Risk control index calculation method, system and device and storage medium

Abstract

The invention discloses a risk control index calculation method, system and device and a storage medium. The method comprises the steps of obtaining a target risk control index corresponding to a target risk control scene and a target expression and a target rule corresponding to the target risk control index; constructing a target feature value dependency tree of the target risk control index according to the target expression; acquiring risk control event data corresponding to the target risk control scene; determining a target index value of the target risk control index according to the risk control event data, the target expression and the target feature value dependency tree; and determining a target risk control identification result according to the target index value and the target rule, and displaying the target risk control identification result through a preset visual system. Therefore, the risk control index is calculated by constructing the feature value dependence tree, and accurate calculation of the risk control index is realized.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method, system, device and storage medium for calculating risk control indicators. Background Technology

[0002] With the rapid development of financial services and the increasing demands for risk management, risk control systems are playing an increasingly important role in financial institutions. However, traditional risk control systems are prone to errors and are difficult to manage when handling complex transaction scenarios and massive amounts of data. Furthermore, the use of different indicator sets in different projects and scenarios can lead to problems such as chaotic management of business indicator versions, low operational security, and high labor costs associated with deployment.

[0003] Therefore, a risk control indicator calculation method is needed to achieve more accurate indicator calculation capabilities. Summary of the Invention

[0004] This application provides a method, system, device, and storage medium for calculating risk control indicators. By constructing a feature value dependency tree to calculate risk control indicators, accurate calculation of risk control indicators is achieved.

[0005] In a first aspect, embodiments of this application provide a method for calculating risk control indicators, the method comprising: Obtain the target risk control indicators corresponding to the target risk control scenario, as well as the target expressions and target rules corresponding to the target risk control indicators; Based on the target expression, identify the dependent indicators in the dependency relationship of the target risk control indicator to obtain n dependent indicators, and determine the dependency relationship corresponding to the n dependent indicators to obtain n dependency relationships; n is a natural number; Determine the initial feature value corresponding to the target risk control indicator; determine the n dependent feature values ​​corresponding to the n dependent indicators; Construct a target feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependency relationships; Obtain the risk control event data corresponding to the target risk control scenario; The target indicator value of the target risk control indicator is determined based on the risk control event data, the target expression, and the target feature value dependency tree. The target risk control identification result is determined based on the target indicator value and the target rule, and the target risk control identification result is displayed through a preset visualization system.

[0006] Secondly, embodiments of this application provide a risk control indicator calculation system, which includes: a risk control indicator determination unit, a dependent indicator determination unit, a feature value calculation unit, a feature value dependency tree construction unit, a risk control event data acquisition unit, an indicator value calculation unit, and a risk control identification result display unit, wherein... The risk control indicator determination unit is used to obtain the target risk control indicator corresponding to the target risk control scenario, as well as the target expression and target rule corresponding to the target risk control indicator; The dependency indicator determination unit is used to identify the dependency indicators in the dependency relationship of the target risk control indicator according to the target expression, to obtain n dependency indicators, and to determine the dependency relationship corresponding to the n dependency indicators, to obtain n dependency relationships; n is a natural number; The feature value calculation unit is used to determine the initial feature value corresponding to the target risk control indicator; and to determine the n dependent feature values ​​corresponding to the n dependent indicators; The feature value dependency tree construction unit is used to construct a target feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependency relationships; The risk control event data acquisition unit is used to acquire risk control event data corresponding to the target risk control scenario; The indicator value calculation unit is used to determine the target indicator value of the target risk control indicator based on the risk control event data, the target expression, and the target feature value dependency tree. The risk control identification result display unit is used to determine the target risk control identification result based on the target indicator value and the target rule, and to display the target risk control identification result through a preset visualization system.

[0007] Thirdly, embodiments of this application provide an electronic device, including: a processor, a memory, a communication interface, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the programs include instructions for performing the steps in the first aspect of embodiments of this application.

[0008] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program for electronic data interchange, wherein the computer program causes a computer to perform some or all of the steps described in the first aspect of embodiments of this application.

[0009] Fifthly, embodiments of this application provide a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps described in the first aspect of embodiments of this application. The computer program product may be a software installation package.

