Compliance risk early warning detection method for bank-enterprise direct connection cross-border payment

Abstract

The invention relates to the technical field of finance, and particularly discloses a compliance risk early-warning detection method for bank-enterprise direct-connection cross-border payment, and the method comprises the steps: obtaining an original transaction sequence of bank-enterprise cross-border payment, and synchronously collecting transaction object information, fund flow data and an error-related supervision parameter sequence; obtaining a transaction mode component according to the original transaction sequence; generating a preliminary risk early warning parameter according to the transaction mode component and the error-related supervision parameter sequence; and obtaining an accumulated number of payment services, and generating an early warning threshold factor and a risk weight factor based on the accumulated number and the fund flow data. According to the method, the hidden association of the amount, the purpose and the time is mined by means of the feature association index, the risk is quantified in combination with the mode compliance deviation degree, and the early warning threshold value and the weight factor which are dynamically generated are matched, so that the risk identification accuracy is greatly improved, the missing report rate and the false report rate are reduced, and particularly, the hidden risk in complex cross-border payment can be accurately captured.

Description

Technical Field

[0001] This invention relates to the field of financial technology, and in particular to a method for early warning and detection of compliance risks in cross-border payments directly between banks and enterprises. Background Technology

[0002] Against the backdrop of global economic integration and the deep integration of digital finance, direct bank-enterprise cross-border payment has become the core model for foreign trade enterprises and multinational groups to conduct cross-border fund settlement due to its advantages of seamless system integration and automated transaction processing, significantly improving payment efficiency and account management convenience.

[0003] The existing bank-enterprise direct connection system is fragmented with the bank's regulatory system and external data service providers, failing to effectively integrate multi-source information such as credit data of transaction objects and data on fund flow levels, resulting in a lack of a panoramic perspective for risk identification. Furthermore, it relies heavily on manual intervention, with a large amount of compliance review being completed manually under the traditional model. This is not only inefficient but also prone to false alarms and omissions due to subjective judgment biases, making it difficult to meet the real-time early warning needs of transactions with a scale of tens of millions. Summary of the Invention

[0004] The purpose of this invention is to provide a compliance risk warning and detection method for direct bank-enterprise cross-border payments, so as to solve the technical problems mentioned in the background.

[0005] To achieve the above objectives, the present invention provides the following technical solution: Compliance risk warning and detection methods for direct bank-enterprise cross-border payments include: Obtain the original transaction sequence of cross-border payments between banks and enterprises, and simultaneously collect transaction object information, fund flow data, and error-related regulatory parameter sequences; Obtain the transaction pattern component based on the original transaction sequence; Preliminary risk warning parameters are generated based on the transaction pattern components and error-related regulatory parameter sequence. Obtain the cumulative number of payment transactions, and generate an early warning threshold factor and a risk weight factor based on the cumulative number of transactions and fund flow data; The initial risk warning parameters, warning threshold factors, and risk weight factors are adaptively fused to generate a transaction risk level. Determine whether the transaction risk level is lower than the preset level; If the value is less than the threshold, the transaction is considered to be within a safe range, corresponding to the transaction risk level. If the risk level is greater than the risk level, the transaction is deemed to be at risk, and an early warning message is generated based on the risk level.

[0006] Preferably, the step of obtaining the transaction pattern component based on the original transaction sequence includes: A classification segmentation sequence is obtained based on the original transaction sequence, and transaction data groups are obtained based on the classification segmentation sequence; The transaction field set is obtained by grouping the transaction data, wherein the transaction field set includes transaction amount information, payment purpose information, and instruction initiation time information; Transaction feature fragments are obtained based on transaction amount information, payment purpose information, and instruction initiation time information, and transaction data is split into transaction time sequence feature information according to fixed time windows; The feature correlation index is obtained based on the transaction timing characteristics and the error-related regulatory parameter sequence. Based on the aforementioned feature correlation index, risk characteristic assessment information corresponding to transaction amount information, payment purpose information, and instruction initiation time information is obtained; The transaction pattern component is obtained based on the risk characteristic assessment information corresponding to the transaction amount information, payment purpose information, instruction initiation time information, and preset screening criteria.

[0007] Preferably, the step of generating preliminary risk warning parameters based on the transaction pattern components and error-related regulatory parameter sequence includes: Obtain a preset regulatory parameter mapping table, break down the constituent elements of the transaction mode components, and extract matching regulatory parameters to form a subset by comparing the correspondence between elements and parameters in the preset regulatory parameter mapping table. Based on the subset of associated regulatory parameters, obtain the completeness and format compliance of the required fields for each subset, and generate parameter compliance results based on the completeness and format compliance of the required fields. The actual characteristic description of the transaction pattern component is compared with the compliance characteristic requirements corresponding to the associated regulatory parameter subset to obtain the pattern compliance deviation degree; Obtain the corresponding risk statement based on the degree of deviation from the compliance of the pattern, and generate a preliminary risk description based on the risk statement; Match the preliminary risk descriptions with the risk scenarios in the risk level correspondence table, and extract the risk quantification identifiers corresponding to the matching scenarios; Based on the preliminary risk coefficient and parameter dynamic adjustment instructions, and combined with the latest changes in regulatory parameters, a revised preliminary risk coefficient is generated, and then preliminary risk warning parameters are obtained based on the revised preliminary risk coefficient.

[0008] Preferably, the step of obtaining the cumulative number of payment transactions and generating a warning threshold factor and a risk weight factor based on the cumulative number of transactions and fund flow data includes: The cumulative number of payment transactions and transaction timestamps are obtained from the bank-enterprise direct connection transaction logs. Each transaction record in the transaction log is traversed, the total number of records is counted to obtain the cumulative number of transactions, and the time field of each record is extracted to obtain the transaction timestamp. The transaction timestamps are categorized into corresponding time windows according to preset time intervals, and the number of transaction records in each window is counted to obtain the transaction frequency per unit time. Based on the order of account receipt in the fund flow data, and combined with the account flow trajectory map, the fund flow hierarchy data is obtained by marking the levels of intermediary accounts and the number of nodes through which the funds pass. The cumulative number of transactions and the frequency of transactions per unit time are compared with similar data in the threshold benchmark dataset, and the threshold benchmark corresponding to the matching data is extracted as the early warning threshold factor. Based on the complexity of the fund transfer data at each level and the frequency of regulatory parameter updates, risk weight factors are generated by determining the proportion of each level in accordance with regulations.

[0009] Preferably, the step of generating a transaction risk level by adaptively fusing preliminary risk warning parameters with warning threshold factors and risk weight factors includes: Standardized risk parameters are obtained based on the preliminary risk warning parameters and parameter normalization standards. By comparing the dimensional conversion requirements of each risk indicator in the parameter normalization standards, the different dimensions of the preliminary risk warning parameters are converted into values ​​within a unified range. Extract the indicator values ​​from the standardized risk parameters, find the corresponding threshold ranges of similar indicators in the early warning threshold factors, record the threshold intervals where the indicator values ​​are located, and obtain the threshold comparison results. Based on the proportion of each dimension in the risk weighting factor, the risk information of different dimensions in the threshold comparison results is integrated to generate weighted risk information; Based on the adjustment experience of similar risks in historical early warning results, the detailed description of weighted risk information is optimized in accordance with the standard to obtain dynamically corrected risk information; By comparing the core risk indicators in the dynamically revised risk information with the risk level classification range, the corresponding level interval is determined, and the level interval attribution information is obtained. Based on the correspondence between the grade range attribution information and the grade name, the range attribution is converted into a transaction risk level.

[0010] Preferably, the steps of obtaining the corresponding risk statement based on the degree of deviation from compliance of the pattern, and generating a preliminary risk description based on the risk statement, include: Based on the standards for decomposing the deviation degree and deviation characteristics of the pattern, detailed deviation data is obtained; Based on the deviation details data and the corresponding risk description table, obtain the basic risk description data; Obtain the preset statement association logic, and obtain the associated risk statement data based on the basic risk statement data and the preset statement association logic; Obtain the preset deviation integrity checklist, and obtain the supplementary risk statement data based on the associated risk statement data and the preset deviation integrity checklist; Remove duplicate statements from the structured risk description data, adjust the coherence and conciseness of the statements according to the description simplification standard, and generate a preliminary risk description.

[0011] Preferably, the steps for generating risk weighting factors by determining the proportion of each level according to regulations, based on the complexity of the fund flow hierarchy data and the frequency of regulatory parameter updates, include: Based on the hierarchical data of fund transfers, the account node sequence is obtained, and the number of transfer accounts, account attributes, and transfer interval information in the account node sequence are extracted and integrated to form hierarchical detailed data.

[0012] Obtain hierarchical complex representation data based on hierarchical detailed data; Obtain statistical records of frequency interval data and regulatory parameter update frequency, and classify the statistical records of frequency interval data and regulatory parameter update frequency into the corresponding intervals according to the time span division in the frequency segmentation standard to form frequency interval data; Obtain updated impact description data based on frequency range data; Initial percentage data are obtained based on hierarchical complex representation data and updated impact description data; Risk weight factors are obtained based on the initial percentage data and factor generation specifications.

