Transaction risk early warning control method and device based on risk measurement model

Abstract

The invention provides a transaction risk early warning control method and device based on a risk measurement model, and relates to the technical field of computer data processing, and the method comprises the steps: obtaining related party data and related transaction data corresponding to a related party; respectively inputting the associated party data and the associated transaction data corresponding to the associated party into an associated party risk measurement model and an associated transaction risk measurement model, and analyzing to obtain an associated party risk control threshold and an associated transaction risk control threshold; multiplying the risk control threshold value of the associated party by a risk control threshold value corresponding to the associated transaction to determine a transaction risk control threshold value; and setting an early warning condition according to the transaction risk control threshold, and when the associated transaction data corresponding to the associated party satisfies the early warning condition, generating an early warning prompt and controlling the associated transaction according to the early warning prompt.

Description

Technical Field

[0001] This invention relates to the field of computer data processing technology, and more particularly to a method and apparatus for transaction risk early warning and control based on a risk measurement model. Background Technology

[0002] This section is intended to provide background or context for embodiments of the present invention. The description herein is not intended to imply that it is prior art simply because it is included in this section.

[0003] In the course of daily operations and management, enterprises proactively verify and confirm related parties and incorporate them into their IT systems for management. Based on the condition that the counterparty is a registered related party, enterprises match and extract related transactions from the transaction data recorded in all business systems and also incorporate them into their IT systems for management. The extracted related transaction information includes: counterparty name, identification information, business type, transaction date, contract amount, transaction amount, transaction balance, and transaction price.

[0004] In existing technologies, the management of related-party transactions requiring monitoring mainly involves manual statistical analysis following steps such as exporting related-party transaction details, filtering transactions of target business types, and performing calculations and analyses to determine whether the conduct of various transactions complies with regulatory requirements. Existing manual statistical analysis methods have the following shortcomings: processing is untimely and inaccurate. In practical applications, manual processing may not be able to meet regulatory requirements regarding review and disclosure in a timely manner, and it is inefficient and prone to errors. Each regulation on related-party transaction management from various regulatory agencies requires separate statistical analysis, resulting in a huge workload and numerous steps in the statistical process. Any errors can lead to deviations in calculation results, affecting judgment and making it difficult to ensure the compliance of related-party transactions.

[0005] Because related-party transaction regulatory agencies are multiple entities, each with its own definition of related parties and the scope of related-party transactions, as well as its management requirements, are diverse and complex. Manual statistical analysis requires distinguishing between regulatory agencies, designing separate calculation rules for each regulation, and performing calculations separately, resulting in a massive workload and low efficiency. Furthermore, the numerous steps in the statistical process mean that any errors can lead to biased results, affecting judgment. In addition, manual statistical analysis cannot be conducted daily or in real-time; it can only be performed at the current point in time, lacking timeliness and accuracy. When special circumstances arise between two manual statistical analyses, such as a sudden surge in transaction volume that directly triggers regulatory requirements regarding review and disclosure, there will be a lack of timely warnings, leading to non-compliant related-party transactions and failure to meet regulatory requirements.

[0006] In summary, there is an urgent need for a technical solution that can overcome the above-mentioned shortcomings, effectively analyze the risks in transactions, improve the accuracy of statistical analysis, and increase processing efficiency. Summary of the Invention

[0007] To address the problems existing in the prior art, this invention proposes a transaction risk early warning and control method and device based on a risk measurement model.

[0008] In a first aspect of this invention, a transaction risk early warning and control method based on a risk measurement model is proposed, comprising:

[0009] Obtain related party data and related party transaction data;

[0010] The related party data and the related party transaction data are respectively input into the related party risk measurement model and the related party transaction risk measurement model. After analysis, the related party risk control threshold and the related party transaction risk control threshold are obtained. The related party risk measurement model and the related party transaction risk measurement model are pre-trained in the following manner:

[0011] For the related party risk measurement model, related party feature data is collected and the risk level of the related party is marked. The marked sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. The machine learning model is mapped from related party to risk level by combining the risk level of the related party. The model is tested using the test set to obtain the related party risk measurement model. The related party risk control threshold is determined based on the risk level of the related party output by the model.

[0012] For the related-party transaction risk measurement model, related-party transaction feature data is collected, and the risk control thresholds corresponding to the related-party transactions are marked. The marked sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. The machine learning model is then mapped from related-party transactions to risk control thresholds by combining the risk control thresholds corresponding to the related-party transactions. Finally, the model is tested using the test set to obtain the related-party transaction risk measurement model.

[0013] The transaction risk control threshold is determined by multiplying the related party risk control threshold and the risk control threshold corresponding to the related transaction.

[0014] Early warning conditions are set according to the transaction risk control threshold. When the related transaction data of the related party meets the early warning conditions, an early warning prompt is generated and the related transaction is controlled according to the early warning prompt.

[0015] In a second aspect of the present invention, a transaction risk early warning and control device based on a risk measurement model is proposed, comprising:

[0016] The data acquisition module is used to acquire data from related parties and related transaction data of those related parties.

[0017] The risk measurement and analysis module is used to input the related party data and the related party transaction data into the related party risk measurement model and the related party transaction risk measurement model, respectively, and obtain the related party risk control threshold and the related party transaction risk control threshold through analysis; wherein, the related party risk measurement model and the related party transaction risk measurement model are pre-trained in the following manner:

[0018] For the related party risk measurement model, related party feature data is collected and the risk level of the related party is marked. The marked sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. The machine learning model is mapped from related party to risk level by combining the risk level of the related party. The model is tested using the test set to obtain the related party risk measurement model. The related party risk control threshold is determined based on the risk level of the related party output by the model.

[0019] For the related-party transaction risk measurement model, related-party transaction feature data is collected, and the risk control thresholds corresponding to the related-party transactions are marked. The marked sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. The machine learning model is then mapped from related-party transactions to risk control thresholds by combining the risk control thresholds corresponding to the related-party transactions. Finally, the model is tested using the test set to obtain the related-party transaction risk measurement model.

[0020] The risk control threshold determination module is used to multiply the related party risk control threshold and the risk control threshold corresponding to the related transaction to determine the transaction risk control threshold;

[0021] The transaction early warning module is used to set early warning conditions based on the transaction risk control threshold. When the related transaction data of the related party meets the early warning conditions, an early warning prompt is generated and the related transaction is controlled according to the early warning prompt.

[0022] In a third aspect of the present invention, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a transaction risk early warning and control method based on a risk measurement model.

[0023] In a fourth aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, which, when executed by a processor, implements a transaction risk early warning and control method based on a risk measurement model.

[0024] In a fifth aspect of the present invention, a computer program product is proposed, the computer program product comprising a computer program, which, when executed by a processor, implements a transaction risk early warning and control method based on a risk measurement model.