[0010] It can be seen that the embodiments of this application have the following beneficial effects: By implementing the embodiments of this application, target risk control indicators corresponding to the target risk control scenario, as well as target expressions and target rules corresponding to the target risk control indicators, are obtained; dependent indicators in the dependency relationships of the target risk control indicators are identified according to the target expressions, resulting in n dependent indicators, and the dependency relationships corresponding to the n dependent indicators are determined, resulting in n dependency relationships; initial feature values ​​corresponding to the target risk control indicators are determined; n dependent feature values ​​corresponding to the n dependent indicators are determined; a target feature value dependency tree is constructed according to the initial feature values, the n dependent feature values, and the n dependency relationships; risk control event data corresponding to the target risk control scenario is obtained; target indicator values ​​of the target risk control indicators are determined according to the risk control event data, the target expressions, and the target feature value dependency tree; the target risk control identification result is determined according to the target indicator values ​​and the target rules, and the target risk control identification result is displayed through a preset visualization system. It can be seen that by constructing a feature value dependency tree to calculate risk control indicators, accurate calculation of risk control indicators is achieved. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the background art, the accompanying drawings used in the embodiments of this application or the background art will be described below.

[0012] Figure 1 This is a flowchart illustrating a risk control indicator calculation method provided in an embodiment of this application; Figure 2 This is a schematic diagram of a feature value dependency tree without circular dependencies provided in an embodiment of this application; Figure 3 This is a schematic diagram of a feature value dependency tree with circular dependencies provided in an embodiment of this application; Figure 4 This is a schematic diagram of a circular dependency detection process provided in an embodiment of this application; Figure 5 This is a schematic diagram of node update of a feature value dependency tree provided in an embodiment of this application; Figure 6 This is an application scenario diagram of a risk control indicator calculation method provided in an embodiment of this application; Figure 7This is a schematic diagram of the structure of a risk control indicator calculation system provided in an embodiment of this application; Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0014] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0015] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0016] The following describes the relevant content, concepts, meanings, technical issues, technical solutions, and beneficial effects involved in the embodiments of this application.

[0017] First, let me explain some of the terms used in this application: AviatorScript Expression Engine: The AviatorScript Expression Engine is a lightweight, high-performance expression engine designed for Java applications. Its core function is to dynamically parse and execute expressions, such as mathematical calculations, logical judgments, and business rules, allowing developers to adjust business logic in real time by configuring expressions without modifying the code.

[0018] Please see Figure 1 , Figure 1This is a flowchart illustrating a risk control indicator calculation method provided in an embodiment of this application. The method includes, but is not limited to, the following steps: S101. Obtain the target risk control indicators corresponding to the target risk control scenario, as well as the target expressions and target rules corresponding to the target risk control indicators.

[0019] In this embodiment of the application, the target risk control scenario refers to a specific business scenario selected or created by the user in the visualization system based on actual risk management needs, such as anti-fraud transaction monitoring scenario, credit approval scenario, etc.

[0020] In this embodiment, the target risk control indicator refers to a specific calculation unit used to quantify the degree of risk or business behavior characteristics under the target risk control scenario, such as the transaction amount in the past hour or the number of abnormal account logins in a single day. The target expression refers to a formulaic description of the calculation logic of the target risk control indicator. For example, the expression for calculating the transaction amount in the past hour can be set as sum(amount where time between now()-3600 and now()), where amount represents the transaction amount field and time represents the transaction time field. The target rule refers to the business logic specification for risk judgment based on the calculation results of the target risk control indicator, such as triggering an alarm when the transaction amount in the past hour is greater than 10,000, or freezing the account when the number of abnormal account logins in a single day is greater than 3.

[0021] In a specific embodiment, the user first determines the target risk control scenario through the scenario management interface of the visual system. The user can either directly select the target risk control scenario that meets the requirements from the list of historical risk control scenarios stored in the system, or create a target risk control scenario by entering information such as scenario name and business scope through the scenario creation function provided by the interface.

[0022] After determining the target risk control scenario, the system will automatically associate it with the resource management table. Based on the business identifier (such as scenario ID) corresponding to the scenario, the system will filter out the resource set bound to the scenario from the resource management table, and then extract the target risk control indicators that belong to the risk control indicator category.

[0023] Next, the system can obtain the target expression corresponding to the target risk control indicator based on the associated records in the feature value table. This target expression was previously written by the user through the visual interface during the indicator configuration phase and bound to the target risk control indicator for storage. Simultaneously, the system can also query the associated risk judgment logic in the resource management table based on the identifier information of the target risk control indicator, thereby obtaining the corresponding target rules.