[0013] Preferably, the step of obtaining risk characteristic assessment information corresponding to transaction amount information, payment purpose information, and instruction initiation time information based on the feature correlation index includes: Based on the feature correlation index and information dimension classification criteria, obtain multi-dimensional correlation data; Based on the multi-dimensional correlation data, obtain the relevant content related to the transaction amount, and based on the relevant content related to the transaction amount, obtain the specific characteristics such as the correlation performance and correlation strength between the amount and other transaction elements, and generate the amount correlation feature data based on the specific characteristics such as the correlation performance and correlation strength.

[0014] Based on the usage-related fragments and usage feature extraction criteria in the multidimensional related data, usage-related feature data is obtained; Based on the time-related segments and time feature extraction criteria in the multidimensional related data, time-related feature data is obtained; Based on the data related to monetary amount, usage, and time, obtain risk factor data for sub-information. Risk characteristic assessment information is obtained based on risk factor data.

[0015] Preferably, the step of obtaining risk element data based on monetary value-related characteristic data, usage-related characteristic data, and time-related characteristic data includes: Obtain data on monetary risk elements based on monetary correlation characteristic data; Obtain application risk factor data based on application-related characteristic data; Obtain time risk element data based on time-related characteristic data; Based on the data on monetary risk factors, usage risk factors, and time risk factors, sub-information risk factor data is generated.

[0016] Preferably, the step of generating sub-information risk element data based on amount risk element data, purpose risk element data, and time risk element data includes: Obtain structured monetary risk element data based on monetary risk element data; Obtain structured use risk factor data based on use risk factor data; Obtain structured time risk element data based on time risk element data; Structured amount risk factor data, structured purpose risk factor data, and structured time risk factor data are mapped to transaction amount, payment purpose, and instruction initiation time, respectively, and sub-information risk factor data are generated based on the transaction amount, payment purpose, and instruction initiation time.

[0017] The beneficial effects of this application are as follows: This invention achieves deep integration of multi-source data by simultaneously collecting transaction sequences, object information, fund flows, and regulatory parameters, combined with multi-dimensional structured processing, thus solving the data fragmentation problem and laying a high-quality data foundation for risk assessment. By leveraging feature correlation indices to uncover implicit correlations in amounts, uses, and times, and combining this with quantifiable risk measures based on compliance deviations, coupled with dynamically generated warning thresholds and weighting factors, the accuracy of risk identification is significantly improved, reducing false negatives and missed positives, especially in accurately capturing hidden risks in complex cross-border payments. Full-process automated processing replaces traditional manual review, achieving an end-to-end closed loop from data extraction and feature analysis to warning generation, significantly shortening risk response time and reducing labor costs. Through incremental learning and dynamic parameter adjustment, it can adapt to regulatory policy updates and changes in transaction patterns in real time, enhancing the system's long-term applicability.

[0018] Ultimately, it provides financial institutions and enterprises with accurate and efficient compliance risk warning tools to help prevent risks such as money laundering and sanctions violations, and to ensure the compliance and security of cross-border payment business. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of a method flow according to an embodiment of this application.

[0020] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0022] like Figure 1 As shown, this application provides a compliance risk early warning and detection method for direct bank-enterprise cross-border payments, including: S1. Obtain the original transaction sequence of cross-border payments between banks and enterprises, and simultaneously collect transaction object information, fund flow data and error-related regulatory parameter sequences; S2. Obtain the transaction pattern component based on the original transaction sequence; S3. Generate preliminary risk warning parameters based on the transaction pattern components and error-related regulatory parameter sequence; S4. Obtain the cumulative number of payment transactions, and generate an early warning threshold factor and a risk weight factor based on the cumulative number of transactions and fund flow data; S5. The preliminary risk warning parameters, warning threshold factors, and risk weight factors are adaptively fused to generate a transaction risk level. S6. Determine whether the transaction risk level is lower than the preset level; If the value is less than the threshold, the transaction is considered to be within a safe range, corresponding to the transaction risk level. If the risk level is greater than the risk level, the transaction is deemed to be at risk, and an early warning message is generated based on the risk level.

[0023] As described in steps S1-S6 above, cross-border payments involve a variety of complex transaction data and regulatory requirements. Therefore, how to efficiently extract valuable risk information from transaction data and dynamically adjust risk assessments based on real-time data has become a core issue that financial institutions urgently need to address in cross-border payments. Traditional risk assessment methods often rely on static rules and manual review, lacking flexibility and timeliness, and are unable to effectively handle large-scale real-time payment data.

[0024] This invention constructs a compliance risk early warning and detection method for direct bank-enterprise cross-border payments. It aims to utilize data analysis technology to monitor potential risks in the cross-border payment process in real time, providing accurate early warning information based on data correlation and risk assessment models, thereby offering compliance assurance for financial institutions and related enterprises. Specifically, through a risk identification and early warning mechanism, it effectively identifies and assesses potential financial transaction risks, thereby ensuring the compliance and security of cross-border payment operations.

[0025] The system obtains the original transaction sequences of cross-border payments between banks and enterprises from payment platforms. These sequences contain basic information about each transaction, such as the transaction amount, payment purpose, initiator, and recipient. Then, it simultaneously collects information on the transaction counterparties, including the account and identity information of both the payer and payee. Furthermore, it needs to collect fund flow data to clarify the path and stages of fund movement in each transaction, ensuring the ability to trace the source and destination of funds and prevent money laundering and other illegal activities.

[0026] The original transaction sequence is processed using a feature extraction algorithm to extract the transaction pattern components. These transaction pattern components refer to the key feature components in the transaction data, mainly including transaction amount, payment purpose, transaction time, and transaction counterparty. These components reveal the basic characteristics and patterns of the transactions.

[0027] By comparing trading patterns and error monitoring data, preliminary risk warning parameters are obtained, providing data support for subsequent risk level assessment. Dynamically capturing risk characteristics in trading enables a precise quantitative description of trading risk.

[0028] The early warning threshold factor and risk weight factor are determined by analyzing the cumulative number of payment transactions and fund flow data. First, all transaction records in cross-border payments between banks and enterprises are statistically analyzed to calculate the cumulative number of payment transactions. Then, based on the fund flow data, the path of fund transfer for each transaction and the account nodes involved are analyzed to obtain hierarchical data on fund flow.

[0029] An adaptive fusion processing algorithm combines preliminary risk warning parameters with warning threshold factors and risk weight factors to generate a final transaction risk level. This algorithm is a dynamic processing method based on data flow and real-time feedback mechanisms. By continuously optimizing the parameters and weights during the fusion process, the final risk level more accurately reflects the actual risk of the transaction.

[0030] The generated transaction risk level is assessed to determine if it is lower than the preset safety level. If the risk level is lower than the preset safety range, the transaction is considered to be within the safe range; if the risk level is higher than the safety range, the transaction is considered to be risky.

[0031] In one embodiment, the step of obtaining the transaction pattern component based on the original transaction sequence includes: S201. Obtain a classification segmentation sequence based on the original transaction sequence, and obtain transaction data groups based on the classification segmentation sequence; S202. Obtain a transaction field set based on the transaction data grouping, wherein the transaction field set includes transaction amount information, payment purpose information, and instruction initiation time information; S203. Obtain transaction feature fragments based on transaction amount information, payment purpose information, and instruction initiation time information, and split the transaction data into transaction time sequence feature information according to a fixed time window; S204. Obtain the feature correlation index based on the transaction time sequence feature information and the error-related regulatory parameter sequence; S205. Obtain risk characteristic assessment information corresponding to transaction amount information, payment purpose information, and instruction initiation time information based on the aforementioned feature correlation index; S206. Obtain the transaction pattern component based on the risk characteristic assessment information corresponding to the transaction amount information, payment purpose information, instruction initiation time information, and preset screening criteria.

[0032] As described in steps S201-S206 above, the present invention classifies and segments the original transaction data, transforming the diverse and complex original transaction data into manageable and analyzable subsets, reducing the complexity of subsequent processing. Traditional methods process transaction data roughly, ignoring inherent category differences, resulting in inaccurate analysis. In terms of technical implementation, the original transaction sequence is first analyzed, and the transaction data is divided into several categories according to predetermined classification standards such as amount range, payment purpose, and time range. Then, a classification and segmentation sequence is generated and further grouped according to this sequence.

[0033] Key fields are extracted from transaction data groups to provide foundational data for feature extraction and pattern analysis. Key transaction information is crucial for transaction pattern recognition. Extracting these fields can effectively conduct transaction analysis. Traditional methods rely on simple data aggregation or single-dimensional indicators, neglecting the interaction between different dimensions, which leads to blind risk assessment. In the technical implementation, key fields such as transaction amount, payment purpose, and instruction initiation time are extracted from each group of transaction data. These field values ​​represent the core characteristics of the transaction and ensure high consistency of each type of transaction feature set.