[0025] The present invention proposes a transaction risk early warning and control method and device based on a risk measurement model. This method and device acquires related party data and related transaction data corresponding to those related parties. The related party data and related transaction data are input into a related party risk measurement model and a related transaction risk measurement model, respectively. After analysis, the related party risk control threshold and the related transaction risk control threshold are obtained. The related party risk measurement model and the related transaction risk measurement model are pre-trained as follows: For the related party risk measurement model, related party feature data is collected, and the risk levels of the related parties are labeled. The labeled sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. Combined with the risk levels of the related parties, the machine learning model performs a mapping learning from related parties to risk levels. The model is then tested using the test set to obtain the related party risk measurement model. The related party risk control threshold is determined based on the risk levels of the related parties output by the model. For the related transaction risk measurement model, related transaction feature data is collected, and the risk control thresholds of the related transactions are labeled. The labeled sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. By combining the risk control thresholds corresponding to the related-party transactions, the machine learning model is mapped from related-party transactions to risk control thresholds, and tested using the test set to obtain a related-party transaction risk measurement model. The related-party risk control threshold and the risk control threshold corresponding to the related-party transaction are multiplied to determine the transaction risk control threshold. Early warning conditions are set according to the transaction risk control threshold. When the related-party transaction data corresponding to the related party meets the early warning conditions, an early warning prompt is generated, and the related-party transaction is controlled according to the early warning prompt. The overall solution automatically completes the early warning analysis of the related-party transactions that need to be monitored, significantly reducing the workload of personnel in related-party transaction management positions, reducing statistical errors, avoiding statistical caliber errors caused by personnel handover or replacement, improving control accuracy, and enhancing work efficiency. This invention uses a related-party risk measurement model and a related-party transaction risk measurement model, which can distinguish different risk levels and set corresponding control thresholds with quantitative indicators. By combining the two control thresholds, the actual control threshold for each related party in various types of related-party transactions that need to be monitored can be determined individually according to the actual risk situation. Through automatic calculation and timely early warning, the pre-control of related-party transactions that need to be monitored is achieved, ensuring early warning, timely response, and reducing the risk of violations, providing strong technical support for risk early warning. Attached Figure Description

[0026] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram of a transaction risk early warning and control method based on a risk measurement model according to an embodiment of the present invention.

[0028] Figure 2 This is a schematic diagram illustrating the specific process of information collection according to an embodiment of the present invention.

[0029] Figure 3 This is a schematic diagram illustrating the specific process of related party data preprocessing according to an embodiment of the present invention.

[0030] Figure 4 This is a schematic diagram of the training process of a related party risk measurement model according to an embodiment of the present invention.

[0031] Figure 5 This is a schematic diagram illustrating the specific process of related party data preprocessing according to an embodiment of the present invention.

[0032] Figure 6 This is a schematic diagram of the training process of a related-party transaction risk measurement model according to an embodiment of the present invention.

[0033] Figure 7 This is a schematic diagram illustrating the specific process of transaction risk warning and control according to an embodiment of the present invention.

[0034] Figure 8 This is a flowchart illustrating an embodiment of the present invention.

[0035] Figure 9 This is a schematic diagram of the architecture of a transaction risk early warning and control device based on a risk measurement model according to an embodiment of the present invention.

[0036] Figure 10 This is a schematic diagram of the architecture of a transaction risk early warning and control device based on a risk measurement model, according to another embodiment of the present invention.

[0037] Figure 11 This is a schematic diagram of a computer device structure according to an embodiment of the present invention. Detailed Implementation

[0038] The principles and spirit of the invention will now be described with reference to several exemplary embodiments. It should be understood that these embodiments are given merely to enable those skilled in the art to better understand and implement the invention, and are not intended to limit the scope of the invention in any way. Rather, these embodiments are provided to make this disclosure more thorough and complete, and to fully convey the scope of this disclosure to those skilled in the art.

[0039] Those skilled in the art will recognize that embodiments of the present invention can be implemented as a system, apparatus, device, method, or computer program product. Therefore, this disclosure can be specifically implemented in the following forms: entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0040] According to an embodiment of the present invention, a transaction risk early warning and control method and device based on a risk measurement model are proposed, which relates to the field of computer data processing technology.

[0041] In the embodiments of the present invention, the following terms need to be explained:

[0042] Regulatory agencies: Agencies that promulgate regulations and systems related to the management of related-party transactions;

[0043] Regulatory rules: Specific provisions regarding the management of related-party transactions in the rules and regulations issued by regulatory authorities concerning related-party transactions;

[0044] Related parties: Natural persons, legal persons, or unincorporated organizations that are related to the enterprise as defined in the regulatory rules that the enterprise is required to comply with;

[0045] Related party transactions refer to transactions between an enterprise and its subsidiaries and related parties involving the transfer of resources, services, obligations or other benefits.

[0046] The principles and spirit of the present invention will be explained in detail below with reference to several representative embodiments.

[0047] Figure 1 This is a schematic flowchart of a transaction risk early warning and control method based on a risk measurement model, according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0048] S101, Obtain related party data and related party transaction data;

[0049] S102, the related party data and the related transaction data corresponding to the related parties are respectively input into the related party risk measurement model and the related transaction risk measurement model, and the related party risk control threshold and the related transaction risk control threshold are obtained through analysis;

[0050] The related party risk measurement model and the related party transaction risk measurement model are pre-trained in the following manner:

[0051] For the related party risk measurement model, related party feature data is collected and the risk level of the related party is marked. The marked sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. The machine learning model is mapped from related party to risk level by combining the risk level of the related party. The model is tested using the test set to obtain the related party risk measurement model. The related party risk control threshold is determined based on the risk level of the related party output by the model.

[0052] For the related-party transaction risk measurement model, related-party transaction feature data is collected, and the risk control thresholds corresponding to the related-party transactions are marked. The marked sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. The machine learning model is then mapped from related-party transactions to risk control thresholds by combining the risk control thresholds corresponding to the related-party transactions. Finally, the model is tested using the test set to obtain the related-party transaction risk measurement model.

[0053] S103, Multiply the related party risk control threshold and the risk control threshold corresponding to the related transaction to determine the transaction risk control threshold;

[0054] S104, set early warning conditions according to the transaction risk control threshold, and when the related transaction data of the related party meets the early warning conditions, generate an early warning prompt and control the related transaction according to the early warning prompt.

