition services (e.g., AWS Rekognition). This analysis checks for document tampering and inconsistencies. 【0700】 When a user applies for a new contract, the server evaluates the entered information based on existing fraud patterns. If an anomaly is detected and a risk assessment is performed, the server automatically sends a notification to the responsible person using a notification system. This notification is sent via email or the company's internal messaging service. 【0701】 As a concrete example, in a bank loan application, if the information provided by the user resembles past fraud patterns, the system detects the anomaly and immediately sends an alert to the bank representative. An example of a prompt message to the generated AI model in this case would be, "Evaluate the new loan application information and check for similarities to past fraud patterns." 【0702】 This invention reduces the risk of fraudulent activity in the contract process and makes it possible to provide users with safer and more appropriate services. 【0703】 The flow of the specific processing in Example 1 will be explained using Figure 11. 【0704】 Step 1: 【0705】 The server collects past contract information and fraud cases from a database. This database includes basic information about the contractor, detailed contract terms, and past fraud patterns. Using the collected data, the server trains a machine learning model using frameworks such as TensorFlow or PyTorch. This results in a model that has learned the characteristics of fraudulent contracts. 【0706】 Step 2: 【0707】 When a user initiates a new contract, the device inputs subscriber information and identity verification documents and sends them to the server. This input includes name, address, contact information, and scanned images of identity verification documents. The server receives this information, analyzes the text data using NLP libraries such as spaCy or NLTK, and evaluates the accuracy of the information. The resulting analyzed text information is then output. 【0708】 Step 3: 【0709】 The server receives images of identity verification documents sent from the terminal and analyzes them using image recognition technologies such as OpenCV and AWS Rekognition. The server checks the integrity of the images and detects tampering, outputting an evaluation result of their legitimacy. This process confirms that the documents are genuine. 【0710】 Step 4: 【0711】 The server uses the results of text and image analysis to compare the input information with learned fraud patterns. By matching it against previously learned fraud patterns, it identifies anomalies and assesses the risk. The server then calculates a risk score and outputs the result. 【0712】 Step 5: 【0713】 Based on the detection results of an anomaly, the server automatically sends a notification to the responsible person if it determines that notification is necessary according to internal evaluation criteria. The notification content is conveyed via email or messaging system. This notification includes the specific nature of the anomaly and recommended additional actions. The notified information is output. 【0714】 Step 6: 【0715】 The final results and feedback from the contract process are stored in a database. Based on this feedback, the server adjusts and improves the machine learning model to address new fraudulent methods. As a result, an improved model is output. 【0716】 (Application Example 1) 【0717】 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". 【0718】 The problem that this invention aims to solve is to improve security and facilitate smooth procedures for legitimate use by detecting potential fraudulent activity in real time during online transactions and account registration, and warning users and related parties in advance. 【0719】 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. 【0720】 This invention includes a server comprising: machine learning means for recognizing patterns of fraudulent activity by collecting and learning from past transaction information and fraud cases; data analysis means for analyzing data received during the transaction process in real time using natural language processing and visual data recognition technologies and evaluating the validity of the data; and information provision means for warning of potential fraud in real time through the user's digital device. This makes it possible to quickly and automatically assess the risk of fraudulent activity and notify users and stakeholders in advance. 【0721】 "Transaction information" refers to data related to past commercial activities or negotiations, including details such as date, time, conditions, and participants. 【0722】 "Fraudulent activity cases" are actual examples of fraudulent acts that have occurred in the past, and are used as data to learn patterns of fraud and deception. 【0723】 "Machine learning" is a technology in which computers automatically learn patterns from large amounts of data to make future predictions and classifications. 【0724】 "Natural language processing" is a technology that enables computers to understand, analyze, and generate human language. 【0725】 "Visual data recognition technology" is a technology that analyzes visual data such as images and videos and extracts information from it. 【0726】 "Data analysis means" refers to technologies and methods that analyze received data in real time to evaluate its validity and detect anomalies. 【0727】 "Information provision means" refers to methods and technologies for providing users with real-time warnings and guidance regarding potential fraud. 