[0034] Based on the extracted transaction fields, transaction feature fragments are generated and transaction data is split. The transaction time series characteristics are analyzed to reveal timeliness and trend characteristics. Transactions contain time series information. By splitting the data into time windows, the differences in patterns over different time periods can be found, revealing potential risk characteristics. Existing technologies usually ignore transaction time series characteristics and fail to utilize time series data for risk analysis. In terms of technical implementation, transaction data is split into multiple time segments based on fixed time windows. Each segment represents the transaction behavior in a specific time period. The transaction data of each time period is then combined with information such as amount and purpose to generate transaction time series characteristics.

[0035] By analyzing the relationship between transaction time-series characteristics and regulatory parameters, a feature correlation index is generated to provide a basis for subsequent risk assessment. The implicit correlation between time-series characteristics and regulatory parameters can reveal potential risk signals, and the feature correlation index can serve as a comprehensive reference standard. Traditional methods ignore the dynamic impact of regulatory parameters and cannot adjust the risk assessment model in real time. In the technical implementation, transaction time-series characteristics are matched with corresponding regulatory parameters, and a feature correlation index is generated using methods such as weighted average or correlation coefficient calculation. This index reflects the correlation between transaction behavior and regulatory parameters, as well as compliance risks.

[0036] Specific risk characteristics are extracted based on the feature correlation index, providing a risk assessment basis for identifying transaction pattern components. Potential risks are judged by the relationship between related transaction characteristics and regulatory parameters. The index is transformed into specific assessment information, which can provide a quantitative basis for risk handling. Traditional methods only conduct single-dimensional assessments, which are difficult to comprehensively consider multiple factors. In terms of technical implementation, risk analysis is carried out on dimensions such as transaction amount, payment purpose, and instruction initiation time based on the feature correlation index, generating risk assessment information for each transaction.

[0037] By analyzing risk characteristic assessment information and combining it with preset screening criteria, transaction pattern components are obtained. By integrating previously acquired transaction characteristic information, transaction pattern components are identified to support subsequent risk warning and compliance monitoring. Traditional methods rely on simple rule matching and lack in-depth analysis of complex transaction patterns. In the technical implementation, transaction risk characteristics are comprehensively assessed, and preset screening criteria such as risk level thresholds and regulatory requirements are combined to identify the pattern components of each transaction. These components represent different patterns of transaction behavior, providing basic data for risk prediction and compliance monitoring.

[0038] In one embodiment, the step of generating preliminary risk warning parameters based on the transaction pattern components and error-related regulatory parameter sequence includes: S301. Obtain the preset regulatory parameter mapping table, decompose the constituent elements of the transaction mode component, and extract the matching regulatory parameters to form a subset by referring to the correspondence between the elements and parameters in the preset regulatory parameter mapping table. S302. Obtain the completeness and format compliance of the required fields for each subset of the associated regulatory parameters, and generate parameter compliance results based on the completeness and format compliance of the required fields. S303. Compare the actual characteristic description of the transaction pattern component with the compliance characteristic requirements corresponding to the associated regulatory parameter subset to obtain the pattern compliance deviation. S304. Obtain the corresponding risk statement based on the degree of deviation from the compliance of the model, and generate a preliminary risk description based on the risk statement; S305. Match the preliminary risk description with the risk scenarios in the risk level correspondence table, and extract the risk quantification identifiers corresponding to the matching scenarios; S306. Based on the preliminary risk coefficient and the explanation of dynamic parameter adjustment, and combined with the latest changes in regulatory parameters, a revised preliminary risk coefficient is generated, and then preliminary risk warning parameters are obtained based on the revised preliminary risk coefficient.

[0039] As described in steps S301-S306 above, this invention generates preliminary risk warning parameters to detect compliance risks in cross-border payments directly between banks and enterprises. Specifically, it generates compliance risk warning information by progressively processing the transaction mode components and error-related regulatory parameter sequences, thereby effectively identifying potential risks, providing early warnings, ensuring the compliance of payment transactions, and avoiding legal and financial risks.

[0040] A pre-defined regulatory parameter mapping table is obtained, and a subset of regulatory parameters is extracted by decomposing transaction pattern components. The transaction pattern components and the contents of the regulatory parameter mapping table are closely related. Transaction patterns themselves represent the characteristics of different payment scenarios (such as amount, time, purpose, etc.), while the regulatory parameter mapping table provides the standards that need to be considered for these characteristics in compliance checks (such as compliance requirements set by regulatory agencies). Therefore, obtaining the mapping table and decomposing the constituent elements of transaction pattern components can help extract relevant data from regulatory requirements, forming a basic dataset for compliance checks. Existing technologies typically only focus on a single transaction element (such as amount or time), neglecting the comprehensive relationship between various dimensions of the transaction. This step, by decomposing the transaction pattern components, comprehensively considers all dimensions of the transaction (such as amount, payment purpose, time, etc.), thereby providing a more comprehensive and accurate compliance assessment. In practice, the original transaction data is first parsed to extract key elements such as transaction amount, payment purpose, and initiation time, ensuring that each dimension is fully considered. Next, a preset regulatory parameter mapping table is obtained, which defines compliance requirements related to transaction elements, such as the upper limit of the amount and the compliance of the payment purpose. Finally, based on the elements of each transaction mode component, the corresponding compliance requirements in the mapping table are matched to extract the relevant regulatory parameters, forming a subset of regulatory parameters for subsequent compliance checks.

[0041] The extracted subset of regulatory parameters undergoes a compliance check to determine whether required fields are complete and whether their content conforms to regulations. The compliance of each transaction item is reflected not only in the existence of fields but also in whether their content complies with relevant regulations. If the field content does not comply, the transaction may fail compliance review or even trigger a risk warning. Therefore, checking the completeness and format compliance of each field is a crucial step in ensuring transaction compliance. Traditional methods often rely on simple field completeness checks, neglecting the compliance of field content format. This invention comprehensively checks both field completeness and content format to ensure that each transaction item meets regulatory requirements. Specifically, a mandatory field completeness check is first performed, verifying each regulatory parameter to ensure that all mandatory fields have been filled in and that the content is not empty. Next, a field format compliance check is conducted to verify whether the field content conforms to predetermined format requirements, such as whether the amount field is a number and whether the time field conforms to time format specifications. Finally, based on the check results, a parameter compliance result is generated, indicating whether the fields in each subset are compliant. If non-compliant fields are found, they will impact risk assessment.

[0042] This invention compares the actual characteristics of transaction pattern components with the compliance requirements of the regulatory parameter subset to calculate the compliance deviation. Each transaction pattern has its actual characteristics, such as the actual transaction amount, purpose, and time. These characteristics must comply with relevant regulatory requirements. The compliance deviation measures the degree of deviation between the actual characteristics and the compliance requirements. A high deviation indicates a significant compliance risk. Traditional compliance inspection methods simply determine whether a certain field meets the regulations without considering the gap between the actual data and the regulations. This invention quantifies risk and provides a more detailed compliance analysis by calculating the compliance deviation. The specific implementation process is as follows: First, extract the actual characteristic descriptions from the transaction data, such as transaction amount, payment purpose, and instruction initiation time. Then, compare the extracted actual characteristics with the compliance requirements of the regulatory parameter subset item by item, such as whether the transaction amount exceeds the limit and whether the payment purpose complies with regulations. Finally, based on the comparison results, calculate the compliance deviation for each element. The deviation can be calculated by comparing the difference between the actual value and the compliance requirements or by quantifying it through other scoring standards.

[0043] Based on the compliance deviation, a corresponding risk statement is generated, forming a preliminary risk description. Transactions with high compliance deviations often imply higher risks. Therefore, generating corresponding risk statements based on deviations helps relevant personnel understand the risk characteristics of transactions and take appropriate measures to address them. Traditional methods typically only provide the results of compliance checks without further explaining the risk implications behind the deviation. By generating detailed risk statements, this invention provides more intuitive and easily understood risk information. In specific implementation, firstly, based on the calculated compliance deviation, a corresponding risk statement is generated with reference to a preset risk statement template. For example, a transaction with a high deviation might be described as "high risk," while a transaction with a low deviation might be described as "low risk." Then, the generated risk statement is matched with the relevant risk level to obtain a preliminary risk description, which can provide a basis for subsequent risk quantification.

[0044] Based on the initial risk description, the risk scenarios are matched with those in the risk level correspondence table, and corresponding risk quantification identifiers are extracted. Risk quantification provides data support for subsequent processing. By quantifying the initial risk description, specific risk levels can be assigned to transaction risks, facilitating the development of handling measures. Traditional methods lack a comprehensive risk quantification system, resulting in a lack of objectivity and standardization in risk assessment. Matching with the risk level correspondence table can improve the accuracy and operability of risk assessment. In practice, firstly, based on the initial risk description, the risk level correspondence table is queried to find the matching risk scenarios; then, based on the matched scenarios, risk quantification identifiers, such as specific risk level scores or risk indices, are extracted from the correspondence table.