[0024] It should be noted that the resource management table refers to a dedicated data table established in this application embodiment to achieve the binding management of scenarios and resources. Since various indicators, rules and other resources are required in scenario products, and different resources are in different tables, in order to ensure that each resource table only focuses on recording the information of the resource itself and does not perform scenario product management work, the resource management table can be used to uniformly manage the relationship between resources and scenarios and products, while supporting multi-version management, validity identification and anomaly tracking of resources.

[0025] It is evident that by acquiring core elements such as target risk control indicators, target expressions, and target rules for the target risk control scenario, we can ensure that subsequent risk control indicator calculations and risk decisions are all centered around specific business scenarios. This effectively avoids resource confusion in different scenarios, provides a clear data foundation for subsequent accurate and efficient risk control indicator calculations, and enhances the relevance and reliability of the risk control process.

[0026] Optionally, the above step of obtaining the target risk control indicators corresponding to the target risk control scenario, as well as the target expressions and target rules corresponding to the target risk control indicators, specifically includes the following steps: The target risk control indicator bound to the target risk control scenario, as well as the target expression and target rule corresponding to the target risk control indicator, are obtained from the preset resource management table. The resource management table includes the binding relationship between scenarios and resources. Each binding relationship includes: scenario ID, resource type, and resource ID. The resource type includes one of the following: indicator, expression, and rule.

[0027] In this embodiment, the pre-configured resource management table refers to a dedicated data table pre-configured to store the association between scenes and various resources. It contains built-in scene-resource binding relationships to enable targeted resource invocation. The resource management table includes scene-resource binding relationships, and each binding relationship includes scene ID, resource type, resource ID, etc. The resource type includes one of the following: metric, expression, or rule.

[0028] Among them, the scenario ID is a unique identifier assigned to each risk control scenario; the resource ID is a unique identifier assigned to each type of resource (including indicators, expressions, and rules), used to locate the specific resource instance in the corresponding resource table; the resource type is a clear division of the category to which the resource belongs, specifically including three categories: indicators, expressions, and rules, which are consistent with the name of each resource table. For example, if the rule table is RULE, then the resource type is RULE.

[0029] In a specific embodiment, when obtaining the target risk control indicators, target expressions, and target rules corresponding to the target risk control scenario, the scenario ID corresponding to the target risk control scenario can be used as a search condition to perform a matching query in a preset resource management table, filtering out all binding relationship records containing that scenario ID. The resource ID of the corresponding resource can be determined based on the binding relationship record containing that scenario ID.

[0030] Based on the resource type recorded in each binding relationship record, and combined with the corresponding resource ID, further location can be performed.

[0031] When the resource type is an indicator, the corresponding target risk control indicator is retrieved from the indicator table using the resource ID; when the resource type is an expression, the target expression corresponding to the target risk control indicator is retrieved from the expression table using the resource ID; when the resource type is a rule, the target rule corresponding to the target risk control indicator is retrieved from the table using the resource ID. Through the above hierarchical retrieval and positioning based on scenario ID, resource type, and resource ID, the target risk control indicator corresponding to the target risk control scenario, as well as the target expression and target rule corresponding to the target risk control indicator, are finally obtained.

[0032] As can be seen, by performing hierarchical retrieval based on the association between scenario ID, resource type and resource ID in the preset resource management table, the core resources required for the target risk control scenario can be accurately and targeted, effectively avoiding confusion and mismatch in the resource call process, and improving the efficiency and accuracy of resource acquisition.

[0033] S102. Identify the dependent indicators in the dependency relationship of the target risk control indicator according to the target expression, obtain n dependent indicators, and determine the dependency relationship corresponding to the n dependent indicators, thus obtaining n dependency relationships.

[0034] In this embodiment, dependent indicators refer to other indicators that need to be referenced during the calculation of the target risk control indicator. These indicators provide basic data support for the calculation of the target risk control indicator and are prerequisites for the generation of the target risk control indicator. Dependency relationship refers to the reference association logic between the target risk control indicator and dependent indicators, and between dependent indicators and other dependent indicators. It is used to clarify which dependent indicator results need to be used as input when the indicator is calculated, and the calculation association method between the two.