【0728】 A "warning" is a notification that prompts users to take action by informing them in advance of potential risks or fraudulent activities. 【0729】 In the system that implements this application, the server first collects past transaction information and fraud cases, and uses this to train a machine learning model. The server leverages machine learning libraries such as TensorFlow and PyTorch to learn patterns of fraudulent activity from large amounts of data. This process creates a model that can recognize potential patterns of fraud. 【0730】 Next, when a user conducts a transaction, the device (such as a smartphone) collects the transaction information and identity verification data entered by the user. The device sends the entered data to the server, which processes the received data in real time. At this stage, text data is analyzed using natural language processing tools such as Hugging Face Transformers, and the validity of visual data is verified using image recognition libraries such as OpenCV. 【0731】 If the server analyzes transaction information and the data entered by the user resembles a learned fraud pattern, the server performs a risk assessment. If an anomaly is detected, the user is warned in real time through the terminal's information provision system. This notification allows the user to immediately review their input information. In short, the system provides a mechanism to quickly and effectively prevent fraudulent activity. 【0732】 For example, when a user creates a new account while shopping online, if the name and address information they enter matches past fraud cases identified on the server, a warning will be displayed on their smartphone, and the user will be prompted to verify the information. 【0733】 Examples of prompt statements when using a generative AI model are as follows: 【0734】 "Please enter your username, address, and identification documents to assess the risk of fraudulent contracts." 【0735】 Based on this prompt, the AI ​​model evaluates the security of contracts and transactions, helping to proactively prevent fraud risks. 【0736】 The flow of a specific process in Application Example 1 will be explained using Figure 12. 【0737】 Step 1: 【0738】 The user initiates a transaction and enters the necessary information into the terminal. This information includes name, address, payment information, and scanned images of identification documents. The terminal then prepares to send this data to the server. 【0739】 Step 2: 【0740】 The server receives input data sent from the terminal. After receiving the data, it uses natural language processing tools (such as Hugging Face Transformers) to analyze the text data and check for any information similar to malicious patterns. The processed data is then used in the next processing step. 【0741】 Step 3: 【0742】 The server uses visual data recognition technology (such as OpenCV) to analyze the scanned images of identification documents. This process verifies that the images have not been tampered with and that the documents are authentic. The legitimacy of the processed image data is evaluated, and the results are sent to the next step. 【0743】 Step 4: 【0744】 The server analyzes the data and compares it to pre-trained fraud patterns to detect anomalies. A risk score is then calculated for the input data. If the risk score is high, it is marked as an anomaly, and a warning is deemed necessary. 【0745】 Step 5: 【0746】 If an anomaly is detected, the server will send a warning message to the terminal in real time. This message will contain details of the information that needs to be checked. Upon receiving this notification, the user can review the information they entered and make corrections or reconfirmations as needed. 【0747】 Step 6: 【0748】 After all transactions are completed, the execution results and feedback are saved to the server. This data helps identify new fraudulent patterns and areas for improvement, which in turn helps train the generative AI model. As a result, the accuracy of the system will improve in future transactions. 【0749】 In this way, users and servers can work together to provide a secure trading environment while establishing protection against fraudulent activity. 【0750】 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. 【0751】 This invention provides a comprehensive judgment capability that takes into account the user's emotional state by combining an emotion engine with a system that effectively detects and prevents fraudulent contracts. An embodiment thereof is shown below. 【0752】 Data collection and learning 【0753】 First, the server collects historical contract data and past fraud cases. This dataset includes detailed information about contractors, contract terms, and cases marked as fraudulent. Machine learning algorithms are used to learn patterns of fraudulent contracts based on this data. This allows the system to recognize existing fraud patterns and identify risks in new contracts. 【0754】 Real-time data analysis and emotion recognition 【0755】 During the contract process, the terminal not only receives contract information and identity verification documents provided by the user, but also captures the user's intent and emotions through voice or text. This information is transmitted to a server and analyzed by a natural language processing engine. Image recognition technology is also used to verify the authenticity of the submitted documents. 