Business risk evaluation method and device

Abstract

The invention provides a business risk evaluation method and device, which can be applied to the technical field of big data. The business risk evaluation method comprises the following steps: in response to received transaction data of a financial transaction, scene feature data associated with a current business scene of the financial transaction is generated according to the transaction data and business characteristic parameters, the transaction data comprises basic transaction data and equipment use condition data, and the business characteristic parameters are determined according to the business scene; comprising a data acquisition window for transaction data and calculation logic; and inputting the scene feature data as risk identification features into the business risk evaluation model to obtain a risk score representing the size of the business risk.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of big data, and in particular to a business risk evaluation method and device. BACKGROUND

[0002] In the financial business scenario, as the business volume of the banking financial market quotation and transaction system increases, the operation and maintenance personnel need to set different monitoring for the increasing business types and scenarios. The existing monitoring method only uses a single index or a simple combination of indexes when evaluating risks, and the risk score accuracy is low.

[0003] Therefore, a new business risk evaluation method is needed. SUMMARY

[0004] In view of the above problems, the present application provides a business risk evaluation method and device.

[0005] According to a first aspect of the present application, a business risk evaluation method is provided, characterized in that it comprises: in response to receiving transaction data of a financial transaction, generating scene feature data associated with the current business scenario of the financial transaction according to the transaction data and business characteristic parameters, the transaction data comprising basic transaction data and device usage data, the business characteristic parameters being determined according to the business scenario and comprising a data collection window and a calculation logic of the transaction data; inputting the scene feature data as a risk identification feature into a business risk evaluation model to obtain a risk score representing the size of the business risk.

[0006] According to an embodiment of the present application, in response to receiving transaction data of a financial transaction, generating scene feature data associated with the current business scenario of the financial transaction according to the transaction data and business characteristic parameters comprises: determining the transaction data within the data collection window according to the data collection window and the transaction data, wherein the data collection window is determined according to the time period of the transaction data fluctuation required for evaluating the business risk in the business scenario; determining the scene feature data according to the calculation logic and the transaction data within the data collection window, wherein the scene feature data comprises characteristic parameters representing the possibility of the occurrence of corresponding business risks in the business scenario.

[0007] According to an embodiment of this application, the business scenario has transaction data that fluctuates over a time period. The step of generating scenario feature data associated with the current business scenario of the financial transaction based on the transaction data and business characteristic parameters in response to receiving financial transaction data further includes: determining the business data for the current time period and the transaction data for the previous time period based on the current time period, the previous time period, and the received transaction data; and performing difference or ratio calculations on the transaction data for the current time period and the transaction data for the previous time period to obtain at least one of the following parameters: rolling average parameter, rolling standard deviation parameter, rolling maximum value parameter, rolling minimum value parameter, month-on-month growth acceleration parameter, and month-on-month decline acceleration parameter, as the scenario feature data.

[0008] According to an embodiment of this application, the business scenario includes a scenario in which a financial event occurs. The step of responding to the receipt of transaction data for a financial transaction and generating scenario feature data associated with the current business scenario of the financial transaction based on the transaction data and business characteristic parameters further includes: determining transaction data for the first time period and transaction data for the second time period based on a first time period before the financial event occurs, a second time period after the financial event occurs, and the received transaction data; determining the difference value based on the transaction data for the first time period and the transaction data for the second time period to obtain scenario feature data describing the impact caused by the financial event.

[0009] According to an embodiment of this application, the method further includes: determining first scene feature data and second scene feature data from multiple scene feature data; calculating a correlation parameter between the first scene feature data and the second scene feature data; and, if the correlation parameter is greater than a preset threshold, using one of the first scene feature data and the second scene feature data as recommended scene feature data, so that the recommended scene feature data can be used as a risk identification feature to input into a business risk assessment model to obtain a contextualized risk score representing the magnitude of business risk.

[0010] According to an embodiment of this application, the method further includes: determining a second weight for the scenario feature data as a risk identification feature based on the business needs for evaluating business risks in the business scenario and / or the correlation between the scenario feature data and business risks; modifying the first weight set for the scenario feature data as a risk identification feature in the business risk evaluation model to the second weight to obtain a new business risk evaluation model, so as to obtain a contextualized risk score based on the new business risk evaluation model.

[0011] According to an embodiment of this application, the business risk assessment model is further configured with a scenario correction factor for scaling the scenario-based risk score based on the fluctuation of the transaction data and / or the occurrence of financial events. The step of inputting the scenario feature data as risk identification features into the business risk assessment model to obtain a scenario-based risk score representing the magnitude of business risk includes: the business risk assessment model determining an initial risk score based on the scenario feature data; and the business risk assessment model determining a risk score based on the initial risk score and the scenario correction factor.