[0035] In a specific embodiment, an expression execution engine (such as AviatorScript) can be invoked to perform syntax parsing and semantic extraction on the target expression, identifying other indicator identifiers necessary for generating the target risk control indicator from the calculation logic of the target expression. Based on these indicator identifiers, the corresponding indicators can be determined from the resource management table, obtaining the first-level dependent indicators of the target risk control indicator. Since the first-level dependent indicators also have their corresponding dependent indicators, i.e., the second-level dependent indicators, the corresponding second-level dependent indicators can be determined based on the expression of the first-level dependent indicators. This process is repeated recursively to obtain n dependent indicators, where n is a natural number. These n dependent indicators all serve as dependent indicators of the target risk control indicator. For example, when the target expression can complete the calculation without referencing other indicators, n is taken as 0.

[0036] When n dependent indicators are obtained, each dependent indicator has its corresponding dependency relationship, resulting in n dependency relationships. For example, if the target risk control indicator depends on indicator 1 and indicator 2, and indicator 1 depends on indicator 3, for a total of 3 dependent indicators, then the dependency relationship is: target risk control indicator depends on indicator 1, target risk control indicator depends on indicator 2, and indicator 1 depends on indicator 3.

[0037] It is evident that by accurately identifying the dependent indicators and corresponding dependencies of the target risk control indicators through parsing the target expression, the prerequisite data requirements for the calculation of the target risk control indicators can be clearly defined. This provides a core basis for subsequently constructing feature value dependency trees, determining the order of indicator calculation, and detecting cyclic dependencies, ensuring the logic and accuracy of the risk control indicator calculation process.

[0038] S103. Determine the initial feature value corresponding to the target risk control indicator; determine the n dependent feature values ​​corresponding to the n dependent indicators.

[0039] In this embodiment, a feature value is an abstract concept describing an indicator. Each indicator corresponds to a unique feature value. Based on the dependencies between indicators, i.e., the dependencies between feature values, a dependency tree can be formed. For example, if the target risk control indicator is the transaction amount in the past hour, then the feature value is a value used to describe the transaction amount in the past hour, which can be represented by the identifier OneHourAmount. The value of OneHourAmount is calculated by the expression sum(amount where time between now()-3600 and now()), where amount is the dependent indicator, the feature value identifier of the transaction amount.

[0040] In a specific embodiment, a search can be performed in a preset feature value table to determine the initial feature value corresponding to the target risk control indicator, and to determine the n dependent feature values ​​corresponding to the n dependent indicators. The preset feature value table stores the binding relationship between indicators and feature values ​​in advance, and includes feature value IDs, feature value identification information, etc.

[0041] It is evident that by determining the initial feature values ​​and dependent feature values, risk control indicators can be transformed into standardized data units that the system can identify and associate, which can then be used to construct the feature value dependency tree.

[0042] S104. Construct a target feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependency relationships.

[0043] In this embodiment of the application, the target feature value dependency tree refers to a tree-like data structure constructed with the initial feature value and n dependent feature values ​​as nodes and n dependency relationships as connection logic. Its core function is to intuitively represent the hierarchical relationship between the initial feature value and each dependent feature value, and to provide a visual and structured basis for determining the order of subsequent indicator calculations.

[0044] In a specific embodiment, the initial feature value is used as the root node of the target feature value dependency tree, and this root node corresponds to the target risk control indicator. Then, according to the dependency relationship of the target risk control indicator, the dependent feature values ​​corresponding to the initial feature value are associated with the root node, and each dependent feature value is connected to the dependency tree as a direct child node of the root node according to its corresponding dependency relationship.

[0045] If a certain dependent feature value has its own lower-level dependencies, that is, the dependent indicator corresponding to the dependent feature value still has other dependent indicators, the system recursively queries the lower-level dependent feature values ​​and dependency relationships corresponding to the dependent feature value, and continues to connect the lower-level dependent feature values ​​as child nodes of the current dependent feature value into the tree, until all dependent feature values ​​are hierarchically associated according to the dependency relationship.

[0046] When n is 0, the target feature value dependency tree contains only the initial feature value as the root node, and has no child nodes.

[0047] It is evident that by constructing a target feature value dependency tree, the association logic between the initial feature value and the dependent feature value is presented in a structured form, clearly defining the order of indicator calculation and ensuring the orderliness and accuracy of the indicator calculation process.

[0048] Optionally, the above step of constructing a target feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependency relationships specifically includes the following steps: A401. Construct an initial feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependency relationships; A402. Perform cyclic dependency detection on the initial feature value dependency tree to obtain the target detection result; A403. If the target detection result indicates the existence of a circular dependency, then determine the circular dependency node based on the target detection result; A404. Correct the initial feature value dependency tree based on the cyclic dependency nodes to obtain the target feature value dependency tree; A405. If the target detection result is that there is no circular dependency, then the initial feature value dependency tree is used as the target feature value dependency tree.