【0756】 Furthermore, an emotion engine analyzes the user's emotional state. This engine analyzes the tone of voice, word choice, and conversational context to recognize the user's emotions in real time. For example, if a user is clearly experiencing stress during a contract, that information could influence the risk assessment. 【0757】 Anomaly detection and notification 【0758】 Based on the analyzed information, the server compares it to previously learned fraud cases and calculates an anomaly risk score for the contract. During this process, user sentiment analysis results are also integrated, and particular attention is paid to any unusual patterns. If the risk is high, the system detects the anomaly and sends a notification to alert the responsible party. 【0759】 Improvement and response 【0760】 In situations deemed high-risk, the server automatically notifies the responsible party of the details and countermeasures. For example, it provides specific suggestions if it determines that additional information should be requested from the user or if a different identity verification procedure is necessary. 【0761】 Continuous model improvement 【0762】 The server collects all feedback from the contract process and uses it to improve the overall model, including the sentiment engine. Newly emerging fraud methods and data on user sentiment improve the model's accuracy and responsiveness. This enables the system to perform more comprehensive fraud detection and prevention, thereby enhancing the security of the service. 【0763】 In this way, by incorporating emotion recognition, the present invention functions as a system that, in addition to detecting fraudulent contracts, improves the user experience and enables the realization of more accurate services. 【0764】 The following describes the processing flow. 【0765】 Step 1: 【0766】 The server collects past contract information and fraud cases from a database, inputs them into a machine learning algorithm, and begins analysis. This builds a dataset that learns common patterns and features in fraudulent contracts. 【0767】 Step 2: 【0768】 The device receives contract information and identity verification documents provided by the user at the start of the contract process. At this time, the user's spoken voice and entered text data are also captured and used for emotion recognition. 【0769】 Step 3: 【0770】 The server processes contract information sent from the terminal in real time. A natural language processing engine analyzes the text data and formats information such as the contract holder's name and address. It also uses image recognition technology to verify whether the submitted identification documents are genuine. 【0771】 Step 4: 【0772】 The server operates an emotion engine that recognizes emotions from the user's voice or written words. It analyzes the user's tone of voice, word choice, and speech rhythm to understand the user's emotional state. This information is also incorporated into the risk assessment of the entire contract process. 【0773】 Step 5: 【0774】 The server compares the analyzed contract information and sentiment recognition results with fraud patterns learned in the past. This comparison calculates a risk score to assess the likelihood that the currently ongoing contract is fraudulent. 【0775】 Step 6: 【0776】 If the server exceeds a certain risk score, it determines that there is a high probability of fraud and automatically sends a notification to the responsible party. This notification includes additional recommended actions based on the user's contract details, detected risks, and emotional state. 【0777】 Step 7: 【0778】 The server collects all feedback after the contract process is complete. This data, including new fraudulent practices and user sentiment, is used to improve the machine learning models and sentiment engine. This process enhances the overall accuracy and reliability of the system. 【0779】 (Example 2) 【0780】 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". 【0781】 In modern contract processes, the risk of fraudulent contracts is increasing, and existing detection methods based on existing technologies are insufficient to prevent them. Furthermore, there is a lack of technology that considers the emotional state of users when determining fraud risk, thus necessitating a more comprehensive contract management approach. Additionally, there is a need for technology that analyzes the emotional state of users during contract procedures and utilizes that information to improve the accuracy of services. 【0782】 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. 【0783】 This invention includes a server comprising: a machine learning means that recognizes patterns of contracts that violate rules by collecting and learning information on past contracts and fraud cases; a data analysis means that analyzes data received during the contract procedure in real time using natural language processing and image recognition technology and evaluates the validity of the information; and an emotion recognition means that analyzes the emotional state from the user's voice and text and reflects that emotional information in the contract risk assessment. This enables comprehensive anomaly detection that takes into account the user's emotional information, which was lacking in conventional fraudulent contract detection technology, thereby reducing fraud risk and improving the accuracy of the service. 【0784】 "Machine learning methods" are technologies that collect information on past contracts and cases of fraud, and use that information to recognize contract patterns that violate regulations. 