[0012] A second aspect of this application provides a business risk assessment device, characterized in that it includes: a scenario feature data determination module, used to generate scenario feature data associated with the current business scenario of the financial transaction based on the transaction data and business characteristic parameters received from a financial transaction, wherein the transaction data includes basic transaction data and equipment usage data, and the business feature parameters are determined according to the business scenario, including a data acquisition window and calculation logic for the basic transaction data and equipment usage data; and a risk score determination module, used to input the scenario feature data as risk identification features into a business risk assessment model to obtain a risk score representing the magnitude of business risk.

[0013] A third aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0014] A fourth aspect of this application also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0015] The fifth aspect of this application also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described method. Attached Figure Description

[0016] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0017] Figure 1 The illustrations depict application scenarios of the business risk assessment method, apparatus, device, medium, and program products according to embodiments of this application.

[0018] Figure 2 A schematic diagram illustrating a business risk assessment method according to an embodiment of this application is shown.

[0019] Figure 3 A schematic diagram of a business risk assessment device according to an embodiment of this application is shown.

[0020] Figure 4 A block diagram of an electronic device suitable for implementing a business risk assessment method according to an embodiment of this application is shown schematically. Detailed Implementation

[0021] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0022] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0023] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0024] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0025] It should be noted that the business risk assessment methods, apparatus, equipment, media, and program products defined in this application can be used in the fields of big data technology and fintech, and can also be used in a variety of other fields besides those mentioned above. The application fields of the business risk assessment methods, apparatus, equipment, media, and program products provided in the embodiments of this application are not limited.

[0026] In the technical solution of this application, the user information (including but not limited to user data, user personal information, user image information, user device information, such as location information) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with relevant laws, regulations, and standards, and necessary measures have been taken to ensure that they do not violate public order and good morals. Corresponding operation entry points are provided for users to choose to authorize or refuse.

[0027] In scenarios involving automated decision-making using personal information, the methods, devices, and systems provided in this application all offer users corresponding entry points for choosing to agree to or reject the automated decision-making results. If the user chooses to reject, the process proceeds to the expert decision-making stage. Here, "automated decision-making" refers to the activity of automatically analyzing and evaluating an individual's behavioral habits, interests, or economic, health, and credit status through computer programs, and then making a decision. Here, "expert decision-making" refers to the activity of making decisions by personnel who specialize in a particular field, possess specialized experience, knowledge, and skills, and have reached a certain level of professional expertise.

[0028] Embodiments of this application provide a business risk assessment method, apparatus, device, medium, and program product.

[0029] Figure 1 The illustration shows an application scenario diagram of the business risk assessment method, apparatus, device, medium, and program product according to embodiments of this application.

[0030] like Figure 1 As shown, application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0031] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only).

[0032] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0033] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0034] It should be noted that the business risk assessment method provided in this application embodiment can generally be executed by server 105. Correspondingly, the business risk assessment device provided in this application embodiment can generally be located in server 105. The business risk assessment method provided in this application embodiment can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the business risk assessment device provided in this application embodiment can also be located in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105.

[0035] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0036] The following will be based on Figure 1 The described scene, through Figure 2 The business risk assessment method according to the embodiments of this application will be described in detail.

[0037] Figure 2 A schematic diagram illustrating a business risk assessment method according to an embodiment of this application is shown. Figure 2 As shown, the business risk assessment method 200 of this embodiment includes steps 210-220, and the business risk assessment method 200 can be executed in an electronic device.

[0038] In step 210, in response to receiving transaction data of a financial transaction, scenario feature data associated with the current business scenario of the financial transaction is generated based on the transaction data and business characteristic parameters.

[0039] According to one implementation, the received transaction data may include basic transaction data and device usage data. Basic transaction data includes multi-source data, such as market data stream data, transaction data stream data, event and public opinion data stream data, and contextual data.

[0040] According to one implementation, market data stream data includes real-time market data for the underlying asset, such as trading price and trading volume; transaction data stream data includes trading instructions, transaction records, and order status generated by users or the system; event and public opinion data stream data includes data that impacts market trading, such as financial news and information, listed company announcements, regulatory documents, social media sentiment analysis results, and the release of macroeconomic indicators; contextual data includes data describing market trading conditions, including current market status labels (volatility labels determined based on volatility indices, etc.), a calendar of major events, holiday information, and underlying asset attribute information (industry, market capitalization, beta coefficient, etc.).