[0045] Based on the preliminary risk coefficient and parameter dynamic adjustment instructions, revised preliminary risk warning parameters are generated. These parameters need to be adjusted according to actual conditions to ensure they better comply with current regulatory requirements. The dynamic adjustment instructions take into account the latest regulatory information and can update the risk coefficient in real time, improving the accuracy of warnings. Traditional methods lack a real-time update mechanism, causing warning parameters to be unable to adapt to the rapidly changing regulatory environment. This invention, by dynamically adjusting risk warning parameters, ensures the timeliness and accuracy of risk warnings. In specific implementation, a preliminary risk coefficient is first calculated based on the risk quantification identifier; then, based on the latest regulatory parameter changes, the preliminary risk coefficient is adjusted to generate revised risk warning parameters.

[0046] In one embodiment, the step of obtaining the cumulative number of payment transactions and generating an early warning threshold factor and a risk weight factor based on the cumulative number of transactions and fund flow data includes: S401. Obtain the cumulative number of payment transactions and transaction timestamps from the bank-enterprise direct connection transaction logs. Iterate through each transaction record in the transaction logs, count the total number of records to obtain the cumulative number of transactions, and extract the time field from each record to obtain the transaction timestamp. S402. Classify the transaction timestamps into the corresponding time windows according to the preset time interval, and count the number of transaction records in each window to obtain the transaction frequency per unit time. S403. Based on the account receiving order in the fund flow data, and combined with the account flow trajectory map, mark the transfer account levels and number of nodes through which the funds pass to obtain fund flow level data. S404. Compare the cumulative number of transactions and the transaction frequency per unit time with similar data in the threshold benchmark dataset, and extract the threshold benchmark corresponding to the matching data as the early warning threshold factor. S405. Based on the complexity of the fund flow hierarchy data and the frequency of regulatory parameter updates, risk weight factors are generated by determining the proportion of each hierarchy in accordance with regulations.

[0047] As described in steps S401-S405 above, this invention generates early warning threshold factors and risk weight factors by conducting in-depth analysis of multiple information such as the cumulative number of transactions, transaction timestamps, fund flow data, and regulatory parameters of cross-border payment transactions directly connected between banks and enterprises. This helps risk monitoring accurately identify potential payment risks, issue timely warnings in case of abnormal transactions, and thus enhance the security and compliance of payments.

[0048] In cross-border payment scenarios involving direct bank-enterprise connections, fund flows span multiple accounts and regions, and the number of payment transactions is typically large, making risk assessment for each transaction crucial. To ensure payment compliance and mitigate potential risks, in-depth analysis of transaction activities is essential, particularly focusing on key information such as transaction patterns, fund flows, and transaction frequency. This technology analyzes core factors such as the cumulative number of payment transactions, transaction timestamps, and fund flow hierarchy data to provide effective parameters reflecting transaction security, and generates risk warning factors and weighting factors accordingly. In practical applications, the amount, purpose, and timing of cross-border payments are important criteria for risk assessment; the cumulative number of transactions and transaction frequency reflect payment activity levels; and fund flow hierarchy data reveals the trajectory of fund flows. By extracting and analyzing this information, a complete and targeted risk warning mechanism can be built, improving both payment transaction compliance and overall payment security.

[0049] By analyzing the transaction logs of direct bank-enterprise connections, the cumulative number of payment transactions and transaction timestamps are extracted as basic data. Transaction logs are the fundamental data source for recording payment transaction activities. Each record contains key information such as transaction amount, payment purpose, and initiation time. Traversing the logs allows calculation of the cumulative number of transactions and extraction of transaction timestamps for the current time period. In the technical implementation, the transaction log database of direct bank-enterprise connections is first accessed. Each record is traversed, and the specific time of the transaction is extracted through the timestamp field. The total number of records is counted to obtain the cumulative number of transactions. The transaction timestamps provide crucial support for subsequent time window division and transaction frequency statistics. For example, if a bank has 5,000 transactions in a day, log analysis can yield a cumulative number of 5,000 transactions and the specific timestamp of each transaction (e.g., September 21, 2025, 14:30:45).

[0050] Transaction timestamps are categorized into corresponding time windows according to preset time intervals to obtain the transaction frequency within each window, providing support for risk assessment. Physically, payment transaction frequency is a crucial indicator for risk assessment; high transaction frequency may suggest abnormal risks such as money laundering. Statistical frequency by time window can assess business activity and provide quantitative evidence for early warning. In technical implementation, timestamp data is grouped according to preset intervals (e.g., hours, days, weeks), each transaction record is categorized into the corresponding window, and the number of records in each window is counted to obtain the transaction frequency per unit time. For example, if each hour is pre-defined as a group, the transaction at 14:30:45 on September 21, 2025, is categorized into the 14:00-15:00 window, ultimately yielding the frequency per unit time.

[0051] By analyzing fund flows and tracing their paths and levels, layered fund flow data is obtained, providing information on the complexity of fund flows for risk assessment. This data reveals the trajectory of fund flows, especially the intermediary accounts and nodes in cross-border payments, enabling the identification of abnormal fund paths or risky accounts, and the discovery of potential money laundering and other illegal activities. In the technical implementation, the order of account receipts in the fund flow data is combined with the account fund flow path to mark each intermediary account and node. The layer increases with each intermediary account, while simultaneously recording account attributes and flow intervals. This constructs a layered fund flow graph and extracts layered complexity features. For example, if a fund flows through three intermediary accounts A, B, and C to the receiving account, layered data is generated (layer 1: A, layer 2: B, layer 3: C), and account attributes and flow intervals are recorded.

[0052] The cumulative number of transactions and the frequency of transactions per unit time are compared with similar data in the threshold benchmark dataset, and the threshold benchmark corresponding to the matching data is extracted as the warning threshold factor. Physically, the threshold benchmark is an important reference for risk warning, determining whether transactions exceed the normal range. Abnormal cumulative number of transactions or frequency may indicate risk. Comparing historical data with the threshold benchmark can accurately identify anomalies and generate warning threshold factors. In technical implementation, the current cumulative number of transactions and transaction frequency are compared with similar data in the historical dataset to find the closest historical transaction record. The corresponding threshold benchmark is extracted as the warning threshold factor. For example, if the current cumulative number of transactions is 10,000 and the frequency is 500 transactions per hour, the matching normal frequency range (450-550 transactions / hour) is extracted after comparing with historical data as the threshold benchmark.

[0053] Based on the complexity of the fund flow hierarchy data and the frequency of regulatory parameter updates, the proportion of each hierarchy is determined according to regulations, generating risk weight factors. The more complex the fund flow hierarchy and the more frequently the regulatory parameters are updated, the higher the potential transaction risk. By quantifying the proportion of their impact on risk, factors that accurately reflect risk weights can be generated. In implementation, the complexity is first determined based on the number of intermediary accounts, account attributes, and flow intervals in the fund flow hierarchy data. Then, the frequency is obtained by statistically analyzing the number of updates per unit time based on the regulatory parameter update records. Subsequently, the impact of both on risk is converted into a proportion of each hierarchy according to preset regulations, ultimately integrating them into risk weight factors. For example, if the fund flow hierarchy is level 3 (high complexity) and the regulatory parameters are updated 3 times per week (high frequency), the proportions of both are allocated according to regulations, and corresponding risk weight factors are generated.

[0054] In one embodiment, the step of generating a transaction risk level by adaptively fusing preliminary risk warning parameters with warning threshold factors and risk weight factors includes: S501. Obtain standardized risk parameters based on the preliminary risk warning parameters and parameter normalization standards. In accordance with the dimension conversion requirements of each risk indicator in the parameter normalization standards, convert the different dimension indicators in the preliminary risk warning parameters into values ​​within a unified range. S502. Extract the indicator values ​​from the standardized risk parameters, find the corresponding threshold range of the same type of indicator in the early warning threshold factor, record the threshold interval where the indicator value is located, and obtain the threshold comparison result. S503. Based on the proportion of each dimension in the risk weighting factor, integrate the risk information of different dimensions in the threshold comparison results to generate weighted risk information. S504. Based on the adjustment experience of similar risks in historical early warning results, optimize the detailed description of weighted risk information according to the standard to obtain dynamically corrected risk information; S505. Compare the core risk indicators in the dynamically corrected risk information with the risk level classification range to determine the corresponding level interval and obtain the level interval attribution information. S506. Based on the correspondence between the grade range attribution information and the grade name, convert the range attribution into a transaction risk level.

[0055] As described in steps S501-S506 above, the present invention generates a transaction risk level by adaptively fusing preliminary risk warning parameters with warning threshold factors and risk weight factors, thereby comprehensively assessing the risk level of the transaction and providing effective risk warning information for financial institutions or relevant regulatory agencies.

[0056] In complex financial transaction scenarios such as cross-border payments and direct bank-enterprise connections, factors such as transaction amount, payment purpose, and instruction initiation time exhibit significant diversity, making it difficult for traditional risk assessment methods to comprehensively and effectively integrate multiple pieces of information for evaluation. To address this issue, this invention proposes a technical solution for standardizing, weighting, and correcting risks across different dimensions, thereby achieving more accurate and dynamic risk assessment.