[0055] This invention can automatically perform early warning analysis of related-party transactions that need to be monitored, significantly reducing the workload of personnel in related-party transaction management positions, reducing statistical errors, avoiding statistical caliber errors caused by personnel handover or replacement, improving control accuracy, and enhancing work efficiency. This invention adopts a related-party risk measurement model and a related-party transaction risk measurement model, which can distinguish different risk levels and set corresponding control thresholds with quantitative indicators. By combining the two control thresholds, the actual control threshold for each related party in various types of related-party transactions that need to be monitored can be determined in a personalized manner according to the actual risk situation. Through automatic calculation and timely early warning, the invention achieves advance control of related-party transactions that need to be monitored, ensuring early warning and timely response, reducing the risk of violations, and providing strong technical support for risk early warning.

[0056] To provide a clearer explanation of the above-mentioned transaction risk early warning and control method based on the risk measurement model, each step will be explained in detail below.

[0057] In one embodiment, reference Figure 2This is a schematic diagram illustrating the specific process of information collection according to an embodiment of the present invention, such as... Figure 2 As shown, the method includes:

[0058] S201, Collect a list of related party customers;

[0059] S202, Based on the aforementioned list of related party customers, compile related party information and related party transaction information corresponding to the related parties.

[0060] Enterprises build IT systems to collect lists of related-party customers. Based on the condition that the counterparty is a registered related party, they proactively identify related-party transactions and incorporate them into the IT system for management. Based on this information, and in accordance with related-party transaction regulatory rules, combined with the enterprise's registered related-party and transaction information, machine learning methods can be used to establish separate risk measurement models for related-party transactions and related parties, thereby enabling effective analysis, early warning, and control of transaction risks.

[0061] Specifically, enterprises establish IT systems capable of storing and managing detailed information on individuals and organizations related to the enterprise, processing and storing various data on related parties, such as transaction data and business relationships. Through the IT system, enterprises collect information on all related parties and create a comprehensive customer list. This list includes all natural persons, legal persons, or unincorporated organizations with direct or indirect relationships to the enterprise. Counterparties are identified as related parties according to regulatory rules. Any transaction with these related parties is considered a related-party transaction. After identifying related-party transactions, this transaction information also needs to be incorporated into the IT system for management. Management includes recording detailed transaction information, such as counterparty name, identification information, business type, transaction date, and contract amount. The accuracy and timeliness of information can be ensured by regularly updating related-party and related-party transaction data. Data maintenance includes data cleaning, deduplication, updating, and archiving. The IT system is used to monitor related-party transactions, ensuring all transactions comply with regulatory requirements, automatically detecting potential compliance issues, and prompting management intervention.

[0062] In one embodiment, reference Figure 3 This is a schematic diagram illustrating the specific process of related party data preprocessing according to an embodiment of the present invention, as follows: Figure 3 As shown, the method includes:

[0063] S301, In accordance with the transaction supervision requirements of regulatory agencies, collect historical transaction information of related parties under different calibers, including at least transaction size and frequency, and establish a related party data mart;

[0064] S302, filter and clean related parties, including extracting related party data with related transactions in historical transaction information, removing invalid and redundant data, and unifying and standardizing the cleaned related party data.

[0065] S303. Based on the related party data mart, the related party risk measurement score is determined through multiple dimensions, including the number of enterprises corresponding to the related party, the frequency of transactions, the number of related party transactions with the highest single amount exceeding the set limit within a set time period, the probability of related party transactions of a preset business type, and the fair value of related party transactions within a set time period.

[0066] S304. Based on the related party risk measurement score, multiple risk levels are obtained for the related parties. For each risk level, a corresponding related party risk control threshold is set.

[0067] Specifically, for the related party risk measurement model, related party feature data is collected, and the risk level of the related party is marked. The marked sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. The machine learning model is then mapped from related party to risk level using the risk level of the related party. The model is tested using the test set to obtain the related party risk measurement model. The related party risk control threshold is determined based on the risk level of the related party output by the model.

[0068] refer to Figure 4 The specific process includes:

[0069] S401, For the related party risk measurement model, collect related party characteristic data corresponding to the related parties. Among them, feature extraction and analysis are performed through business dimension feature values ​​including related party role, business type, transaction scale, and transaction frequency to determine the related party characteristic data, which serves as the sample set for model training.

[0070] S402, use Bayesian algorithm to train the model, construct the classification prediction of the risk level and the characteristic data of the related parties, and obtain the related party risk measurement model.

[0071] S403, the related party risk measurement model is tested, the feature dimensions of the related parties are automatically corrected according to the test results, the feature data of the related parties are optimized, and the model parameters are adjusted manually.

[0072] Specifically, in accordance with the related-party transaction regulatory rules of various regulatory agencies, basic data on related parties are established based on the specific circumstances of related-party transactions that need to be monitored in previous years, including transaction size and frequency.

[0073] Data cleaning is performed on related parties (extracting related party data with related party transactions within the historical sample period and removing invalid and redundant data). Then, the cleaned related party data is unified and standardized.

[0074] Leveraging a data mart, the system automatically calculates a risk measurement score K for each related party based on five dimensions: the number of companies within the related party's group, the frequency of past transactions, the number of related party transactions in the past 12 months with a maximum single amount exceeding a set amount (e.g., 200 million yuan), the likelihood of significant related party transactions within the related party's group, and the fairness of related party transactions in the past 12 months. This score can be manually queried and reviewed, and adjusted as needed. Based on the score, related parties are categorized and tagged, for example, into three levels: Risk Level 1 - Low Risk, Risk Level 2 - Medium Risk, and Risk Level 3 - High Risk. Furthermore, a related party control threshold Y2 is determined for each different risk level.

[0075] We establish business dimension features such as related party roles (e.g., counterparties, issuers, underwriters, etc.), business types (credit, service, profit transfer, etc.), transaction size (amount), and transaction frequency (number of transactions within one year) for feature extraction and analysis, and construct a correlation matrix of dimensional features to prepare for subsequent training of related party risk score models.

[0076] The Bayesian algorithm is used to train the model, constructing a classification prediction of related party risk levels and related party chapter dimension features, obtaining risk level classification results, and determining the related party control threshold Y2. Simultaneously, based on human settings and verification results, the characteristic dimensions of related parties are automatically adjusted (e.g., transaction frequency is categorized by month), optimizing the related party risk measurement parameters. Human-machine training is strengthened, and subsequent adjustments are made manually based on machine learning. When the existing hierarchical management system cannot meet the requirements of refined management, the risk level can be further subdivided into n levels. Referring to Table 1, the correspondence between the related party risk measurement score interval, the related party risk level, and the related party control threshold is illustrated.

[0077] Table 1

[0078] Related party risk measurement score range Related Party Risk Level Related party control threshold k1= <K<=100 Risk Level 1 Y2-1 k2=<K<k1 Risk Level 2 Y2-2 … … … k9= <K<k8 Risk level 9 Y2-9 0=<K<k9 Risk level 10 Y2-10

[0079] Related party risk levels and thresholds are recorded separately by regulatory agencies, as shown in Table 2, which illustrates the corresponding relationships.