[0041] According to one implementation, the device usage data includes operational data of servers, electronic devices, etc. used for transactions, including performance indicators (such as CPU, memory, etc. utilization), network status (including latency data, etc.), database load, and critical service operation logs, etc.

[0042] According to one implementation, this application can determine business characteristic parameters based on a business scenario. The business characteristic parameters include a data acquisition window for transaction data and calculation logic. These parameters are used to perform calculations on the transaction data to obtain scenario characteristic data.

[0043] According to one implementation, business characteristic parameters are used to describe the specificity of a business scenario, that is, how risk assessment differs from other business scenarios when conducted in the current business scenario; wherein, the data acquisition window describes how long the data needs to be collected in this business scenario, and which time periods the data is collected in order to better conduct risk assessment; and the calculation logic describes how to process the transaction data to obtain more advanced, more business-specific, and more risk-aware scenario characteristic data.

[0044] According to one implementation, generating scenario characteristic data associated with the current business scenario of a financial transaction based on transaction data and business characteristic parameters includes: determining the transaction data within a data acquisition window based on the data acquisition window and the transaction data, wherein the data acquisition window is determined based on the time period of transaction data fluctuation required to evaluate business risk in the business scenario; and determining scenario characteristic data based on calculation logic and the transaction data within the data acquisition window, wherein the scenario characteristic data includes characteristic parameters characterizing the probability of corresponding business risks occurring in the business scenario. Since some transaction data is volatile, to avoid the impact of transaction data volatility on risk assessment, it is necessary to determine a data acquisition time window and use calculation logic to process the data within this time window to eliminate transaction data volatility, making the obtained scenario characteristic data more reflective of whether the transaction data has business risks and improving the accuracy of the determined risk score.

[0045] According to one implementation, the business scenario has transaction data that fluctuates over a time period. In response to receiving transaction data for a financial transaction, generating scenario feature data associated with the current business scenario of the financial transaction based on the transaction data and business characteristic parameters further includes: determining the business data for the current time period and the transaction data for the previous time period based on the current time period, the previous time period, and the received transaction data; performing difference or ratio calculations on the transaction data for the current time period and the transaction data for the previous time period to obtain at least one of the following parameters: rolling average parameter, rolling standard deviation parameter, rolling maximum value parameter, rolling minimum value parameter, month-on-month growth acceleration parameter, and month-on-month decline acceleration parameter, as the scenario feature data.

[0046] According to one implementation, the rolling mean, rolling standard deviation, rolling maximum value, and rolling minimum value are determined based on the transaction data of the current time period; the rolling mean, rolling standard deviation, rolling maximum value, and rolling minimum value are also determined based on the transaction data of the previous time period; the difference between the rolling mean of the current time period and the rolling mean of the previous time period is then calculated to obtain the rolling mean parameter; similarly, the difference calculation can be performed to obtain the rolling standard deviation parameter, rolling maximum value parameter, and rolling minimum value parameter, which serve as scene feature data.

[0047] According to one implementation, the time period can be determined as 7 days. The transaction data in the previous time period includes the consumption amount data of last week; the transaction data of this week includes the consumption amount data of this week; the rolling average determined by the transaction data in the previous time period includes the average consumption amount data of last week; the rolling average determined by the transaction data in the current time period includes the average consumption amount data of this week; and the rolling standard deviation parameter includes the change in the average consumption amount data determined based on the average consumption amount data of last week and the average consumption amount data of this week, as the rolling average parameter.

[0048] According to one implementation method, different business scenarios have different business objectives, risk points, and user behavior patterns, requiring different scenario characteristic data to represent different business scenarios, such as the rate of change in abnormal transactions in financial risk control. For data that fluctuates over time, attention can be paid to trends, stability, abnormal fluctuations, and comparisons with benchmarks, thereby effectively avoiding the impact of periodic fluctuations in technical data on risk assessment, and thus accurately evaluating business risks.

[0049] According to one implementation, a proportional calculation can be performed based on the transaction data of the current time period and the transaction data of the previous time period to obtain the acceleration parameters of month-on-month growth and month-on-month decline, and the acceleration parameters of month-on-month growth and month-on-month decline can be used as scene feature data.