【0785】 "Data analysis means" refers to a technology that uses natural language processing and image recognition technologies to analyze data received during contract procedures in real time and evaluate the validity of the information. 【0786】 "Emotion recognition means" refers to technology that analyzes a user's emotional state from their voice or text data and incorporates that emotional information into the risk assessment of a contract. 【0787】 An "anomaly detection method" is a technology that detects anomalies and calculates risk by comparing the analyzed information with learned fraud patterns. 【0788】 A "notification method" is a technology that assesses the possibility of a rule violation and automatically notifies the responsible person based on that possibility. 【0789】 "Model improvement methods" refer to techniques for continuously improving the overall model based on the feedback received. 【0790】 This invention provides a system for effectively detecting and preventing fraudulent contracts, enabling comprehensive judgment that takes into account the user's emotional state. By combining machine learning and sentiment analysis technologies, this system aims to reduce the risk of fraud at each stage of the contract process while improving the user experience. 【0791】 The server collects historical contract information and data on fraudulent cases, and uses this data to train machine learning models. Specifically, it uses machine learning libraries such as TensorFlow and PyTorch to train the models. The models are used to recognize fraudulent contract patterns and identify risks in new contracts. 【0792】 The terminal receives contract information and identity verification documents provided by the user during the contract process. The terminal digitizes the documents using optical character recognition (OCR) technology and converts the user's voice into text using a speech recognition system. This information is transmitted to the server in real time and analyzed by a natural language processing engine. 【0793】 Based on the analyzed audio and text data, the server uses an emotion engine to evaluate the user's emotional state. Specifically, it analyzes the tone of voice, word choice, and conversational context to determine what emotions the user is experiencing. The Hugging Face emotion analysis model is used for this evaluation. 【0794】 This system further compares the analyzed information against patterns of fraudulent contracts, detects anomalies, and calculates a risk score. Scikit-learn's anomaly detection algorithm is used for this purpose. If a high risk is determined, the system sends a notification to the responsible party and suggests additional countermeasures. 【0795】 A concrete example is a mortgage application. When a user attempts to sign a mortgage contract, the terminal detects stress from the user's voice. The server evaluates the detected stress level as high risk and sends a notification to the person in charge stating, "The user may be experiencing stress; please consider additional identity verification procedures." In this case, a prompt such as, "Suggest countermeasures to take when a specific emotional state (e.g., anxiety, stress) is detected during the mortgage contract review process," is used as an example input to the generating AI model. 【0796】 By implementing the invention in this way, it is possible to simultaneously achieve improved fraud detection capabilities and enhanced user experience. 【0797】 The flow of the specific processing in Example 2 will be explained using Figure 13. 【0798】 Step 1: 【0799】 The server collects historical contract information and fraud cases from the contract database. The input consists of the contractor's personal information, contract details, and records of fraud cases. The server executes SQL queries to extract this information as a dataset. The output is a fraudulent contract dataset in an analyzable format. 【0800】 Step 2: 【0801】 The server trains a machine learning model using the collected dataset. The input is the fraudulent contract dataset obtained in Step 1. Specifically, the server uses TensorFlow or PyTorch to preprocess the data, select features, and train the model. The output is a trained model that recognizes fraudulent contract patterns. 【0802】 Step 3: 【0803】 The terminal receives contract information, voice recordings, and identification documents entered by the user during the contract process. The input consists of this information provided by the user. The terminal uses OCR technology to digitize documents and applies speech recognition technology to convert speech to text. The output is data in a digital format suitable for processing on the server. 【0804】 Step 4: 【0805】 The server receives digital data transmitted from the terminal and analyzes its content using a natural language processing engine. The input is the text data processed in step 3. The server analyzes the text content and extracts information to determine the legitimacy and potential fraud of the contract. The output is a dataset containing the analysis results. 【0806】 Step 5: 【0807】 The server analyzes the user's emotional state using an emotion engine. The input is the audio and text data obtained in step 3. Specifically, it uses the Hugging Face emotion analysis model to analyze voice tone, word choice, and context. The output is data indicating the user's emotional state. 