[0049] In this embodiment of the application, the initial feature value dependency tree refers to a tree structure initially constructed based on the initial feature value, n dependent feature values ​​and n dependency relationships. This structure has not yet been verified for cyclic dependency and is only used as the basic carrier for subsequent correction and determination of the target feature value dependency tree.

[0050] Cyclic dependency detection refers to the process of traversing and verifying the logical relationships between feature value nodes in an initial feature value dependency tree to determine whether there are closed-loop relationships such as node A depending on node B, and node B depending on node A. A cyclic dependency node is the core node in the set of feature value nodes that form a closed-loop relationship in the initial feature value dependency tree, leading to the cyclic dependency problem.

[0051] In a specific embodiment, the initial feature value is used as the root node of the tree. According to the n dependencies, the corresponding n dependent feature values ​​are respectively connected as child nodes under the root node. If a certain dependent feature value has a lower-level dependent feature value, the lower-level dependent feature value is recursively connected as the child node of the corresponding dependent feature value to continue to be connected in the tree, thereby initially constructing the initial feature value dependency tree.

[0052] Cyclic dependency detection is performed on the initial feature value dependency tree. All nodes of the initial feature value dependency tree and the dependency links between nodes are traversed. By verifying the dependency source and dependency pointer of each node, it is determined whether there is a closed loop of mutual dependency between nodes in the tree, and the target detection result is obtained. The target detection result includes one of the following: cyclic dependency exists, or cyclic dependency does not exist.

[0053] Please see Figure 2 , Figure 2This is a schematic diagram of a feature value dependency tree without circular dependencies provided in an embodiment of this application. As shown in the figure, each circular node represents a feature value, and the direction of the arrows represents the dependency relationship between feature values ​​(the direction from which the arrow originates is the dependent party, and the direction the arrow points is the dependent party). Feature value A is the root node of the feature value dependency tree, directly dependent on feature values ​​B and C, and feature value C further depends on feature value D. This dependency tree has a hierarchical tree structure, with feature values ​​B and D as leaf nodes. They have no lower-level dependencies, and the dependency links between feature values ​​are not closed loops. There is no situation where the dependency relationship of a feature value ultimately points to itself.

[0054] Please see Figure 3 , Figure 3 This is a schematic diagram of a feature value dependency tree with circular dependencies provided in an embodiment of this application. As shown in the figure, feature value A is the root node of the feature value dependency tree, which directly depends on feature value B and feature value C. Feature value C depends on feature value D and feature value E, while feature value E in turn depends on feature value A. This constitutes a circular dependency chain of ACEA, making the feature value dependency tree have a clear dependency loop.

[0055] If the target detection result indicates the existence of a circular dependency, the circular dependency nodes that form a closed loop are located by tracing back the dependency links recorded during the circular dependency detection process. Then, the association relationship of the circular dependency nodes in the initial feature value dependency tree is adjusted according to the preset correction strategy (such as dynamically disassembling weak dependency links and injecting virtual feature values ​​to replace the closed loop association). After eliminating the dependency loop, the target feature value dependency tree is obtained.

[0056] If the target detection result shows that there is no circular dependency, then there is no need to adjust the initial feature value dependency tree; it can be directly determined as the target feature value dependency tree.

[0057] As can be seen, by initially constructing the initial feature value dependency tree, performing circular dependency detection, and then correcting it as needed, it is ensured that the final generated target feature value dependency tree has no circular dependency problem, effectively avoiding the calculation dead loop caused by circular dependency, and ensuring the stability and accuracy of risk control indicator calculation.

[0058] Optionally, the above step of performing cyclic dependency detection on the initial feature value dependency tree to obtain the target detection result specifically includes the following steps: B401. Traverse each node in the initial feature value dependency tree, determine the ancestor node list and child node list of each node, and obtain n ancestor node lists and n child node lists. B402. Determine the target detection result based on the list of n ancestor nodes and the list of n child nodes.

[0059] In this embodiment, the ancestor node list refers to the set of all ancestral nodes that can be traced upwards along the dependency chain of a given node in the initial feature value dependency tree. The child node list refers to the set of all subordinate nodes that can be traced downwards along the dependency chain of a given node in the initial feature value dependency tree.