[0050] According to one implementation, the growth rate of transaction data in the previous time period and the growth rate of transaction data in the current time period can be calculated. A proportional calculation is performed based on the growth rates of transaction data in the previous and current time periods to obtain the acceleration parameter for month-on-month growth. Similarly, the decline rate of transaction data in the previous and current time periods can be calculated. A proportional calculation is performed based on the decline rates of transaction data in the previous and current time periods to obtain the acceleration parameter for month-on-month decline.

[0051] According to one implementation, the scene feature data can also be set as a parameter to be compared with a benchmark value. For example, the benchmark value is set as the 95th percentile of historical transaction data, the data collection window is set to three consecutive days, and the scene feature data is set as transaction users, transaction items, etc. that exceed the 95th percentile of historical transaction data for three consecutive days.

[0052] According to one implementation, the benchmark value can also be set to twice the average volatility of the past 30 days, the data collection window can be set to 1 day, and the scenario characteristic data can be set to trading users or trading projects whose current trading volatility exceeds twice the average volatility of the past 30 days.

[0053] According to one implementation, the business scenario includes a scenario in which a financial event occurs. In response to receiving transaction data of a financial transaction, generating scenario feature data associated with the current business scenario of the financial transaction based on the transaction data and business characteristic parameters further includes: determining transaction data for the first time period and transaction data for the second time period based on a first time period before the financial event occurs, a second time period after the financial event occurs, and the received transaction data; determining the difference value based on the transaction data for the first time period and the transaction data for the second time period to obtain scenario feature data describing the impact caused by the financial event.

[0054] According to one implementation method, financial events include company listings, successful financing, and failed financing, etc. This application does not limit the specific type of financial event. The first time period before the financial event can be set as 1 hour or 1 day before the financial event occurs; the second time period after the financial event can be set as 1 hour or 1 day after the financial event is sent. Scene characteristic data includes differences in transaction amounts and transaction quantities between the first and second time periods, etc., to assess the impact of the financial event and whether it has triggered business risks. By comparing data before and after the financial event, the business risk caused by the financial event can be accurately evaluated.

[0055] According to one implementation method, if it is believed that business risks need to be evaluated during certain special time periods, data collection windows can be set, such as setting the data collection window to a preset time period for the opening call auction, a preset time period before the closing, and the window period for major events, etc.

[0056] According to one implementation, when determining scenario characteristic data from device usage data, the device load change rate is used as scenario characteristic data to describe the changes in device load at different times, thereby describing the changes in the number of user transaction operations. If the number of user transaction operations increases abnormally, it indicates that there may be business risks.

[0057] Subsequently, step 220 is executed, whereby the scenario feature data is input as risk identification features into the business risk assessment model to obtain a risk score representing the magnitude of business risk. In this application, by determining more accurate and relevant scenario feature data during risk assessment, the risk score determined by the model is made more accurate, reducing the likelihood of missed business risk reports.

[0058] According to one implementation, scene feature data can be selected, for example, removing scene feature data with a high data missing rate and scene feature data with a high proportion of single values.

[0059] According to one embodiment, scenario feature data can be further selected based on correlation, and the selected scenario feature data is input into the model. When multiple scenario feature data are determined, a first scenario feature data and a second scenario feature data can be determined from the multiple scenario feature data; the first scenario feature data and the second scenario feature data are different scenario feature data. Subsequently, a correlation parameter between the first scenario feature data and the second scenario feature data is calculated; if the correlation parameter is greater than a preset threshold, one of the first scenario feature data and the second scenario feature data is selected as the recommended scenario feature data, so that the recommended scenario feature data can be used as a risk identification feature input into the business risk assessment model to obtain a contextualized risk score representing the magnitude of business risk. This application can optimize feature selection and improve model performance by analyzing correlation, such as improving the efficiency and accuracy of risk assessment output. According to one embodiment, the preset threshold of the correlation parameter can be set to a value such as 95%, and those skilled in the art can set it as needed. When selecting one scenario feature data from the first scenario feature data and the second scenario feature data as the recommended scenario feature data, the importance of each scenario feature data can be determined according to the degree of influence of the first scenario feature data and the second scenario feature data on risk assessment, and the most important scenario feature data is determined according to the importance of each scenario feature data as the recommended scenario feature data.

[0060] According to one implementation method, a high-level model, namely a business risk assessment model, can be pre-trained for risk assessment. Specifically, the high-level model may be a neural network model, and this application does not limit the specific implementation method of the high-level model. When training the business risk assessment model, multiple samples can be constructed, each sample having a risk score and at least one scenario feature data. The samples are then input into the neural network model for training to obtain the business risk assessment model.