【0808】 Step 6: 【0809】 The server detects anomalies by comparing the analyzed contract data and sentiment state with a trained model and calculates a risk score. The inputs are the analyzed data from step 4 and the sentiment data from step 5. The server uses Scikit-learn to evaluate the similarity to known fraudulent contract patterns and quantifies the risk of the contract. The output is the risk score. 【0810】 Step 7: 【0811】 The server notifies the responsible person based on the risk score. The input is the risk score obtained in step 6. Notifications are sent via email or the company's internal messaging system, and the responsible person is notified if additional verification work is required based on the risk. The output is the notification message sent to the responsible person. 【0812】 Step 8: 【0813】 The server collects feedback obtained after the contract process and uses it to improve the model. Inputs are actual contract results and user feedback obtained from the entire system. The server incorporates new fraud patterns and sentiment data into the model to improve the system's adaptability. The output is the updated, trained model. 【0814】 (Application Example 2) 【0815】 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". 【0816】 In modern electronic transactions, fraudulent transactions are on the rise, posing a significant security challenge. Furthermore, conventional fraud detection systems often focus solely on transaction patterns and fail to consider the user's emotional state, resulting in inaccurate detection. Therefore, the objective of this invention is to improve the accuracy of fraud detection and realize a safer trading environment. 【0817】 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. 【0818】 In this invention, the server includes machine learning means for collecting and recognizing past transaction information and fraudulent activity cases; data analysis means for analyzing data received during transaction procedures in real time using natural language processing and image recognition technologies and evaluating the legitimacy of the data; and emotion recognition means for analyzing user voice data and evaluating emotional state. This enables highly accurate detection of fraudulent transactions and enhanced security that takes user experience into consideration. 【0819】 "Past transaction information" refers to data on previously conducted transactions that are collected for the purpose of detecting fraudulent activity. 【0820】 "Examples of fraudulent activity" refers to data that shows specific examples of fraudulent transactions or actions that have occurred in the past. 【0821】 "Machine learning methods" are algorithms and techniques that recognize patterns based on data, enabling prediction and classification. 【0822】 "Natural language processing methods" are technologies used to analyze text and speech and understand their meaning and context. 【0823】 "Image recognition technology" is a technology that analyzes image data to recognize objects and characters, and to extract features. 【0824】 "Data analysis methods" refer to techniques and systems for processing various types of data using statistics and algorithms to gain insights. 【0825】 "Emotion recognition methods" are technologies for inferring and evaluating human emotions from things like voice and facial expressions. 【0826】 An "anomaly detection method" is a system for detecting patterns or behaviors that deviate from normal and for evaluating the associated risks. 【0827】 A "notification mechanism" refers to a function or device used by the system to inform relevant parties of anomalies or important information it has detected. 【0828】 "Model improvement methods" are techniques for continuously improving machine learning models based on new data and feedback, thereby increasing their accuracy. 【0829】 The system that implements this application is server-centric and consists of multiple means with various roles. The server first collects past transaction information and fraudulent activity cases, and uses machine learning to recognize patterns of fraudulent transactions based on this data. During the transaction process, the server analyzes the data received from the terminal in real time using natural language processing and image recognition technologies to evaluate the legitimacy of the data. 【0830】 Furthermore, the terminal acquires the user's voice data, and an emotion recognition system evaluates the user's emotional state. This allows the server to compare the input information with learned fraud patterns, detect anomalies, and calculate the risk. If the calculated risk exceeds a certain level, the server immediately and automatically notifies the responsible person through a notification system. 【0831】 The hardware used will primarily consist of devices such as smartphones and tablets, while the software will utilize Google Cloud's Natural Language API and various speech analysis libraries. 【0832】 As a concrete example, the terminal receives audio from the user during payment, the server analyzes the audio data to read the emotion, and if it determines that there is a high probability of fraud, an alert is immediately sent to the person in charge. An example of a prompt sentence to be input into the generating AI model is: "Explain how to proactively detect fraud by analyzing the emotion from the user's voice and context during smartphone payments." 【0833】 The flow of a specific process in Application Example 2 will be explained using Figure 14. 