[0080] Table 2

[0081]

[0082]

[0083] In one embodiment, reference Figure 5 This is a schematic diagram illustrating the specific process of related party data preprocessing according to an embodiment of the present invention, as follows: Figure 5 As shown, the method includes:

[0084] S501, in accordance with the transaction supervision requirements of regulatory authorities, determine the transaction type, amount, business type to be included in the calculation, and calculation mechanism of related transactions;

[0085] S502, Based on the aforementioned calculation mechanism, and in conjunction with the transaction type, amount, and business type corresponding to the transaction type of the related transaction that needs to be included in the calculation, the historical transaction data is analyzed to determine the transaction scale and increase / decrease of the related transaction, and a related transaction data mart is established.

[0086] S503 filters and cleans the business data of related transactions, including removing invalid and redundant data, and performing data transformation, integration, discretization and normalization.

[0087] S504. Based on the related party transaction data mart, determine the risk control threshold corresponding to the related party transaction through multiple dimensions, including related party role, business type, transaction scale, risk limit and increase / decrease range.

[0088] Specifically, for the related-party transaction risk measurement model, related-party transaction feature data is collected, and the risk control thresholds corresponding to the related-party transactions are marked. The marked sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. Combined with the risk control thresholds corresponding to the related-party transactions, the machine learning model is mapped from related-party transactions to risk control thresholds. The model is then tested using the test set to obtain the related-party transaction risk measurement model.

[0089] refer to Figure 6 The specific process includes:

[0090] S601, For the related-party transaction risk measurement model, collect related-party transaction characteristic data corresponding to related-party transactions. Among them, feature extraction analysis is performed through business dimension feature values, including related party roles, business types, transaction scale, risk limits and increase / decrease ranges, to determine related-party transaction characteristic data, which serves as the sample set for model training.

[0091] S602 trains the model using the support vector machine algorithm, uses a Gaussian kernel as the kernel function to optimize the model's accuracy and generalization ability, and uses the F1 score to evaluate the model.

[0092] Specifically, based on the related-party transaction regulatory rules of various regulatory agencies, the types of related-party transactions that need to be monitored, the corresponding amounts, the business categories that need to be included in the calculation under the types of related-party transactions that need to be monitored, and the calculation rules are clearly defined.

[0093] Based on the calculation rules corresponding to various regulatory provisions, and combined with the specific details of all business types under the related-party transaction types that financial institutions need to monitor in previous years, including basic data dimensions such as transaction size and increase / decrease, as well as external financial information such as industrial and commercial information and banking and insurance information, a related-party transaction risk data mart is established on the basis of a big data cloud platform.

[0094] Based on the related-party transaction risk data mart, and the related party list and related-party transaction business type parameters established by the enterprise in the IT system, business data is filtered and cleaned (invalid and redundant data is removed) to obtain related-party transaction data that meets business needs.

[0095] Transform and integrate the data to ensure data quality, accuracy, and reliability; then, discretize and reduce the integrated data (according to business management requirements, such as discretizing by time intervals or by frequency discretizing by the number of transactions to obtain concise and effective related transaction data), unify and standardize the basic data of related transactions, and improve the quality and usability of related transaction data.

[0096] By extracting and selecting features of related-party transactions, business dimension feature values ​​such as related party roles, business types, transaction scale, risk limits, and increases / decreases are established. This reduces the dimensionality of the dataset, lowers the computational complexity and memory consumption of subsequent model training, and prepares for the entire feature engineering and model training of related-party transactions.

[0097] Support Vector Machines (SVMs) excel at handling high-dimensional and non-linear data, and since related transaction data often exhibits non-linear characteristics, this invention uses the SVM algorithm for model training. Because linear kernel functions are unsuitable for handling non-linear data, and the sigmoid kernel performs poorly, this invention employs a Gaussian kernel to optimize the model's accuracy and generalization ability. The F1 score is then used as the evaluation metric to assess the model.

[0098] The overall backtest of the SVM strategy for related transactions is as follows:

[0099] 1. Transaction backtesting settings;

[0100] 1.1 The types, amounts, and control thresholds for related-party transactions that need to be monitored;

[0101] 1.2 Backtesting period: 20220101-20241231;

[0102] 1.3 Backtesting frequency: 1 month;

[0103] 2. Transaction type selection;

[0104] Tests were conducted using four main transaction types selected from the related transaction system: credit transactions, service transactions, benefit transfer transactions, and other transactions.

[0105] 3. Trading strategy signals;

[0106] Feature selection: Monthly related transactions are aggregated as features to calculate the control thresholds corresponding to the types of related transaction businesses;

[0107] 4. Transaction model evaluation;

[0108] Model evaluation method: F1 score;

[0109] In practical applications, it's crucial to avoid overlooking any genuine related transaction types. Omitting even one could lead to non-compliance with regulatory requirements, and low accuracy could result in legitimate transactions being incorrectly labeled. F1, the harmonic mean of precision and recall, effectively balances these two factors. Therefore, this invention employs the F1 rating model for performance. The F1 score is an evaluation metric used in classification problems to measure model performance; it comprehensively considers both precision and recall, providing a single value to assess model accuracy. The F1 score is defined as the harmonic mean of precision and recall, calculated as follows:

[0110]

[0111] Precision: The proportion of cases predicted as positive that are actually positive, calculated using the following formula: TP stands for True Positives, and FP stands for False Positives.

[0112] Recall: The proportion of positive classes that are actually predicted to be positive. The formula is: FN stands for False Negatives.

[0113] The F1 score balances precision and recall, making it suitable for imbalanced positive and negative sample situations. In such cases, precision or recall alone may not be sufficient to comprehensively evaluate model performance. In multi-class classification problems, an F1 score can be calculated for each class, and then the average of these scores is used to obtain an overall performance evaluation of the model. During model training, the F1 score can serve as an evaluation metric to help select the optimal model parameters and architecture. By tuning the model to maximize the F1 score, a model that achieves the best balance between precision and recall can be found. The requirements for precision and recall may differ in different business scenarios. For example, in some scenarios, avoiding misclassifying positive classes as negative classes (i.e., improving recall) may be more important than accurately identifying positive classes (i.e., improving precision). By adjusting the calculation of the F1 score, these different business needs can be adapted. The F1 score provides an intuitive numerical value, allowing operators to directly understand the model's performance.

[0114] Based on the related-party transaction risk measurement model, the corresponding control threshold Y1 is automatically generated. These thresholds are calculated and recorded separately according to the regulatory authorities, as shown in Table 3.