[0057] Standardized risk parameters are obtained based on preliminary risk warning parameters and parameter normalization standards. During the technical implementation process, preliminary risk warning parameters may include indicators with different dimensions and numerical ranges, such as transaction amounts expressed in monetary units, payment purposes in categorized data formats, and time information in timestamp format. These parameters directly affect the accuracy of subsequent risk assessments; therefore, it is necessary to unify the dimensions according to parameter normalization standards. Different risk indicators have significantly different dimensions and units; direct comparison or weighting will lead to inaccurate results. Only after unifying the values ​​of each indicator can the risk level of each indicator be truly reflected in the subsequent weighted fusion process. Technically, standardization enables various indicators to be calculated and compared within the same numerical range, ensuring that risk information from different dimensions is processed under the same standard and avoiding deviations caused by inconsistent dimensions.

[0058] The process involves extracting the numerical values ​​of standardized risk parameters and then comparing them with the corresponding threshold ranges of similar indicators in the early warning threshold factor. The threshold range in which each indicator value falls is recorded, resulting in a threshold comparison result. Technically, the standardized risk parameters are first extracted and then compared with the relevant threshold ranges in the early warning threshold factor. The early warning threshold factor, based on past transaction data and historical risk assessment results, includes risk level classifications and corresponding thresholds for different risk factors. Standardized risk parameter values ​​are mapped to their corresponding threshold ranges, and the range corresponding to each indicator value is recorded after the comparison. Determining the risk level range for each indicator clarifies its risk scope. Simple standardization results cannot reflect the severity of risk; comparison with preset thresholds is necessary to determine the specific risk level. Technically, by comparing with preset threshold ranges, each standardized risk parameter can be clearly mapped to its corresponding risk level range, ensuring that the assessment results of each risk factor meet the actual risk differentiation requirements.

[0059] Based on the proportion of each dimension in the risk weighting factors, risk information from different dimensions in the threshold comparison results is integrated to generate weighted risk information. In the technical implementation process, the risk of each dimension is weighted according to preset risk weighting factors. These risk weighting factors are determined based on historical data and industry standards, reflecting the importance of different dimensions to the overall risk assessment. Then, the risk information is merged according to the proportion of each dimension in the overall assessment to generate weighted comprehensive risk information. Each dimension's risk indicator has a different impact on the overall assessment; setting appropriate weights ensures that key dimensions' risk information has a greater proportion, accurately reflecting the contribution of different dimensions to the total risk and avoiding any one dimension being overemphasized or ignored. Technically, the weighting process achieves a comprehensive assessment of multiple risk information, ensuring that the contribution of each dimension's risk to the final result is consistent with reality, and improving the accuracy and rationality of the risk assessment.

[0060] By combining historical early warning results with experience in adjusting similar risks, and optimizing the detailed description of weighted risk information according to standards, dynamically revised risk information is obtained. In technical implementation, by analyzing historical early warning results and combining them with experience in handling historical risk events, the weighted risk information is dynamically revised, including optimizing the details of risk information to ensure that the assessment results better reflect actual risk performance. As financial markets continuously develop and trading models constantly change, traditional risk assessment standards may not be suitable for new forms of risk performance, necessitating adjustments based on historical data to ensure the dynamism and timeliness of risk information. Technically, combining historical adjustment experience optimizes the weighted risk information, making it more accurate and reliable. Dynamic revision ensures that risk assessment can adapt to new market environments, improving the real-time performance and accuracy of the assessment model.

[0061] The core risk indicators in dynamically revised risk information are compared with the risk level classification ranges to determine the corresponding level intervals, thus obtaining level interval assignment information. In technical implementation, the core risk indicators of the dynamically revised risk information are compared with the preset risk level classification ranges, which are set based on historical data and industry standards. The final level interval assignment is obtained through matching. Matching actual risk information with the preset risk level system ensures that the risk of each transaction is classified into an accurate level, guaranteeing that the risk assessment results reflect the true risk level of the transaction. Technically, this comparison process ensures that the risk assessment results are consistent with the preset risk level system, providing a scientific and accurate basis for subsequent decision-making.

[0062] Based on the correspondence between risk level range information and risk level names, the range is converted into a transaction risk level. In the technical implementation, according to the rules governing the correspondence between risk level range information and risk level names, the determined risk range is transformed into an actual transaction risk level. For example, if a transaction assessment result falls within the "high-risk" range, it is marked as "high-risk." All risk assessment information is ultimately transformed into a clear risk level, facilitating financial institutions or regulators to take appropriate risk control measures. Technically, this conversion ensures the operational feasibility of the risk assessment results, providing a direct basis for further risk warnings and control.

[0063] In one embodiment, the steps of obtaining a corresponding risk statement based on the degree of deviation from compliance with the pattern, and generating a preliminary risk description based on the risk statement, include: S3041. Obtain detailed deviation data based on the deviation degree and deviation characteristic decomposition standards of the pattern compliance; S3042. Obtain basic risk statement data based on the deviation details data and the risk statement correspondence table; S3043. Obtain the preset statement association logic, and obtain the associated risk statement data based on the basic risk statement data and the preset statement association logic; S3044. Obtain the preset deviation integrity checklist, and obtain the supplementary risk statement data based on the associated risk statement data and the preset deviation integrity checklist; S3045. Remove duplicate expressions from the structured risk description data, adjust the coherence and conciseness of the expressions according to the description simplification standard, and generate a preliminary risk description.

[0064] As described in steps S3041-S3045 above, this invention generates accurate, concise, and compliant preliminary risk descriptions based on the degree of deviation from compliance patterns, providing fundamental data for risk management and decision-making, and ultimately offering effective compliance risk early warning tools for enterprises, financial institutions, and regulatory agencies. In practical applications, compliance checks and risk warnings heavily rely on the identification and analysis of deviations from transaction patterns. The compliance level of different transaction behaviors and the identification of abnormal patterns directly affect the effectiveness of risk identification and management. Therefore, generating effective risk descriptions through in-depth analysis of the degree of deviation from compliance patterns is crucial for subsequent compliance judgments, potential risk assessments, and early warnings. If accurate, comprehensive, and easily understandable risk descriptions cannot be achieved, it will be difficult to support relevant decisions, potentially leading to risk management errors or excessive intervention. However, existing risk assessment methods largely rely on manual judgment, which suffers from strong subjectivity and low accuracy in identifying and describing compliance deviations. There is an urgent need to transform compliance deviations into standardized and accurate risk descriptions through automated and standardized processes.

[0065] Transforming compliance deviations into specific deviation details is crucial. Since the calculation results of pattern compliance deviations are mostly quantitative or qualitative indicators, they cannot directly provide sufficient details. Therefore, it is necessary to subdivide the deviation data according to established deviation characteristic decomposition standards to obtain information such as the deviation magnitude, nature of the deviation, and differences from the standard under each risk dimension. Technically, this relies on the structured processing of deviation data. By establishing deviation characteristic decomposition standards and setting rules such as deviation magnitude ranges and types, calculation algorithms are used to map pattern compliance deviations into specific deviation details. Specific operations include data querying, comparison, and automated calculation based on database rules to obtain detailed deviation data related to each transaction or behavior. This is achieved by decomposing the deviation.

[0066] A risk description mapping table is used to generate basic risk descriptions based on deviation detail data. Since the deviation detail data only describes the degree of deviation and lacks a clear risk description, a pre-defined risk description mapping table is needed to map different deviation details to corresponding basic risk descriptions. In the technical implementation, the risk description mapping table lists different deviation details and their corresponding risk descriptions; for example, a higher degree of deviation corresponds to "high risk," and financial inconsistency corresponds to "financial anomaly." An algorithm quickly queries the mapping table to generate a basic risk description for each transaction. Mapping deviation details to specific risk descriptions makes the risk situation of each transaction or behavior more specific and structured, improves the standardization of risk descriptions, reduces human subjective bias, and ensures the uniformity and consistency of risk assessment.

[0067] By refining and adjusting the basic risk statements through correlation logic, the potential simplification of the basic statements is addressed, resulting in more comprehensive and correlated risk statements. Technically, the preset correlation logic encodes the rules governing the relationships between risk statements. For example, "large transaction amount and unclear source" must be correlated with the statement "suspicious source of funds." Encoding these rules allows for the expansion and refinement of the basic statements. Through correlation logic processing, a more targeted, detailed, and complete risk description is built upon the basic statements, more accurately reflecting the risk characteristics behind the transaction.

[0068] A pre-defined deviation completeness checklist is used to verify the completeness of associated risk descriptions and to fill any gaps in the descriptions. This checklist contains the necessary elements for describing various risks. In practice, the system automatically checks the completeness of associated risk description data based on the checklist and supplements missing content using templates or a rule engine. This ensures that risk descriptions are complete and compliant with regulations, resulting in a comprehensive and accurate description that provides a solid foundation for subsequent decision-making.

[0069] Risk descriptions are streamlined and optimized by eliminating redundant content, ensuring concise and coherent descriptions, and improving report readability and effectiveness. Technically, text processing algorithms are used to remove duplicate risk descriptions, and the coherence and conciseness of the expression are optimized according to simplification standards. For example, natural language processing techniques are used to remove duplicate content and adjust sentence structure. While retaining key information, the readability of the descriptions is improved, providing decision-makers with a more efficient risk assessment tool.