【0834】 Step 1: 【0835】 The server collects past transaction information and fraud cases from a database. Using this data as input, it learns patterns of fraudulent transactions using a machine learning algorithm and outputs the results as a model. This model becomes the new standard for evaluating transaction data. 【0836】 Step 2: 【0837】 The terminal receives data entered by the user during the transaction process. This input data includes voice data, text data, and image data. The terminal then sends this data to the server. 【0838】 Step 3: 【0839】 The server uses natural language processing techniques to analyze the meaning of the received audio and text data. The audio data is converted to text using a speech analysis library, and based on the output, an emotion recognition system evaluates the user's emotional state. This result serves as a basis for determining the user's emotions during the transaction. 【0840】 Step 4: 【0841】 The server analyzes the received image data using image recognition technology to verify the authenticity and integrity of the submitted identity verification documents. Based on the results, it determines whether the data has been forged or tampered with and outputs information to assess the risk of fraud. 【0842】 Step 5: 【0843】 The server compares pre-processed transaction data with historical databases based on processed sentiment data and image analysis results, and uses an anomaly detection algorithm to assess the likelihood of fraud. A risk score is calculated and output as a result. If the likelihood of fraud is high, that information is passed on to the next step. 【0844】 Step 6: 【0845】 Based on the calculated risk score, the server immediately notifies the responsible party of the risk using notification methods. The notification includes detailed information about the detected fraud and countermeasures. This notification facilitates a rapid response. 【0846】 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. 【0847】 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. 【0848】 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. 【0849】 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. 【0850】 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. 【0851】 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. 【0852】 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. 【0853】 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. 【0854】 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." 【0855】 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. 【0856】 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. 【0857】 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. 【0858】 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. 【0859】 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. 【0860】 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. 【0861】 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. 【0862】 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. 【0863】 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. 【0864】 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. 【0865】 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. 【0866】 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. 【0867】 The following is further disclosed regarding the embodiments described above. 【0868】 (Claim 1) 【0869】 A machine learning method that recognizes patterns of fraudulent contracts by collecting and learning from past contract information and fraud cases, 【0870】 A data analysis means that analyzes data received during the contract process in real time using natural language processing and image recognition technologies, and evaluates the validity of the data, 【0871】 An anomaly detection means that compares the input information with learned fraud patterns, detects anomalies, and calculates the risk. 【0872】 A notification mechanism that assesses the possibility of fraud and automatically notifies the person in charge based on that possibility, 【0873】 A model improvement method that continuously improves the model based on feedback, 【0874】 A system that includes this. 【0875】 (Claim 2) 【0876】 The system according to claim 1, comprising analyzing user information during the contract process and formatting its contents into a standard format. 【0877】 (Claim 3) 【0878】 The system according to claim 1, further comprising image recognition means for analyzing the entered identity verification document and verifying its integrity. 【0879】 "Example 1" 【0880】 (Claim 1) 【0881】 A machine learning method that uses a data storage device to collect and learn from past contract information and fraud cases to identify the characteristics of fraudulent contracts, 【0882】 An information analysis means that analyzes information received during new contract procedures in real time using text analysis and image processing technologies and evaluates the validity of the information, 【0883】 An anomaly identification means that compares input data with learned fraudulent features to identify anomalies and calculate the risk level, 【0884】 A notification system that assesses the potential for fraud and automatically notifies relevant parties based on that potential, 【0885】 A model tuning mechanism that continuously improves the model based on feedback obtained from the results of the process, 【0886】 A system that includes this. 【0887】 (Claim 2) 【0888】 The system according to claim 1, which includes analyzing user information during contract procedures and organizing its contents into a normative format. 