[0061] According to one implementation, a primary model can also be trained for supplementary risk assessment. During training, multiple samples can be constructed, each with a risk score and at least one transaction data point. The samples are then input into the primary model for training. Transaction data may include commonly used financial technical indicators such as yield, volatility, relative strength index (RSI), moving average, and buying / selling pressure. This application does not limit the specific type of transaction data. The primary model may employ a neural network model, and this application does not limit the specific implementation method of the primary model. The primary risk score obtained by inputting the transaction data into the primary model can be weighted and added to the risk score obtained based on the business risk assessment model to obtain a more accurate business risk assessment.

[0062] According to one implementation, scenario feature data is further determined as a second weight for risk identification features based on business needs for evaluating business risks within a business scenario and / or the correlation between scenario feature data and business risks. The first weight set for scenario feature data as a risk identification feature in the business risk assessment model is then modified to the second weight, resulting in a new business risk assessment model. This new model is used to obtain a contextualized risk score. For example, changes in the business scenario include a shift from active to slowed trading, and business needs include different trading time periods. Different scenario feature data play different roles in risk assessment under different business scenarios, requiring adjustments to the weights of the scenario feature data. The correlation between scenario feature data and business risks changes under different business scenarios or at different time periods within a business scenario.

[0063] By assigning different weights to scenario feature data based on business needs, risk assessments based on scenario feature data can better align with changing business scenarios. In one implementation, different scenario feature data can be input into the risk assessment model as the business scenario, time period, business needs, and the correlation between scenario feature data and business risks change. Through this approach, the risk identification features and their weights corresponding to the scenario feature data are no longer static but dynamically calculated, context-dependent, and their weights adjustable. By adjusting the weights, the expression of risk identification features can better meet business needs, resulting in more accurate risk scores.

[0064] According to one implementation method, the weight adjustment of risk identification features can be determined using linear regression coefficients. The calculated linear regression coefficients naturally represent the direction and strength of the contribution of risk identification features to the prediction results; the larger the coefficient value, the greater the weight. Meanwhile, for some business features, weight adjustment needs to be made based on business experience.

[0065] According to one implementation, the business risk assessment model also has a scenario correction factor for scaling the scenario-based risk score based on the fluctuation of transaction data and / or the occurrence of financial events. The scenario feature data is input into the business risk assessment model as a risk identification feature to obtain a scenario-based risk score that represents the magnitude of business risk. This includes: the business risk assessment model determining an initial risk score based on the scenario feature data; and the business risk assessment model determining a risk score based on the initial risk score and the scenario correction factor.

[0066] According to one implementation method, external environmental factors can be introduced to fine-tune the model output. A scenario correction factor k is defined, where k > 1 indicates that risk needs to be amplified. The scenario correction factor can be set differently for different time periods. A risk score is obtained by combining the scenario correction factor with the initial risk score, and then the risk score is compared with an alarm threshold. If the threshold is exceeded, a structured alarm record is generated, including the triggering reason (including the contribution of important scenario feature data), risk score, confidence level, relevant data snapshot, and scenario information. External environmental factors include: if the transaction data for that period is inherently volatile, a scenario correction factor less than 1 can be set to reduce the risk score; if the transaction data for that period is inherently volatile, a scenario correction factor greater than 1 can be set to amplify the risk score. If a financial event causes significant volatility in the transaction data, a smaller scenario correction factor can be set; if a financial event causes relatively low volatility, a larger scenario correction factor can be set. By setting the scenario correction factor, risk assessment can be made more sensitive or less sensitive in different scenarios, improving the accuracy of the obtained risk score.

[0067] According to one implementation, a two-level triggering of the alarm mechanism can also be set up, including using a primary model as a fast filtering layer, and also using a lightweight rule engine or a simple model (such as rules based on dynamic thresholds) to handle core, easily judged risks that require extremely low latency (such as: violations of hard compliance rules, system crashes, price flash crashes exceeding extreme dynamic thresholds), and output preliminary, high-confidence real-time alarms.

[0068] Set up a high-level model as a deep analysis layer, such as a business risk assessment model trained with a more complex machine learning model, and input scenario feature data. The business risk assessment model can use supervised learning, setting valid alarm labels to 1 and invalid ones to 0. Before training, divide the dataset into training and validation sets. During training, use cross-validation to adjust hyperparameters, focusing on optimizing recall to ensure no alarm events are missed, while controlling precision.