[0070] In one embodiment, the step of determining the proportion of each level and generating risk weight factors according to regulations, based on the complexity of the fund flow hierarchy data and the frequency of regulatory parameter updates, includes: S4051. Obtain the account node sequence based on the fund transfer hierarchy data, extract the number of transfer accounts, account attributes and transfer interval information in the account node sequence, and integrate them to form hierarchical detailed data; S4052. Obtain hierarchical complex representation data based on hierarchical detail data; S4053. Obtain statistical records of frequency interval data and regulatory parameter update frequency, and classify the statistical records of frequency interval data and regulatory parameter update frequency into the corresponding intervals according to the time span division in the frequency segmentation standard to form frequency interval data. S4054. Obtain updated impact description data based on frequency range data; S4055. Obtain initial percentage data based on hierarchical complex representation data and updated impact description data; S4056. Obtain risk weight factors based on the initial percentage data and factor generation specifications.

[0071] As described in steps S4051-S4056 above, this invention generates risk weight factors by determining the proportion of each level according to the complexity of the fund flow hierarchy data and the frequency of regulatory parameter updates, in accordance with regulations. Specifically, through multi-step detailed analysis, relevant characteristics of fund flow hierarchy and information on the frequency of regulatory parameter updates are extracted, and finally, weight factors that can accurately reflect risks are generated. These factors will be further used in risk assessment, decision-making, and early warning to improve the ability to identify risks in financial activities such as cross-border payments. In the financial industry, especially in cross-border payment scenarios, the complexity of fund transfers and the frequent updates to regulatory requirements can affect the compliance and risk control of the payment process. Therefore, it is crucial to design a mechanism for dynamically calculating risk weight factors. Fund transfer hierarchical data contains information from multiple nodes, and the characteristics of each node (such as account attributes) and the transfer interval have different degrees of impact on payment risk. Detailed analysis of these hierarchical details is required. The frequent updates to regulatory parameters necessitate timely adjustments to the allocation of risk factors in a dynamically changing environment. The above process ensures that risk weight factors are reasonably allocated based on actual circumstances during fund transfers, thereby enhancing the accuracy and real-time performance of the risk assessment model.

[0072] Extracting relevant features of account nodes from fund transfer data helps understand the relationships between nodes. Each account node in a fund transfer may represent a different role (such as an intermediary account, the final receiving account, etc.). The number, attributes, and transfer intervals of nodes are crucial for risk assessment. For example, frequent transfers may indicate high risk, and specific attributes such as foreign accounts and non-resident accounts may increase compliance risks. Traditional methods usually only focus on explicit factors such as transaction amount and payment time, ignoring the diversity of account levels and the complexity of transfer paths. This problem is solved by detailed analysis of each node: Technically, by stepping through the fund transfer path, the account nodes in each transaction are identified, and information such as account type (such as personal account, corporate account), transfer interval (time required for fund transfer), and whether intermediary accounts are involved are extracted. After data integration and preprocessing, hierarchical detailed data is formed, providing a foundation for subsequent complex hierarchical representation.

[0073] This invention integrates the characteristics of each node in the fund transfer path to form higher-level complex representation data, revealing the complexity of the transfer. The transfer of each account node is not only related to a single transaction, but may also involve multiple levels of transfer. It is necessary to synthesize node features to understand the path complexity and provide a basis for generating risk factors. Traditional methods cannot integrate various types of information, process account data at each level in isolation, and lack in-depth analysis of the relationships between levels. This invention solves this problem by transforming hierarchical detailed data into hierarchical complex representation data: In terms of technical implementation, it aggregates information such as the attributes and transfer intervals of each account node, analyzes the relationships and interactions between nodes, and comprehensively considers factors such as the number of nodes, path complexity, and the potential contribution of each node to the overall risk to quantify and model the fund transfer path, obtain the overall complexity, and form hierarchical complex representation data.

[0074] Based on the update frequency of regulatory parameters, frequency interval data for fund flow activities in different time periods is established. The update frequency of regulatory parameters reflects the speed of changes in the financial environment. By dividing the frequency intervals, the impact of different update frequencies on the complexity of fund flow paths can be clearly identified, allowing for dynamic adjustments to the flexibility of risk assessment. Traditional methods neglect the impact of regulatory parameter update frequency and use fixed parameter standards. This problem is solved by dividing the frequency intervals: Technically, the update frequency of regulatory parameters is statistically analyzed to obtain update frequency data for each time period. This data is then categorized into corresponding frequency intervals according to a preset time span standard. The risk assessment model is dynamically adjusted through this classification, providing a basis for subsequent risk factor calculations.

[0075] The impact of regulatory parameter updates is analyzed based on frequency interval data, generating descriptive data. Frequently updated regulatory parameters may have different effects on financial behavior, necessitating the generation of accurate data describing the impact of updates to reflect their role in the potential risks of fund flows. Technically, the specific content of regulatory parameter updates within different frequency intervals is analyzed, and combined with update frequency characteristics, data describing the impact of updates is generated, providing a necessary basis for the subsequent generation of risk weighting factors at different levels.

[0076] Taking into account the complexity of the hierarchical structure and the frequency of regulatory parameter updates, preliminary percentage data is calculated to lay the foundation for the calculation of risk weighting factors. By analyzing the complexity of each level and the status of regulatory updates, the generated initial percentage data can provide key support for the final risk assessment. Technically, the hierarchical complexity representation data is combined with update impact description data, and an algorithm is used to calculate the initial percentage for each level. This data reflects the importance of each level in the fund flow path and the impact of update frequency on it.

[0077] Based on the initial proportion data, final risk weight factors are generated according to specifications. These risk weight factors are calculated based on the proportions of each level and reflect the contribution of different levels and update frequencies to the overall risk, serving as an important basis for risk assessment and decision-making. Technically, the initial proportion data is combined with the factor generation specifications to calculate the risk weight factors for each level. These factors will be used in subsequent risk assessments to ensure that the assessment results dynamically reflect the risk situation in cross-border payments.

[0078] In one embodiment, the step of obtaining risk characteristic assessment information corresponding to transaction amount information, payment purpose information, and instruction initiation time information based on the feature correlation index includes: S2051. Obtain multi-dimensional correlation data based on feature correlation index and information dimension classification criteria; S2052. Obtain transaction amount-related related content based on multi-dimensional related data, and obtain specific characteristics such as the correlation performance and correlation strength between the amount and other transaction elements based on the transaction amount-related related content, and generate amount-related feature data based on the correlation performance and correlation strength. S2053. Based on the usage association fragments and usage feature extraction criteria in the multidimensional association data, obtain the usage association feature data; S2054. Obtain time-related feature data based on the time-related segments and time feature extraction criteria in the multi-dimensional related data; S2055. Obtain risk element data based on the amount-related characteristic data, purpose-related characteristic data, and time-related characteristic data; S2056. Obtain risk characteristic assessment information based on the risk element data of the information segment.

[0079] As described in steps S2051-S2056 above, this invention uses a feature correlation index-based analysis method to assess the risk characteristics of cross-border payment transactions. Specifically, it obtains corresponding risk characteristic assessment information through correlation analysis of transaction amount information, payment purpose information, and instruction initiation time information, thereby generating compliance risk warning data and effectively identifying and warning of potential payment risks. This method addresses the complexity and high frequency of cross-border payments; a single data dimension (such as amount, purpose, or time) cannot capture the risk signals behind the transaction. Therefore, it is necessary to form a comprehensive risk assessment system through multi-type data correlation and feature extraction. This multi-type analysis not only improves the accuracy of risk detection but also enables risk warnings to reflect non-compliant or abnormal behaviors in transactions more promptly and accurately.

[0080] By identifying the relationships between different transaction features through feature correlation indices, we can prepare for subsequent risk assessments. Feature correlation indices quantify the correlation between different feature dimensions, and information dimension classification standards ensure that each dimension accurately reflects the transaction situation. Traditional methods ignore feature correlations and only analyze single features (such as only looking at transaction amount). By setting preset feature correlation standards to determine the classification of information dimensions (amount, purpose, time, etc.), we analyze the correlation between each dimension based on feature correlation indices, and dynamically calculate and adjust them in combination with actual transaction data to obtain multi-dimensional correlation data containing correlation information across all dimensions, laying the foundation for comprehensive assessment.

[0081] The correlation and strength of transaction amounts with other elements (purpose, time) are analyzed to determine whether transaction amounts conform to normal patterns or pose potential risks. Traditional amount risk assessment ignores other risk factors behind the amount (such as abnormal purpose or inconsistent timing). This method extracts amount-related content from multi-dimensional correlation data and uses statistical methods such as correlation coefficients and covariance to assess the strength of the correlation between the amount and other elements, generating amount correlation feature data.

[0082] By analyzing the correlation between payment purpose and transaction purpose, abnormal or unusual uses can be identified. Payment purpose is an important component of transactions, and different purposes represent different risk levels. Existing methods treat purpose as a static factor and fail to capture its changes in different scenarios in real time. This method analyzes the purpose-related segments in multi-dimensional correlation data, and combines purpose feature extraction criteria to extract feature data such as purpose type, frequency, and trend of change, thereby identifying transaction purposes that do not conform to normal patterns.