[0060] In a specific embodiment, starting from the root node of the tree (corresponding to the initial feature value), each feature value node in the tree is processed sequentially. For each node, the system traces upwards along the hierarchy of the dependency tree, collecting all parent nodes that provide prerequisite dependencies for that node, and integrating them to form a list of ancestor nodes specific to that node; simultaneously, it searches downwards along the hierarchy of the dependency tree, collecting all child nodes that depend on the calculation results of that node, and integrating them to form a list of child nodes specific to that node. As the traversal operation covers all nodes in the initial feature value dependency tree, a list of n ancestor nodes and a list of n child nodes, matching the number of nodes, are finally obtained.

[0061] For example, in a certain initial feature value dependency tree, the root node A depends on nodes B and C, and node B depends on node D. Then the ancestor node list of the root node A is empty, and the child node list is [B,C,D]. The ancestor node list of node C is [A], and the child node list is [D].

[0062] Next, cross-validation is performed on the list of n ancestor nodes and the list of n child nodes to determine the target detection result. That is, it is to check whether a closed-loop dependency link is formed in which the current node depends on the child node and the child node depends on the current node. If a closed-loop dependency link is formed, the target detection result is determined to be that there is a circular dependency; otherwise, the target detection result is determined to be that there is no circular dependency.

[0063] Please see Figure 4 , Figure 4This is a schematic diagram of a circular dependency detection process provided in an embodiment of this application. As shown in the figure, the process begins by obtaining any node in the circular dependency chain to obtain the node to be verified. Next, it is determined whether the node depends on itself. If the result is yes, the verification fails, indicating a circular dependency exists. If the result is no, the operation of obtaining the node's ancestor node list is performed, collecting all the node's parent nodes. Then, it is determined whether the ancestor node list is empty. If yes, it means the ancestor node list is empty, the node has no parent dependencies, and the verification passes. If no, it means the ancestor node list is not empty, and the operation of obtaining the node's child node list is performed, collecting all the node's child nodes that are depended upon. Finally, it is determined whether the child node list contains nodes from the ancestor node list. If yes, it means the child node list contains nodes from the ancestor node list, the child node and ancestor node form a closed loop, and the verification fails. If no, it means the child node list does not contain nodes from the ancestor node list, and the verification passes. The process then ends.

[0064] It is evident that by traversing and obtaining the list of ancestor nodes and child nodes of a node and performing cross-validation, the circular dependencies in the initial feature value dependency tree can be identified comprehensively and accurately.

[0065] Optionally, the above step of determining the target detection result based on the list of n ancestor nodes and the list of n child nodes specifically includes the following steps: C401. Obtain the target ancestor node list and the target child node list; the target ancestor node list is the list of ancestor nodes corresponding to the target node in the n ancestor node list; the target child node list is the list of child nodes corresponding to the target node in the n child node list; C402. If any child node in the target child node list exists in the target ancestor node list, then the target detection result is determined to have a circular dependency. C403. If any child node in the target child node list does not exist in the target ancestor node list, then the target detection result is determined to be that there is no circular dependency.

[0066] In a specific embodiment, from the obtained list of n ancestor nodes and list of n child nodes, the current feature value node to be verified in the initial feature value dependency tree is selected as the target node, and the ancestor node list (i.e., the target ancestor node list) and child node list (i.e., the target child node list) corresponding to the target node are extracted simultaneously.

[0067] Each child node in the target child node list is checked one by one to determine if it exists in the target ancestor node list. If the check finds that any child node in the target child node list is also in the target ancestor node list, it means that the target node and that child node form a closed-loop dependency chain, where the target node depends on the child node and the child node depends on the target node. Therefore, the target detection result can be directly determined as having a circular dependency. If, after checking all child nodes in the target child node list, no child node is found to exist in the target ancestor node list, the target detection result is determined as not having a circular dependency.