[0069] The final output can be the output of the primary model, the risk score from the business risk model, or a risk score obtained by integrating the primary risk score with certain weights. Then, based on the current business scenario (including market status, events, time periods, etc.), the performance indicators of the business risk model within a certain time window are calculated in real time: recall, precision, false positive rate, etc. When the false positive rate rises continuously above a preset threshold or the business scenario changes, the model's judgment threshold is dynamically adjusted or the score is modified according to the context. For example, during periods of high volatility in the overall market, the threshold for triggering alarms is appropriately increased; during major event windows, a more sensitive model is activated for specific targets.

[0070] According to one implementation, alarm grading can also be set: alarms are automatically classified into different severity levels based on the final risk score, confidence level, and event type (market risk, operational risk, credit risk, system risk). Based on information such as alarm type, level, associated object, and responsible team, alarms are precisely routed to the corresponding personnel or systems. Furthermore, an adaptive suppression mechanism is implemented: it identifies related alarm clusters (multiple alarms triggered by the same root cause). Highly relevant, low-level subsequent alarms are intelligently suppressed or aggregated, retaining only the most core representative alarms. A time- or event-based silent period / cooling-off period is set to avoid repeated alarm bombardment for the same abnormal state.

[0071] According to one implementation, a model feedback loop is also set up: providing a convenient interface for operators to mark alarm handling results, including setting them to "real risk," "false alarm," or "needs adjustment," and allowing for supplementary handling instructions and additional information. Core indicators, including alarm effectiveness and recall rate, are calculated periodically (or triggered), and false alarm and missed alarm cases are analyzed to pinpoint model or rule deficiencies (such as feature failure, unreasonable thresholds, or insufficient context coverage). Using newly labeled data, the business risk assessment model is incrementally or periodically retrained, adjusting feature weights, updating dynamic threshold strategies, and optimizing context correction rules. Newly labeled data can come from user feedback or actual risk event records. Subsequently, the optimized model / rules are seamlessly deployed back to the online system. This ultimately forms a closed loop of "data—alarm—feedback—learning—optimization—re-alarm," driving the model's adaptive evolution and continuously improving alarm accuracy.

[0072] This application constructs a business risk assessment model to improve the effectiveness of risk alarms for financial market businesses. This model overcomes the rigidity of static threshold alarms, enabling adaptation to dynamic market fluctuations and differences in underlying assets. It effectively integrates and correlates multi-source, heterogeneous financial market data to uncover deeper, more accurate anomaly indicators of risk. It incorporates crucial contextual information (such as market conditions, events, and time) into the alarm decision-making process, enhancing the targeting and accuracy of alarms. Furthermore, it establishes a closed-loop feedback mechanism to continuously optimize the alarm model using historical alarm handling experience, avoiding alarm fatigue.

[0073] This application utilizes a multimodal model to improve the effectiveness of alerts in financial market applications. It employs multi-dimensional indicators, combined with contextual patterns and user feedback, to enhance system reliability. Compared to traditional alert detection models, this application offers the following advantages: Significantly reduced false alarm and false negative rates: Through dynamic thresholds, contextual awareness, and multi-dimensional correlation analysis, it effectively distinguishes between normal market fluctuations and genuine anomalies, significantly reducing invalid alerts. Composite features and intelligent models can capture more hidden and complex risk patterns, while contextual awareness ensures that critical risks are not overlooked during special periods. Enhanced alert effectiveness and reduced alert fatigue: Accurately identifying truly relevant risk events ensures high-value alerts. Hierarchical, routing, and suppression mechanisms effectively reduce information overload, and closed-loop learning continuously reduces false alarm sources. Improved response efficiency and optimized resource utilization: Structured and interpretable alert information (including triggering reasons and context) helps users quickly understand the problem and make decisions. A hierarchical model ensures rapid response to core risks. Simultaneously, it reduces the human resource consumption for handling invalid alerts, allowing for a more precise focus on addressing genuine risks. Highly adaptive: Dynamic features, contextual correction, and closed-loop learning enable the system to automatically adapt to changes in the market environment, the emergence of new risk patterns, and the characteristics of different targets.

[0074] Figure 3 A schematic diagram of a business risk assessment apparatus according to an embodiment of this application is shown. The following will be combined with... Figure 3 The device is described in detail. Figure 3 As shown, the business risk assessment device 300 includes a scenario feature data determination module 310 and a risk score determination module 320. The modules in the business risk assessment device 300 work together to assess business risks.