[0115] Table 3

[0116]

[0117] In one embodiment, for S103, the related party risk control threshold and the risk control threshold corresponding to the related transaction are multiplied together to determine the transaction risk control threshold.

[0118] Based on the calculation results of the related-party transaction risk measurement model and the related-party risk measurement model, the actual control threshold Y3 for related-party transactions that need to be monitored is obtained by multiplying the related-party control threshold Y1 by the related-party transaction control threshold Y2. Table 4 illustrates the correspondence between the actual control thresholds for related-party transactions that need to be monitored.

[0119] Table 4

[0120]

[0121] In one embodiment, for S104, an early warning condition is set according to the transaction risk control threshold. When the related transaction data corresponding to the related party meets the early warning condition, an early warning prompt is generated and the related transaction is controlled according to the early warning prompt.

[0122] For details, please refer to Figure 7 This is a schematic diagram illustrating the specific process of transaction risk warning and control according to an embodiment of the present invention. Figure 7 As shown, the method includes:

[0123] S701, Multiply the amount of the related party transactions that need to be monitored by the transaction risk control threshold to obtain the warning line of the related party, which serves as the warning condition;

[0124] S702: When the transaction volume of a related party transaction that needs to be monitored reaches the warning threshold of the related party, an early warning is generated and pushed to the related party transaction processing personnel, who then notify the relevant business departments to carry out risk control and transaction processing.

[0125] Multiplying the amount of related-party transactions requiring monitoring by the actual control threshold Y3 yields the corresponding regulatory warning line for a specific related party. When Company X, a related party under Regulatory Authority A, engages in any business type of transaction requiring monitoring, and the overall transaction volume reaches the corresponding warning line, the IT system will generate a warning alert and push it to the related-party transaction management personnel. Upon receiving the warning alert, the related-party transaction management personnel will notify the relevant business departments to pay attention and develop enhanced risk control plans to ensure the compliant conduct of related-party transactions.

[0126] In one embodiment, the enterprise sets up an IT system to collect a list of related party customers; establishes a related party transaction risk measurement model and a related party risk measurement model according to regulatory rules; trains the models and conducts risk assessments using machine learning methods (SVM and Bayes algorithm); extracts the characteristics of related party transactions and related parties, and classifies them into risk levels; determines the actual control thresholds for related party transactions that need to be monitored based on the model results; the IT system generates early warning prompts, and management personnel take action based on the early warnings.

[0127] refer to Figure 8 This is a schematic diagram of the architectural relationship of a risk measurement model according to an embodiment of the present invention, as shown below. Figure 8 As shown, this architecture utilizes big data and machine learning to manage and control related-party transaction risks. Business channels, including corporate lending, financial markets, interbank asset management, and subsidiaries, fall under the business areas where the company conducts transactions.

[0128] The related-party transaction big data cloud platform is a centralized platform for processing and storing related-party transaction data. Data sources include business data from corporate credit lines, interbank asset management, and subsidiaries, as well as regulatory information from business registration authorities, the People's Bank of China, and the China Banking and Insurance Regulatory Commission.

[0129] In the related-party transaction risk measurement model laboratory, machine learning algorithms (such as SVM) are used for risk assessment and model training of related-party transactions. Related-party transaction risk source data serves as the foundational data source for model training and evaluation. The related-party transaction risk source data undergoes cleaning and format conversion to facilitate analysis and ensure data quality and consistency; standardized and unified basic data is obtained through cleaning and conversion. Further features of related parties and related-party transaction risk dimensions are extracted for risk assessment. Machine learning techniques such as Support Vector Machine (SVM) are used for model training, evaluation, and optimization.

[0130] Through related-party transaction applications, identified related-party transactions are used for further risk control and management. Finally, target business related-party transaction control parameters and thresholds are automatically generated: the system can automatically set control parameters and thresholds for related-party transactions to facilitate supervision and risk control; the overall architecture improves the efficiency and accuracy of related-party transaction management through automation and intelligence.

[0131] According to the specific regulations of the regulatory authorities, the amount of related-party transactions that need to be monitored is determined. By setting up multi-dimensional indicators, risk measurement models for related-party transactions and related-party transactions are established to distinguish different risk levels. Corresponding control thresholds are set with quantitative indicators, and the related-party transactions that need to be monitored are controlled in advance through automatic calculation and timely early warning by the system.

[0132] The transaction risk early warning and control method based on a risk measurement model proposed in this invention enables enterprises to better manage related-party transactions, improve risk management capabilities, and ensure the stable operation of their businesses. The overall solution improves the accuracy and efficiency of identification through automated identification of related parties and related-party transactions, reducing errors and omissions from manual operations. Utilizing big data and machine learning technologies, it can more accurately assess and manage related-party transaction risks, providing enterprises with more effective risk control measures. It can monitor related-party transactions in real time and issue timely warnings once potential risks are detected, helping enterprises respond quickly. Automated generation of control parameters and thresholds helps ensure the compliance of related-party transactions and reduces the occurrence of violations. Quantitative risk assessment provides managers with data support for making more rational business decisions. The automated and intelligent processing flow reduces reliance on manual labor, thereby lowering labor and operational costs.

[0133] This invention's big data platform can process and analyze large amounts of complex data, improving enterprises' data utilization efficiency. Simultaneously, through continuous model training, evaluation, and optimization, it can continuously improve the model's predictive accuracy and generalization ability. It can be adjusted according to different business needs and regulatory requirements, exhibiting good flexibility and scalability. In practical application scenarios, it integrates information from various sources, such as business registration information, People's Bank of China information, and China Banking and Insurance Regulatory Commission information, providing comprehensive data support for risk assessment.

[0134] It should be noted that although the operation of the method of the present invention has been described in a specific order in the above embodiments and figures, this does not require or imply that the operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0135] After introducing the method of exemplary embodiments of the present invention, the following references are made. Figure 9 This invention provides an exemplary embodiment of a transaction risk early warning and control device based on a risk measurement model.

[0136] The implementation of the transaction risk early warning and control device based on the risk measurement model can refer to the implementation of the above method, and the repeated parts will not be described again. The term "module" or "unit" used below can be a combination of software and / or hardware to achieve a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0137] Based on the same inventive concept, this invention also proposes a transaction risk early warning and control device based on a risk measurement model, such as... Figure 9 As shown, the device includes:

[0138] The data acquisition module 910 is used to acquire related party data and related party transaction data.