[0068] Optionally, the above step of correcting the initial feature value dependency tree based on the cyclic dependency nodes to obtain the target feature value dependency tree specifically includes the following steps: D401. Obtain the circular dependency chain containing the circular dependency node; the circular dependency chain includes m nodes; m is an integer greater than 1; D402. Determine the m dependencies in the circular dependency chain; D403. Determine the association weights corresponding to the m dependencies to obtain m association weights; the association weights are used to characterize the importance of the corresponding dependencies in the business. D404. If there is a correlation weight among the m correlation weights that is lower than a preset weight threshold, then obtain the first correlation weight; the first correlation weight is the correlation weight among the m correlation weights that is lower than the preset weight threshold. D405. Obtain the first dependency relationship corresponding to the first association weight; D406. Remove the first dependency relationship from the initial feature value dependency tree to obtain the target feature value dependency tree; D407. If none of the m associated weights is lower than the preset weight threshold, then obtain the second associated weight; the second associated weight is the associated weight with the smallest value among the m associated weights. D408. Obtain the second dependency relationship corresponding to the second association weight; D409. Determine the dependent nodes of the second dependency relationship; D410. Obtain the historical snapshot data corresponding to the dependent node; the historical snapshot data is the valid historical snapshot result when the dependent node has no circular dependencies in the history. D411. Replace the feature values ​​of the dependent nodes in the second dependency relationship with the historical snapshot data to obtain the third dependency relationship; D412. Update the initial feature value dependency tree according to the third dependency relationship to obtain the target feature value dependency tree.

[0069] In a specific embodiment, a circular dependency chain refers to a closed-loop link formed by the interconnected circular dependency nodes in the initial feature value dependency tree. This link contains m nodes, where m is an integer greater than 1, and the nodes are connected by dependency relationships to form a dependency loop.

[0070] Association weight is a numerical value used to quantify the importance of each dependency in a circular dependency chain in a business scenario. The higher the weight, the greater the impact of the dependency on the risk control business logic. The preset weight threshold is a pre-configured critical value for judging the importance of dependencies. Dependencies below this value can be considered weak dependencies with little impact on the business.

[0071] It should be noted that correlation weight refers to converting the impact of dependencies on risk control operations into calculable and comparable numerical values, such as values ​​in the range of 0.1-1.0, through preset business rules. For example, in an anti-fraud transaction monitoring scenario, the preset rules assign weights of 0.8-1.0 to dependencies directly related to core risk control indicators (such as transaction amount and number of abnormal logins), and 0.1-0.3 to dependencies related to auxiliary information (such as transaction notes and device model descriptions).

[0072] The correlation weight applies to dependencies within a circular chain of dependencies, rather than isolated metrics or features. For example, in the circular chain ABCA, node A depends on node B, where node A represents the fraud risk score of the past hour and node B represents the transaction amount of the past hour. Because the risk score is directly correlated, the weight might be set to 0.9. Node B depends on node C, where node C represents the number of transactions in the past hour. Since the number of transactions is a basis for calculating the transaction amount but not a core influencing factor, the weight might be set to 0.6. Node C depends on node A, where node C relies on node A's risk score to filter invalid transactions. Since this is a non-essential auxiliary filter, the weight might be set to 0.2.

[0073] Dependencies with high weight correspond to critical links in risk control operations. Removing them would directly cause core indicator calculations to fail. For example, removing a high-weight dependency on transaction amount for risk scoring would prevent risk scoring from being calculated. Therefore, this embodiment prioritizes removing low-weight dependencies below a preset weight threshold. Essentially, this weighted filtering eliminates circular dependencies while preserving the dependency logic that plays a decisive role in the business to the greatest extent possible.

[0074] When there are no low-weight dependencies to remove in the circular chain, meaning the association weights of all dependencies are higher than a preset weight threshold, this embodiment removes the dependency with the lowest weight and replaces its node feature value with historical snapshot data. The lowest association weight means the least business impact; even with historical data replacement, interference with overall risk control decisions (such as alarm triggering and risk rating) is minimized, ensuring business continuity.

[0075] In a specific embodiment, the identified circular dependency nodes are located from the initial feature value dependency tree, the dependency association links between these nodes are traced, a circular dependency relationship chain consisting of m nodes is determined, and the dependency relationship between the m nodes in the circular dependency relationship chain is determined, resulting in m dependency relationships.

[0076] Based on preset business rules, such as the degree of influence of dependency relationships on risk control decision results and the strength of correlation between node data, a corresponding association weight is assigned to each dependency relationship in the circular dependency chain, resulting in m association weights.

[0077] If any of the m associated weights is below a preset weight threshold, then these weights are selected as the first associated weights. The first dependency relationship corresponding to the first associated weight is found, and this first dependency relationship is removed from the initial feature value dependency tree to eliminate dependency loops, thus obtaining the target feature value dependency tree. It should be noted that after removing the dependency relationship, there is no dependency relationship between the dependent node and the dependent node; therefore, the dependent node is retained, while the dependent node is removed.