[0075] The scenario feature data determination module 310 is used to respond to received transaction data of a financial transaction, and generate scenario feature data associated with the current business scenario of the financial transaction based on the transaction data and business characteristic parameters. The transaction data includes basic transaction data and device usage data. The business feature parameters are determined according to the business scenario, including the data acquisition window and calculation logic for the basic transaction data and device usage data. In one embodiment, the scenario feature data determination module 310 can be used to execute step 210 described above, which will not be repeated here.

[0076] The risk scoring determination module 320 is used to input scenario feature data as risk identification features into the business risk assessment model to obtain a risk score that represents the magnitude of business risk.

[0077] According to one embodiment, the scenario feature data determination module 310 includes a first determination module, used to determine the transaction data within the data acquisition window based on the data acquisition window and the transaction data, wherein the data acquisition window is determined based on the time period of the transaction data fluctuation required to evaluate business risks in a business scenario; and a second determination module, used to determine scenario feature data based on the calculation logic and the transaction data within the data acquisition window, wherein the scenario feature data includes characteristic parameters that characterize the probability of corresponding business risks occurring in the business scenario.

[0078] According to one embodiment, the first determining module includes a first determining submodule, used to determine the business data of the current time period and the transaction data of the previous time period based on the current time period, the previous time period, and the received transaction data; the second determining module includes a second determining submodule, used to determine the business data of the current time period and the transaction data of the previous time period based on the current time period, the previous time period, and the received transaction data.

[0079] According to one implementation, the first determining module further includes a third determining submodule, used to determine the transaction data of the first time period and the transaction data of the second time period based on the first time period before the financial event, the second time period after the financial event, and the received transaction data; and a fourth determining submodule, used to determine the difference value based on the transaction data of the first time period and the transaction data of the second time period to obtain scenario feature data describing the impact caused by the financial event.

[0080] According to one embodiment, the apparatus further includes a correlation parameter module, configured to determine first scene feature data and second scene feature data from multiple scene feature data; calculate a correlation parameter between the first scene feature data and the second scene feature data; and, if the correlation parameter is greater than a preset threshold, use one of the first scene feature data and the second scene feature data as recommended scene feature data, so that the recommended scene feature data can be used as a risk identification feature input into a business risk assessment model to obtain a contextualized risk score representing the magnitude of business risk.

[0081] According to one embodiment, the apparatus further includes a weight update module, which is used to determine the second weight of the scenario feature data as a risk identification feature based on the business needs for evaluating business risks in a business scenario and / or the correlation between scenario feature data and business risks; and to modify the first weight set for the scenario feature data as a risk identification feature in the business risk assessment model to the second weight to obtain a new business risk assessment model, so as to obtain a contextualized risk score based on the new business risk assessment model.

[0082] According to one implementation, the risk score determination module 320 further includes a scenario correction factor module. The business risk assessment model is also configured with a scenario correction factor for scaling the scenario-based risk score based on the fluctuation of transaction data and / or the occurrence of financial events. The scenario correction factor module is used to call the business risk assessment model to determine the initial risk score based on scenario feature data; and to call the business risk assessment model to determine the risk score based on the initial risk score and the scenario correction factor.

[0083] Figure 4 A block diagram of an electronic device suitable for implementing a business risk assessment method according to an embodiment of this application is shown schematically.

[0084] like Figure 4 As shown, an electronic device 400 according to an embodiment of this application includes a processor 401, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 402 or a program loaded from a storage portion 408 into a random access memory (RAM) 403. The processor 401 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 401 may also include onboard memory for caching purposes. The processor 401 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0085] RAM 403 stores various programs and data required for the operation of electronic device 400. Processor 401, ROM 402, and RAM 403 are interconnected via bus 404. Processor 401 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 402 and / or RAM 403. It should be noted that the programs may also be stored in one or more memories other than ROM 402 and RAM 403. Processor 401 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in one or more memories.

[0086] According to embodiments of this application, the electronic device 400 may further include an input / output (I / O) interface 405, which is also connected to a bus 404. The electronic device 400 may also include one or more of the following components connected to the input / output (I / O) interface 405: an input section 406 including a keyboard, mouse, etc.; an output section 407 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN card, modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output (I / O) interface 405 as needed. A removable medium 411, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 410 as needed so that computer programs read from it can be installed into the storage section 408 as needed.

[0087] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0088] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 402 and / or RAM 403 and / or one or more memories other than ROM 402 and RAM 403 described above.

[0089] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code is used to enable the computer system to implement the business risk assessment method provided in the embodiments of this application.