[0139] The risk measurement and analysis module 920 is used to input the related party data and the related party transaction data into the related party risk measurement model and the related party transaction risk measurement model, respectively, and obtain the related party risk control threshold and the related party transaction risk control threshold through analysis; wherein, the related party risk measurement model and the related party transaction risk measurement model are pre-trained in the following manner:

[0140] For the related party risk measurement model, related party feature data is collected and the risk level of the related party is marked. The marked sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. The machine learning model is mapped from related party to risk level by combining the risk level of the related party. The model is tested using the test set to obtain the related party risk measurement model. The related party risk control threshold is determined based on the risk level of the related party output by the model.

[0141] For the related-party transaction risk measurement model, related-party transaction feature data is collected, and the risk control thresholds corresponding to the related-party transactions are marked. The marked sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. The machine learning model is then mapped from related-party transactions to risk control thresholds by combining the risk control thresholds corresponding to the related-party transactions. Finally, the model is tested using the test set to obtain the related-party transaction risk measurement model.

[0142] The risk control threshold determination module 930 is used to multiply the related party risk control threshold and the risk control threshold corresponding to the related transaction to determine the transaction risk control threshold;

[0143] The transaction early warning module 940 is used to set early warning conditions according to the transaction risk control threshold. When the related transaction data of the related party meets the early warning conditions, an early warning prompt is generated and the related transaction is controlled according to the early warning prompt.

[0144] In one embodiment, the data acquisition module 910 is specifically used for:

[0145] Collect a list of related party customers;

[0146] Based on the aforementioned list of related party customers, statistics on related party information and related party transaction information are compiled.

[0147] In one embodiment, reference Figure 10 The device also includes: a related party data processing module 950;

[0148] The related party data processing module 950 is specifically used for:

[0149] In accordance with the transaction supervision requirements of regulatory agencies, historical transaction information of related parties under different criteria will be collected, including at least the transaction size and frequency, and a related party data mart will be established.

[0150] Related parties are filtered and cleaned, including extracting related party data with related transactions in historical transaction information, removing invalid and redundant data, and unifying and standardizing the cleaned related party data.

[0151] Based on the related party data mart, the related party risk measurement score is determined through multiple dimensions, including the number of related parties corresponding to the related party, the frequency of transactions, the number of related party transactions with the highest single amount exceeding the set limit within a set time period, the probability of related parties engaging in related party transactions of a preset business type, and the fair value of related party transactions within a set time period.

[0152] Based on the risk measurement score of the related parties, multiple risk levels are obtained for the related parties. For each risk level, a corresponding related party risk control threshold is set.

[0153] In one embodiment, the risk measurement and analysis module 920 is specifically used for:

[0154] For the related party risk measurement model, related party characteristic data is collected. In this model, feature extraction and analysis are performed using business dimension feature values, including related party role, business type, transaction size, and transaction frequency, to determine the related party characteristic data, which serves as the sample set for model training.

[0155] The Bayesian algorithm is used to train the model, and the risk level and related party feature data of the associated parties are classified and predicted to obtain the associated party risk measurement model.

[0156] The related party risk measurement model is tested, and the feature dimensions of the related parties are automatically corrected based on the test results. The feature data of the related parties are optimized, and the model parameters are adjusted manually.

[0157] In one embodiment, reference Figure 10 The device also includes: a related transaction data processing module 960;

[0158] The related transaction data processing module 960 is specifically used for:

[0159] In accordance with the transaction supervision requirements of regulatory authorities, determine the transaction type, amount, corresponding business type to be included in the calculation, and calculation mechanism of related transactions;

[0160] Based on the aforementioned calculation mechanism, and in conjunction with the transaction type, amount, and business type corresponding to the transaction type that needs to be included in the calculation, historical transaction data is analyzed to determine the transaction scale and increase / decrease of related transactions, and a related transaction data mart is established.

[0161] The business data of related transactions is filtered and cleaned, including the removal of invalid and redundant data, and the data is transformed, integrated, discretized and normalized.

[0162] Based on the related-party transaction data mart, the risk control thresholds corresponding to related-party transactions are determined through multiple dimensions, including the role of the related party, business type, transaction size, risk limit, and the magnitude of increase or decrease.

[0163] In one embodiment, the risk measurement and analysis module 920 is specifically used for:

[0164] For the related-party transaction risk measurement model, related-party transaction characteristic data is collected. In this model, feature extraction and analysis are performed on the business dimension feature values, including the related party role, business type, transaction size, risk limit and increase / decrease range, to determine the related-party transaction characteristic data, which serves as the sample set for model training.

[0165] The model is trained using the support vector machine algorithm, and the Gaussian kernel is used as the kernel function to optimize the model's accuracy and generalization ability. The model is also evaluated using the F1 score.

[0166] In one embodiment, the transaction early warning module 940 is specifically used for:

[0167] Multiply the amount of related transactions that need to be monitored by the transaction risk control threshold to obtain the warning line for related parties, which serves as the warning condition.

[0168] When the transaction volume of a related party transaction that requires monitoring reaches the related party's warning threshold, an early warning is generated and pushed to the related party transaction processing personnel, who then notify the relevant business departments to carry out risk control and transaction processing.

[0169] It should be noted that although several modules of the transaction risk early warning and control device based on the risk measurement model have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of the present invention, the features and functions of two or more modules described above can be embodied in one module. Conversely, the features and functions of one module described above can be further divided and embodied by multiple modules.

[0170] Based on the aforementioned inventive concept, such as Figure 11 As shown, the present invention also proposes a computer device 1100, including a memory 1110, a processor 1120, and a computer program 1130 stored in the memory 1110 and executable on the processor 1120. When the processor 1120 executes the computer program 1130, it implements the aforementioned transaction risk early warning and control method based on the risk measurement model.

[0171] Based on the aforementioned inventive concept, the present invention proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the aforementioned transaction risk early warning and control method based on a risk measurement model.

[0172] Based on the aforementioned inventive concept, this invention proposes a computer program product, which includes a computer program that, when executed by a processor, implements a transaction risk early warning and control method based on a risk measurement model.