[0083] Identifying potential anomalies through transaction time information, time feature analysis can reveal transactions occurring during abnormal time periods (potentially stemming from fraud, money laundering, etc.), and is crucial for comprehensive risk assessment. Current technologies for extracting time features are simplistic, focusing only on specific points in time while ignoring time intervals, periodicity, and other signals. This approach involves extracting time-related segments from multi-dimensional correlated data, obtaining time-related feature data based on time feature extraction standards, analyzing time distribution patterns, and combining this with data from other dimensions to enhance assessment accuracy.

[0084] By aggregating correlational characteristic data across dimensions such as amount, purpose, and time, a multi-faceted risk element data set is formed, comprehensively reflecting the transaction risk level. Existing technologies process data from each dimension independently, failing to create a comprehensive assessment system. By integrating the aforementioned correlational characteristic data of amount, purpose, and time, a comprehensive risk element set encompassing all dimensions of transaction characteristics is constructed.

[0085] By comprehensively analyzing the risk element data of segmented information, risk characteristic assessment information is generated, providing a basis for classifying transaction risk levels. This information is the core data for risk early warning and directly determines subsequent early warning strategies and handling measures. Traditional methods rely on static rules and cannot be dynamically adjusted. By using a risk assessment model to process the risk element data of segmented information, an assessment result containing information such as risk level, potential risk points, and risk type is generated, serving as an important basis for subsequent risk early warning.

[0086] In one embodiment, the step of obtaining risk element data based on amount-related characteristic data, purpose-related characteristic data, and time-related characteristic data includes: S20551. Obtain financial risk element data based on financial correlation characteristic data; S20552. Obtain application risk factor data based on application-related characteristic data; S20553. Obtain time risk element data based on time-related characteristic data; S20554. Generate sub-information risk element data based on amount risk element data, purpose risk element data, and time risk element data.

[0087] As described in steps S20551-S20554 above, the present invention uses a reasonable feature data association and extraction method to analyze and extract risk factors from three types of data: amount-related feature data, purpose-related feature data, and time-related feature data, and finally integrates them to generate information risk factor data.

[0088] Based on the correlation characteristics of monetary amounts, risk factor data is obtained. Extracting monetary risk factors is crucial, as monetary amount is a significant source of risk in cross-border payments. High-value transactions, in particular, indicate greater complexity in fund flows, making accurate extraction essential for assessing regulatory compliance. Technically, the relationship between monetary amount and other correlation characteristics is analyzed first to obtain the distribution of monetary data and determine its position within the overall transaction volume. For large transactions, the focus is on analyzing the payment purpose and timing; for example, unusually large transactions require consideration of purpose and timing to determine if they involve cross-border money laundering or illicit fund flows. Simultaneously, risk coefficients are calculated based on factors such as the strength and trend of monetary correlation. Risk is assessed by establishing a correlation model between monetary amount and other transaction elements. For instance, when the transaction amount exceeds a threshold and the purpose is fund remittance, higher monetary risk factor data is generated.

[0089] This step involves obtaining usage risk factor data based on usage-related characteristic data. The aim is to deduce potential risks by analyzing payment usage characteristics. The legality and compliance of the usage directly relate to the risk level; for example, "personal remittances" have lower risk, while "inter-company fund transfers" with large amounts carry higher risk. Technically, data such as payment purpose, recipient information, and transaction description are first extracted based on the usage type. This data is then analyzed in conjunction with information such as amount and time to identify high-risk usages that may involve illegal activities such as gambling or terrorist financing. By comparing usage data with industry standards and regulatory rules, compliance is determined, and usage risk factor data is generated. For example, large transactions for "import payments" that do not match trade legality or usage descriptions generate higher usage risk factors.

[0090] Time-related risk element data is obtained based on time-related feature data. The goal of this step is to extract time-dimensional risk elements. High-risk transactions often occur during specific time windows such as holidays and nighttime; identifying these time characteristics can improve the accuracy of early warnings. Technically, the transaction time series data is first analyzed to identify abnormal patterns at time points, such as large transactions occurring at 1 AM which have a higher risk level. Then, the transaction time is compared with historical transaction time data to assess whether it matches normal trading patterns. For example, transactions occurring during low-frequency trading periods are identified as high-risk, and time-related risk element data is generated.

[0091] This step generates risk element data based on three risk factors: monetary value, purpose, and time. The goal is to integrate these three types of risk factors to achieve a comprehensive risk assessment. A single-dimensional risk factor cannot reflect the overall risk of a transaction; integration creates a more complete risk profile. Technically, the three types of risk element data are first weighted and fused, with weights adjusted based on factors such as importance, relevance, and transaction scenario (e.g., monetary value and time are more important than purpose in some scenarios, and purpose is more crucial in legality verification). Then, through comprehensive analysis of the fused risk data, an overall transaction risk level is generated, ultimately forming the risk element data.

[0092] In one embodiment, the step of generating sub-information risk element data based on amount risk element data, purpose risk element data, and time risk element data includes: S205541. Obtain structured monetary risk element data based on monetary risk element data; S205542. Obtain structured use risk factor data based on use risk factor data; S205543. Obtain structured time risk element data based on time risk element data; S205544. The structured amount risk element data, structured purpose risk element data, and structured time risk element data are respectively mapped to the transaction amount, payment purpose, and instruction initiation time, and sub-information risk element data are generated based on the transaction amount, payment purpose, and instruction initiation time.

[0093] As described in steps S205541-S205544 above, this invention extracts risk element data from three dimensions—amount, purpose, and time—and performs structured processing to generate sub-information risk element data. This supports cross-border payment compliance risk early warning and detection, thereby more effectively assessing and predicting transaction risks, improving the accuracy of risk identification and early warning, and ultimately achieving automated risk detection and early warning.

[0094] In cross-border payment scenarios involving direct bank-enterprise connections, key elements of transaction information, such as amount, purpose, and time, are inherently correlated and directly impact payment compliance and risk levels. To effectively assess potential risks, it is necessary not only to accurately extract these data elements but also to uncover their relationships. Through classification, correlation, and structured analysis, the potential risks under different transaction models can be revealed.

[0095] Structured financial risk element data is obtained based on financial risk element data. This step aims to transform the raw financial risk element data into a standardized structure. Financial amount, as a core element of transactions, is directly related to compliance and risk level, requiring precise structuring. In the technical implementation, basic information is first extracted from the financial field of transaction logs, covering the size and fluctuations of each transaction. Then, the financial amounts are classified according to preset rules and decomposed and transformed using a specific algorithm to generate structured data. For example, amounts exceeding a threshold are marked as high-risk, and frequent fluctuations within a short period are identified as abnormal fund flows, triggering an early warning signal. This structuring process can accurately identify abnormal transactions at the financial level, providing crucial support for risk assessment.

[0096] Structured payment purpose risk factor data is obtained based on the payment purpose risk factor data. The purpose of payment relates to the flow, source, and purpose of funds. Some purposes (such as sensitive cross-border remittances and suspicious funding) imply high compliance risks and require meticulous structured processing. In practice, payment purpose information is first extracted from the original payment purpose risk factor data (such as the transaction purpose field); then, it is classified and analyzed according to specific standards and algorithms to refine the classification of purposes, label risk levels based on characteristics, and standardize information such as purpose type, objective, and beneficiary into structured data. For example, purposes such as "illegal fundraising" and "money laundering" are directly labeled as high-risk, triggering strict review and monitoring. The structured payment purpose data can help quickly identify high-risk transaction scenarios.

[0097] Structured time risk element data is obtained based on time risk element data. Time characteristics are implicitly correlated with high-risk behaviors (such as money laundering and terrorist financing). High-frequency trading and nighttime trading patterns often imply risks, requiring structured processing to uncover time-dimensional risks. The process first extracts transaction time information from the raw time risk element data; then, combined with transaction time characteristics, algorithms such as time window analysis are used to identify abnormal time patterns. For example, high-frequency short-term trading and large transactions during inactive periods such as early morning are marked as high-risk. Structured time data can improve the risk assessment system and, combined with other factors, enhance prediction accuracy.

[0098] The three types of structured risk element data are mapped to the transaction amount, payment purpose, and instruction initiation time fields, respectively, to generate sub-information risk element data. This step is the core of multi-type data fusion, requiring precise docking and integration between structured risk data and the original transaction fields. First, an association mapping is established: structured amount data is associated with the transaction amount field, structured purpose data is associated with the payment purpose field, and structured time data is associated with the instruction initiation time field, ensuring that the risk characteristics of each transaction are closely bound to the original information; then, the three types of data are integrated through a fusion algorithm to form sub-information risk element data containing multi-type risk information.