[0090] When the computer program is executed by the processor 401, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0091] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via communication section 409, and / or installed from removable medium 411. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0092] In such an embodiment, the computer program can be downloaded and installed from a network via communication section 409, and / or installed from removable medium 411. When the computer program is executed by processor 401, it performs the functions defined in the system of this application embodiment. According to embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0093] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0094] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0095] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A business risk assessment method, characterized in that, include: In response to receiving transaction data for a financial transaction, scenario feature data associated with the current business scenario of the financial transaction is generated based on the transaction data and business characteristic parameters. The transaction data includes basic transaction data and device usage data. The business characteristic parameters are determined according to the business scenario and include a data acquisition window for the transaction data and calculation logic. The scenario feature data is used as risk identification features and input into the business risk assessment model to obtain a risk score that represents the magnitude of business risk.

2. The method according to claim 1, characterized in that, The step of responding to received transaction data for a financial transaction and generating scenario feature data associated with the current business scenario of the financial transaction based on the transaction data and business characteristic parameters includes: The transaction data within the data acquisition window is determined based on the data acquisition window and the transaction data, wherein the data acquisition window is determined based on the time period of transaction data fluctuation required to evaluate business risk in the business scenario. The scenario feature data is determined based on the calculation logic and the transaction data within the data acquisition window. The scenario feature data includes characteristic parameters that characterize the likelihood of corresponding business risks occurring in the business scenario.

3. The method according to claim 2, characterized in that, The business scenario has transaction data that fluctuates over time. The step of responding to received financial transaction data and generating scenario feature data associated with the current business scenario of the financial transaction based on the transaction data and business characteristic parameters further includes: The business data for the current time period and the transaction data for the previous time period are determined based on the current time period, the previous time period, and the received transaction data. The difference or ratio calculation is performed between the transaction data of the current time period and the transaction data of the previous time period to obtain at least one of the following parameters: rolling mean parameter, rolling standard deviation parameter, rolling maximum value parameter, rolling minimum value parameter, month-on-month growth acceleration parameter, and month-on-month decline acceleration parameter, which shall be used as the scene feature data.

4. The method according to claim 2, characterized in that, The business scenarios include scenarios where financial events occur. The step of responding to received transaction data from a financial transaction and generating scenario feature data associated with the current business scenario of the financial transaction based on the transaction data and business characteristic parameters further includes: The transaction data for the first time period and the transaction data for the second time period are determined based on the first time period before the financial event, the second time period after the financial event, and the received transaction data. Based on the transaction data of the first time period and the transaction data of the second time period, the difference value is determined to obtain scenario feature data describing the impact of the financial event.

5. The method according to any one of claims 1-4, characterized in that, Also includes: Determine the first scene feature data and the second scene feature data from multiple scene feature data; Calculate the correlation parameter between the feature data of the first scene and the feature data of the second scene; If the correlation parameter is greater than a preset threshold, one of the first scenario feature data and the second scenario feature data is used as the recommended scenario feature data, so that the recommended scenario feature data can be used as risk identification features to input into the business risk assessment model to obtain a contextualized risk score that represents the magnitude of business risk.

6. The method according to any one of claims 1-4, characterized in that, Also includes: Based on the business needs for evaluating business risks in the business scenario and / or the correlation between the scenario feature data and business risks, the scenario feature data is determined as the second weight of the risk identification feature; The first weight of the scenario feature data as a risk identification feature in the business risk assessment model is modified to the second weight to obtain a new business risk assessment model, so as to obtain a contextualized risk score based on the new business risk assessment model.

7. The method according to any one of claims 1-4, characterized in that, The business risk assessment model also includes a scenario correction factor for scaling the scenario-based risk score based on the fluctuations in the transaction data and / or the occurrence of financial events. The process of inputting the scenario feature data as risk identification features into the business risk assessment model to obtain a scenario-based risk score representing the magnitude of business risk includes: The business risk assessment model determines an initial risk score based on the scenario feature data; The business risk assessment model determines the risk score based on the initial risk score and the scenario correction factor.

8. A business risk assessment device, characterized in that, include: The scenario feature data determination module is used to respond to received transaction data of financial transactions, and generate scenario feature data associated with the current business scenario of the financial transaction based on the transaction data and business characteristic parameters. The transaction data includes basic transaction data and equipment usage data. The business feature parameters are determined according to the business scenario, including a data acquisition window and calculation logic for the basic transaction data and equipment usage data. The risk scoring determination module is used to input the scenario feature data as risk identification features into the business risk assessment model to obtain a risk score that represents the magnitude of business risk.

9. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1-7.

11. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1-7.

Images