[0173] The present invention proposes a transaction risk early warning and control method and device based on a risk measurement model. This method and device acquires related party data and related transaction data corresponding to those related parties. The related party data and related transaction data are input into a related party risk measurement model and a related transaction risk measurement model, respectively. After analysis, the related party risk control threshold and the related transaction risk control threshold are obtained. The related party risk measurement model and the related transaction risk measurement model are pre-trained as follows: For the related party risk measurement model, related party feature data is collected, and the risk levels of the related parties are labeled. The labeled sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. Combined with the risk levels of the related parties, the machine learning model performs a mapping learning from related parties to risk levels. The model is then tested using the test set to obtain the related party risk measurement model. The related party risk control threshold is determined based on the risk levels of the related parties output by the model. For the related transaction risk measurement model, related transaction feature data is collected, and the risk control thresholds of the related transactions are labeled. The labeled sample set is divided to obtain a training set and a test set. The machine learning model is trained using the training set. By combining the risk control thresholds corresponding to the related-party transactions, the machine learning model is mapped from related-party transactions to risk control thresholds, and tested using the test set to obtain a related-party transaction risk measurement model. The related-party risk control threshold and the risk control threshold corresponding to the related-party transaction are multiplied to determine the transaction risk control threshold. Early warning conditions are set according to the transaction risk control threshold. When the related-party transaction data corresponding to the related party meets the early warning conditions, an early warning prompt is generated, and the related-party transaction is controlled according to the early warning prompt. The overall solution automatically completes the early warning analysis of the related-party transactions that need to be monitored, significantly reducing the workload of personnel in related-party transaction management positions, reducing statistical errors, avoiding statistical caliber errors caused by personnel handover or replacement, improving control accuracy, and enhancing work efficiency. This invention uses a related-party risk measurement model and a related-party transaction risk measurement model, which can distinguish different risk levels and set corresponding control thresholds with quantitative indicators. By combining the two control thresholds, the actual control threshold for each related party in various types of related-party transactions that need to be monitored can be determined individually according to the actual risk situation. Through automatic calculation and timely early warning, the pre-control of related-party transactions that need to be monitored is achieved, ensuring early warning, timely response, and reducing the risk of violations, providing strong technical support for risk early warning.

[0174] The acquisition, storage, use, and processing of data in this application comply with relevant laws and regulations.

[0175] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0176] This invention is described with reference to flowchart illustrations and / or block diagrams of methods and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0177] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0178] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0179] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A transaction risk early warning control method based on a risk measurement model, characterized in that: include: Obtain related party data and related transaction data corresponding to the related parties; The related party data and the related transaction data corresponding to the related parties are respectively input into the related party risk measurement model and the related transaction risk measurement model, and the related party risk control threshold and the related transaction risk control threshold are obtained through analysis; wherein the related party risk measurement model and the related transaction risk measurement model are pre-trained in the following manner: For the related party risk measurement model, the related party characteristic data corresponding to the related party is collected, and the risk level corresponding to the related party is marked, and the marked sample set is divided to obtain a training set and a test set; the machine learning model is trained using the training set, and the mapping of the related party to the risk level is learned for the machine learning model in combination with the risk level corresponding to the related party, and the test set is used for testing to obtain the related party risk measurement model, and the related party risk control threshold is determined according to the risk level corresponding to the related party output by the model; For the risk measurement model of related-party transactions, characteristic data of related-party transactions are collected, and risk control thresholds corresponding to related-party transactions are marked, and the marked sample set is divided to obtain a training set and a test set; the machine learning model is trained using the training set, and mapping learning from related-party transactions to risk control thresholds is performed on the machine learning model in combination with the risk control thresholds corresponding to the related-party transactions, and the test set is used for testing to obtain the risk measurement model of related-party transactions; Multiplying the related party risk control threshold and the risk control threshold corresponding to the related transaction to determine the transaction risk control threshold; An early warning condition is set according to the transaction risk control threshold, and when the related transaction data corresponding to the related party meets the early warning condition, an early warning prompt is generated and the related transaction is controlled according to the early warning prompt.

2. The transaction risk early warning control method based on the risk measurement model according to claim 1 is characterized in that: The method includes: Collect a list of related party customers; Based on the related party customer list, related party information and related transaction information corresponding to the related parties are counted.

3. The transaction risk early warning control method based on the risk measurement model according to claim 1 is characterized in that: The method further includes: According to the transaction supervision requirements of the regulatory authorities, historical transaction information of related parties under different calibers is collected, including at least transaction scale and frequency, to establish a related party data mart; Filter and clean related parties, including extracting related party data with related transactions from historical transaction information, eliminating invalid and redundant data, and unifying and standardizing the cleaned related party data; According to the related party data mart, the related party risk measurement score is determined through multiple dimensions including the number of enterprises corresponding to the related parties, the frequency of transactions, the number of related transactions with a single maximum amount exceeding a set limit within a set time period, the probability of related transactions of preset business types occurring between the related parties, and the fair value of related transactions within a set time period; A plurality of risk levels corresponding to the related parties are obtained according to the related party risk measurement scores, and for each risk level, a corresponding related party risk control threshold is set respectively.

4. The transaction risk early warning control method based on the risk measurement model according to claim 3 is characterized in that: For the related party risk measurement model, the related party characteristic data corresponding to the related party is collected, and the risk level corresponding to the related party is marked, and the marked sample set is divided to obtain a training set and a test set; the machine learning model is trained using the training set, and the machine learning model is mapped from the related party to the risk level in combination with the risk level corresponding to the related party, and the test set is used for testing to obtain the related party risk measurement model, and the related party risk control threshold is determined according to the risk level corresponding to the related party output by the model, including: For the related party risk measurement model, the related party feature data corresponding to the related parties are collected, wherein feature extraction and analysis are performed through the business dimension feature values ​​including the related party role, business type, transaction scale, and transaction frequency to determine the related party feature data as the sample set for model training; The Bayesian algorithm is used to train the model, construct the risk level corresponding to the related party and the classification prediction of the related party characteristic data, and obtain the related party risk measurement model; The related party risk measurement model is tested, the characteristic dimensions of the related parties are automatically corrected according to the test results, the characteristic data of the related parties are optimized, and the model parameters are adjusted manually.

5. The transaction risk early warning control method based on the risk measurement model according to claim 1 is characterized in that: The method further includes: According to the transaction supervision requirements of the regulatory authorities, determine the transaction type, amount, business type and calculation mechanism of related transactions that need to be included in the calculation; According to the calculation mechanism, in combination with the transaction type, amount, and business type corresponding to the transaction type that needs to be included in the calculation, historical transaction data is analyzed to determine the transaction scale, increase or decrease of the related transactions, and a related transaction data mart is established; Filter and clean the business data of related transactions, including eliminating invalid and redundant data, converting, integrating, discretizing and normalizing the data; According to the related-party transaction data mart, the risk control threshold corresponding to the related-party transaction is determined through multiple dimensions including the related-party role, business type, transaction scale, risk limit and increase or decrease range.