【0889】 (Claim 3) 【0890】 The system according to claim 1, further comprising image processing means for analyzing an input document relating to personal identification and confirming its consistency. 【0891】 "Application Example 1" 【0892】 (Claim 1) 【0893】 A machine learning method that recognizes patterns of fraudulent activity by collecting and learning from past transaction information and fraud cases, 【0894】 A data analysis means that analyzes data received during the transaction process in real time using natural language processing and visual data recognition technologies, and evaluates the validity of the data, 【0895】 An anomaly detection means that compares the input information with learned fraud patterns, detects anomalies, and calculates the risk. 【0896】 A notification mechanism that assesses the possibility of fraud and automatically notifies the person in charge based on that possibility, 【0897】 A means of providing information that warns of potential fraud in real time through the user's digital device, 【0898】 A model improvement method that continuously improves the model based on feedback, 【0899】 A system that includes this. 【0900】 (Claim 2) 【0901】 The system according to claim 1, comprising analyzing user data during the transaction process and formatting its contents into a standard format. 【0902】 (Claim 3) 【0903】 The system according to claim 1, further comprising visual data recognition means for analyzing the entered identity verification document and verifying its integrity. 【0904】 "Example 2 of combining an emotion engine" 【0905】 (Claim 1) 【0906】 A machine learning method that recognizes patterns in rule-violating contracts by collecting and learning from information on past contracts and cases of misconduct, 【0907】 A data analysis means that analyzes data received during contract procedures in real time using natural language processing and image recognition technologies, and evaluates the validity of the information, 【0908】 An emotion recognition method that analyzes the emotional state from the user's voice and text, and reflects that emotional information in the risk assessment of the contract, 【0909】 An anomaly detection means that compares the analyzed information with learned fraud patterns to detect anomalies and calculate the risk, 【0910】 A notification system that assesses the possibility of rule violations and automatically notifies the responsible person based on that possibility, 【0911】 A model improvement method that continuously improves the overall model based on the feedback obtained, 【0912】 A system that includes this. 【0913】 (Claim 2) 【0914】 The system according to claim 1, which includes analyzing user information during contract procedures and formatting the content into a standard format. 【0915】 (Claim 3) 【0916】 The system according to claim 1, further comprising image recognition means for analyzing the entered documents related to identity verification and confirming their consistency. 【0917】 "Application example 2 when combining with an emotional engine" 【0918】 (Claim 1) 【0919】 A machine learning method that recognizes patterns of fraudulent transactions by collecting and recognizing past transaction information and fraudulent activity cases, 【0920】 A data analysis means that analyzes data received during transaction procedures in real time using natural language processing and image recognition technologies, and evaluates the validity of the data, 【0921】 An emotion recognition means that analyzes the user's voice data and evaluates their emotional state, 【0922】 An anomaly detection means that compares the input information with learned fraud patterns, detects anomalies, and calculates the risk. 【0923】 A notification mechanism that assesses the possibility of fraud and automatically notifies the person in charge based on that possibility, 【0924】 A model improvement method that continuously improves the model based on feedback, 【0925】 A system that includes this. 【0926】 (Claim 2) 【0927】 The system according to claim 1, comprising analyzing user information during a transaction and formatting its contents into a standard format. 【0928】 (Claim 3) 【0929】 The system according to claim 1, further comprising image recognition means for analyzing entered documents related to identity verification and verifying their consistency. [Explanation of symbols] 【0930】 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

[Claim 1] A machine learning method that recognizes patterns of fraudulent contracts by collecting and learning from past contract information and fraud cases, A data analysis means that analyzes data received during the contract process in real time using natural language processing and image recognition technologies, and evaluates the validity of the data, An anomaly detection means that compares the input information with learned fraud patterns, detects anomalies, and calculates the risk. A notification mechanism that assesses the possibility of fraud and automatically notifies the person in charge based on that possibility, A model improvement method that continuously improves the model based on feedback, A system that includes this. [Claim 2] The system according to claim 1, comprising analyzing user information during the contract process and formatting its contents into a standard format. [Claim 3] The system according to claim 1, further comprising image recognition means for analyzing the entered identity verification document and verifying its integrity.

Images