[0099] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent results or equivalent process transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A compliance risk early warning detection method for bank-to-bank cross-border payment, characterized in that, The method comprises the following steps: obtaining a raw transaction sequence of cross-border payment between banks and enterprises, synchronously collecting transaction object information, fund flow direction data and error-related regulatory parameter sequence; obtaining a transaction mode component according to the raw transaction sequence; generating a preliminary risk warning parameter according to the transaction mode component and the error-related regulatory parameter sequence; obtaining the cumulative number of payment transactions, generating a warning threshold factor and a risk weight factor based on the cumulative number of payment transactions and fund flow direction data; generating a transaction risk level by adaptively fusing the preliminary risk warning parameter, the warning threshold factor and the risk weight factor; judging whether the transaction risk level is less than a preset level; if yes, judging that the corresponding transaction risk level is within a safe range; if no, judging that the corresponding transaction risk level has risks, and generating a warning information according to the transaction risk level.

2. The compliance risk early warning detection method for silver enterprise direct connection cross-border payment according to claim 1, characterized in that, The step of obtaining the transaction mode component according to the raw transaction sequence comprises the following steps: obtaining a classification segmentation sequence according to the raw transaction sequence, and obtaining a transaction data group according to the classification segmentation sequence; obtaining a transaction field set according to the transaction data group, wherein the transaction field set comprises transaction amount information, payment purpose information and instruction initiation time information; obtaining a transaction feature segment according to the transaction amount information, the payment purpose information and the instruction initiation time information, and splitting the transaction data according to a fixed time window to obtain transaction time sequence feature information; obtaining a feature correlation index according to the transaction time sequence feature information and the error-related regulatory parameter sequence; obtaining risk feature evaluation information corresponding to the transaction amount information, the payment purpose information and the instruction initiation time information according to the feature correlation index; obtaining the transaction mode component according to the risk feature evaluation information corresponding to the transaction amount information, the payment purpose information and the instruction initiation time information and a preset screening standard.

3. The compliance risk early warning detection method for silver enterprise direct connection cross-border payment according to claim 1, characterized in that, The step of generating the preliminary risk warning parameter according to the transaction mode component and the error-related regulatory parameter sequence comprises the following steps: obtaining a preset regulatory parameter mapping table, disassembling constituent elements of the transaction mode component, and extracting matching regulatory parameters to form a subset by comparing the elements and the parameters in the preset regulatory parameter mapping table; obtaining the completeness of mandatory fields and the compliance of field content format of each subset according to the associated regulatory parameter subset, and generating a parameter compliance result according to the completeness of mandatory fields and the compliance of field content format; comparing the actual feature description of the transaction mode component with the compliance feature requirement corresponding to the associated regulatory parameter subset to obtain a mode compliance deviation; obtaining a corresponding risk expression according to the mode compliance deviation, and generating a preliminary risk description according to the risk expression; matching the preliminary risk description with a risk situation in a risk level corresponding table, and extracting a risk quantization identifier corresponding to the matching situation; generating a revised preliminary risk coefficient by combining the latest change information of the regulatory parameter according to the preliminary risk coefficient and the parameter dynamic adjustment description, and obtaining the preliminary risk warning parameter according to the revised preliminary risk coefficient.

4. The compliance risk early warning detection method for silver enterprise direct connection cross-border payment according to claim 1, characterized in that, The step of obtaining the cumulative number of payment transactions, generating the warning threshold factor and the risk weight factor based on the cumulative number of payment transactions and the fund flow direction data comprises the following steps: According to the bank-enterprise direct connection transaction log, the cumulative number of payment business and the transaction timestamp are obtained, each transaction record in the transaction log is traversed, the total number of records is counted to obtain the cumulative number, and the time field in each record is extracted to obtain the transaction timestamp; According to the preset time interval, the transaction timestamp is classified into the corresponding time window, and the number of transaction records in each window is counted to obtain the transaction frequency per unit time; According to the account receiving sequence in the fund flow direction data, the account transfer level and the number of nodes passed by the fund are marked in combination with the account transfer track, and the fund transfer level data is obtained; The cumulative number and the transaction frequency per unit time are compared with the same type of data in the threshold reference data set, and the threshold reference corresponding to the matching data is extracted as the early warning threshold factor; According to the complexity of the fund transfer level data and the frequency of the regulatory parameter update, the proportion of each level is determined according to the specification to generate the risk weight factor.

5. The compliance risk early warning detection method for silver enterprise direct connection cross-border payment according to claim 1, characterized in that, The steps of generating the transaction risk level by adaptively fusing the preliminary risk early warning parameter, the early warning threshold factor and the risk weight factor include: According to the preliminary risk early warning parameter and the parameter normalization standard, the standardized risk parameter is obtained, and according to the dimension conversion requirement of each risk index in the parameter normalization standard, the different dimension indexes in the preliminary risk early warning parameter are converted into numerical values in a unified range; The index value in the standardized risk parameter is extracted, the threshold range of the same type index in the early warning threshold factor is searched, the threshold interval in which the index value is located is recorded, and the threshold comparison result is obtained; According to the proportion of each dimension in the risk weight factor, the risk information of different dimensions in the threshold comparison result is integrated to generate weighted risk information; According to the adjustment experience of the same type of risk in the historical early warning result, the details of the weighted risk information are optimized according to the specification to obtain the dynamically corrected risk information; The core risk index in the dynamically corrected risk information is compared with the risk level division range to determine the corresponding level interval, and the level interval attribution information is obtained; According to the corresponding relationship between the level interval attribution information and the level name, the interval attribution is converted into the transaction risk level.

6. The compliance risk early warning detection method for silver enterprise direct connection cross-border payment according to claim 3, characterized in that, The steps of generating the preliminary risk description according to the mode compliance deviation degree and the risk expression include: According to the mode compliance deviation degree and the deviation characteristic disassembly standard, the deviation detail data is obtained; According to the deviation detail data and the risk expression corresponding table, the basic risk expression data is obtained; The preset expression association logic is obtained, and the associated risk expression data is obtained according to the basic risk expression data and the preset expression association logic; The preset deviation integrity check list is obtained, and the completed risk expression data is obtained according to the associated risk expression data and the preset deviation integrity check list; The repeated expressions in the structured risk description data are removed, the coherence and simplicity of the expressions are adjusted according to the description simplification standard, and the preliminary risk description is generated.

7. The compliance risk early warning detection method for silver enterprise direct connection cross-border payment according to claim 4, characterized in that, The steps of generating the risk weight factor according to the complexity of the fund transfer level data and the frequency of the regulatory parameter update, and determining the proportion of each level according to the specification include: According to the fund flow level data, an account node sequence is obtained, the number of transfer accounts, account attributes and flow interval information in the account node sequence are extracted, and level detail data is formed by integration; According to the level detail data, level complex representation data is obtained; Frequency interval data and statistical records of regulatory parameter update frequency are obtained, and are classified into corresponding intervals according to the time span division in the frequency segmentation standard to form the frequency interval data; According to the frequency interval data, update influence description data is obtained; According to the level complex representation data and the update influence description data, initial proportion data is obtained; According to the initial proportion data and the factor generation specification, risk weight factors are obtained.

8. The compliance risk early warning detection method for silver enterprise direct connection cross-border payment according to claim 2, characterized in that, According to the feature correlation index, the risk feature evaluation information corresponding to the transaction amount information, the payment purpose information and the instruction initiation time information is obtained, and the steps include: According to the feature correlation index and the information dimension division standard, multi-dimensional correlation data is obtained; According to the transaction amount related correlation content obtained from the multi-dimensional correlation data, the correlation performance, the correlation strength and other specific features of the amount and other transaction elements are obtained, and the amount correlation feature data is generated according to the correlation performance, the correlation strength and other specific features; According to the purpose correlation segment in the multi-dimensional correlation data and the purpose feature extraction standard, purpose correlation feature data is obtained; According to the time correlation segment in the multi-dimensional correlation data and the time feature extraction standard, time correlation feature data is obtained; According to the amount correlation feature data, the purpose correlation feature data and the time correlation feature data, multi-information risk element data is obtained; According to the multi-information risk element data, risk feature evaluation information is obtained.

9. The compliance risk early warning detection method for silver enterprise direct connection cross-border payment according to claim 8, characterized in that, According to the amount correlation feature data, the purpose correlation feature data and the time correlation feature data, multi-information risk element data is obtained, and the steps include: According to the amount correlation feature data, amount risk element data is obtained; According to the purpose correlation feature data, purpose risk element data is obtained; According to the time correlation feature data, time risk element data is obtained; According to the amount risk element data, the purpose risk element data and the time risk element data, multi-information risk element data is generated.

10. The compliance risk early warning detection method for silver enterprise direct connection cross-border payment according to claim 9, characterized in that, According to the amount risk element data, the purpose risk element data and the time risk element data, multi-information risk element data is generated, and the steps include: According to the amount risk element data, structured amount risk element data is obtained; According to the purpose risk element data, structured purpose risk element data is obtained; According to the time risk element data, structured time risk element data is obtained; The structured amount risk element data, the structured purpose risk element data and the structured time risk element data are respectively corresponded to the transaction amount, the payment purpose and the instruction initiation time, and multi-information risk element data is generated according to the transaction amount, the payment purpose and the instruction initiation time.

Images