6. The transaction risk early warning control method based on the risk measurement model according to claim 5 is characterized in that: For the risk measurement model of related-party transactions, the characteristic data of related-party transactions are collected, and the risk control thresholds corresponding to the related-party transactions are marked, and the marked sample set is divided to obtain a training set and a test set; the machine learning model is trained using the training set, and the mapping of related-party transactions to risk control thresholds is learned for the machine learning model in combination with the risk control thresholds corresponding to the related-party transactions, and the test set is used for testing to obtain the risk measurement model of related-party transactions, including: For the related-party transaction risk measurement model, the related-party transaction feature data corresponding to the related-party transaction is collected. The feature extraction and analysis is performed through the business dimension feature values ​​including the related-party role, business type, transaction scale, risk limit and increase / decrease range corresponding to the related-party transaction to determine the related-party transaction feature data as the sample set for model training; The model is trained using the support vector machine algorithm, and the Gaussian kernel is used as the kernel function to optimize the model accuracy and generalization ability, and the F1 score is used to evaluate the model.

7. The transaction risk early warning control method based on the risk measurement model according to claim 1 is characterized in that: Setting a warning condition according to the transaction risk control threshold, generating a warning prompt when the related transaction data corresponding to the related party meets the warning condition, and controlling the related transaction according to the warning prompt, including: The amount of the related-party transaction to be monitored is multiplied by the transaction risk control threshold to obtain the warning line of the related party as the warning condition; When the transaction scale of related-party transactions that need to be monitored reaches the warning line of the related-party, a warning prompt is generated and pushed to the related-party transaction processing personnel, who will notify the corresponding business departments to perform risk control and transaction processing.

8. A transaction risk early warning control device based on a risk measurement model, characterized in that: include: A data acquisition module is used to acquire the data of related parties and the related transaction data corresponding to the related parties; The risk measurement analysis module is used to input the related party data and the related transaction data corresponding to the related party into the related party risk measurement model and the related transaction risk measurement model respectively, and obtain the related party risk control threshold and the related transaction risk control threshold through analysis; wherein the related party risk measurement model and the related transaction risk measurement model are pre-trained in the following manner: For the related party risk measurement model, the related party characteristic data corresponding to the related party is collected, and the risk level corresponding to the related party is marked, and the marked sample set is divided to obtain a training set and a test set; the machine learning model is trained using the training set, and the mapping of the related party to the risk level is learned for the machine learning model in combination with the risk level corresponding to the related party, and the test set is used for testing to obtain the related party risk measurement model, and the related party risk control threshold is determined according to the risk level corresponding to the related party output by the model; For the risk measurement model of related-party transactions, characteristic data of related-party transactions are collected, and risk control thresholds corresponding to related-party transactions are marked, and the marked sample set is divided to obtain a training set and a test set; the machine learning model is trained using the training set, and mapping learning from related-party transactions to risk control thresholds is performed on the machine learning model in combination with the risk control thresholds corresponding to the related-party transactions, and the test set is used for testing to obtain the risk measurement model of related-party transactions; A risk control threshold determination module, used to multiply the risk control threshold of the related party and the risk control threshold corresponding to the related transaction to determine the transaction risk control threshold; The transaction warning module is used to set warning conditions according to the transaction risk control threshold, generate warning prompts when the related transaction data corresponding to the related party meets the warning conditions, and control the related transactions according to the warning prompts.

9. The transaction risk early warning control device based on the risk measurement model according to claim 8 is characterized in that: The data acquisition module is specifically used for: Collect a list of related party customers; Based on the related party customer list, related party information and related transaction information corresponding to the related parties are counted.

10. The transaction risk early warning control device based on the risk measurement model according to claim 8, characterized in that: The device also includes: an associated party data processing module; The associated party data processing module is specifically used for: According to the transaction supervision requirements of the regulatory authorities, historical transaction information of related parties under different calibers is collected, including at least transaction scale and frequency, to establish a related party data mart; Filter and clean related parties, including extracting related party data with related transactions from historical transaction information, eliminating invalid and redundant data, and unifying and standardizing the cleaned related party data; According to the related party data mart, the related party risk measurement score is determined through multiple dimensions including the number of enterprises corresponding to the related parties, the frequency of transactions, the number of related transactions with a single maximum amount exceeding a set limit within a set time period, the probability of related transactions of preset business types occurring between the related parties, and the fair value of related transactions within a set time period; A plurality of risk levels corresponding to the related parties are obtained according to the related party risk measurement scores, and for each risk level, a corresponding related party risk control threshold is set respectively.

11. The transaction risk early warning control device based on the risk measurement model according to claim 10, characterized in that: The risk measurement analysis module is specifically used for: For the related party risk measurement model, the related party feature data corresponding to the related parties are collected, wherein feature extraction and analysis are performed through the business dimension feature values ​​including the related party role, business type, transaction scale, and transaction frequency to determine the related party feature data as the sample set for model training; The Bayesian algorithm is used to train the model, construct the risk level corresponding to the related party and the classification prediction of the related party characteristic data, and obtain the related party risk measurement model; The related party risk measurement model is tested, the characteristic dimensions of the related parties are automatically corrected according to the test results, the characteristic data of the related parties are optimized, and the model parameters are adjusted manually.

12. The transaction risk early warning control device based on the risk measurement model according to claim 8, characterized in that: The device also includes: an associated transaction data processing module; The associated transaction data processing module is specifically used for: According to the transaction supervision requirements of the regulatory authorities, determine the transaction type, amount, business type and calculation mechanism of related transactions that need to be included in the calculation; According to the calculation mechanism, in combination with the transaction type, amount, and business type corresponding to the transaction type that needs to be included in the calculation, historical transaction data is analyzed to determine the transaction scale, increase or decrease of the related transactions, and a related transaction data mart is established; Filter and clean the business data of related transactions, including eliminating invalid and redundant data, converting, integrating, discretizing and normalizing the data; According to the related-party transaction data mart, the risk control threshold corresponding to the related-party transaction is determined through multiple dimensions including the related-party role, business type, transaction scale, risk limit and increase or decrease range.

13. The transaction risk early warning control device based on the risk measurement model according to claim 12, characterized in that: The risk measurement analysis module is specifically used for: For the related-party transaction risk measurement model, the related-party transaction feature data corresponding to the related-party transaction is collected. The feature extraction and analysis is performed through the business dimension feature values ​​including the related-party role, business type, transaction scale, risk limit and increase / decrease range corresponding to the related-party transaction to determine the related-party transaction feature data as the sample set for model training; The model is trained using the support vector machine algorithm, and the Gaussian kernel is used as the kernel function to optimize the model accuracy and generalization ability, and the F1 score is used to evaluate the model.

14. The transaction risk early warning control device based on the risk measurement model according to claim 8, characterized in that: The transaction early warning module is specifically used for The amount of the related-party transaction to be monitored is multiplied by the transaction risk control threshold to obtain the warning line of the related party as the warning condition; When the transaction scale of related-party transactions that need to be monitored reaches the warning line of the related-party, a warning prompt is generated and pushed to the related-party transaction processing personnel, who will notify the corresponding business departments to perform risk control and transaction processing.

15. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

16. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

17. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images