[0078] If none of the m associated weights is below a preset weight threshold, then the associated weight with the smallest value is selected from the m associated weights as the second associated weight. The second dependency relationship corresponding to the second associated weight is determined, the dependent node in the second dependency relationship is found, and the valid historical snapshot data of the dependent node when there was no circular dependency in the historical operation is retrieved. The feature value of the dependent node in the second dependency relationship is replaced with the historical snapshot data to form a third dependency relationship without the risk of closed loop. The initial feature value dependency tree is then updated with the third dependency relationship to eliminate the dependency loop and finally obtain the target feature value dependency tree.

[0079] Please see Figure 5 , Figure 5 This is a schematic diagram of node update of a feature value dependency tree provided in an embodiment of this application. As shown in the figure, when the dependency relationship of the resource (such as an indicator) corresponding to the feature value changes due to operations such as adding or deleting dependencies, it is necessary to adjust the nodes and dependency links of the feature value dependency tree.

[0080] The left figure shows the original eigenvalue dependency tree, where eigenvalue A depends on eigenvalues ​​B and C, and eigenvalue D depends on eigenvalue A.

[0081] The middle plot shows that feature value A depends on feature value B, which needs to be removed, the original feature value C, and the newly added feature value E.

[0082] The right figure shows the feature value dependency tree after dependency relationship reconstruction. The dependency link of feature value D, which originally depended on feature value A, is updated to depend on the new feature value A'. At the same time, A' and feature value C form a new dependency link. The dependency of feature value E is also adjusted accordingly. The dependency link (dashed line) between the original feature values ​​A and D is updated, and finally the node update and dependency relationship reconstruction of the feature value dependency tree are completed.

[0083] Therefore, when indicator A (corresponding to feature value A) is updated, a new indicator record A' (corresponding to feature value A') is actually added. This may introduce a new indicator E (corresponding to feature value E) or adjust the dependency on indicator B (corresponding to feature value B). At this time, it is necessary to query the resources that depend on indicator A (such as indicator D, corresponding to feature value D), adjust their dependency from the original indicator A to the new indicator A', and reconstruct the dependency chain of A' (such as dependencies on C and E), thereby completing the node update and dependency reconstruction of the entire feature value dependency tree.

[0084] It is evident that the hierarchical correction strategy of prioritizing the removal of weak dependencies and replacing them with historical data when there are no weak dependencies can quickly eliminate circular dependencies while ensuring that the core business logic is not affected. It can also maintain business continuity by using historical snapshot data when there are no weak dependencies to remove, ensuring that the corrected target feature value dependency tree has no risk of circular dependencies and improving the accuracy of subsequent indicator calculations.

[0085] S105. Obtain the risk control event data corresponding to the target risk control scenario.

[0086] In this embodiment of the application, risk control event data refers to the actual business data that triggers the calculation of risk control indicators under the target risk control scenario. It directly reflects the business behavior or status under the scenario and is the core input basis for subsequent execution of the target expression and calculation of the target risk control indicators. For example, in the anti-fraud transaction monitoring scenario, risk control event data is the specific transaction amount, transaction time, and transaction account identifier. In the tax risk identification scenario, risk control event data is the specific enterprise declaration amount, invoice issuance data, etc.

[0087] In a specific embodiment, based on the business attributes of the target risk control scenario (such as real-time transaction monitoring, batch tax scanning, etc.), the type of business event that triggers the risk control calculation of the scenario (such as transaction occurrence, declaration submission, etc.) is determined, and business data pushed in real time by the business system is received to obtain risk control event data.

[0088] S106. Determine the target indicator value of the target risk control indicator based on the risk control event data, the target expression, and the target feature value dependency tree.

[0089] In this embodiment of the application, the target indicator value refers to the specific quantitative result obtained by calculating the target risk control indicator through an expression. This result directly reflects the degree of risk or business behavior characteristics under the target risk control scenario.

[0090] In a specific embodiment, the constructed target feature value dependency tree is first loaded, and the calculation order of the indicators is determined according to the hierarchical structure of the dependency tree. Starting from the dependency feature value at the bottom of the dependency tree, the calculation is performed layer by layer upwards to the initial feature value at the top (corresponding to the target risk control indicator).

[0091] Next, the expression execution engine (such as AviatorScript) is invoked to load the target expression corresponding to the target risk control indicator. Preprocessed risk control event data (such as standardized data like transaction amount and time) is encapsulated into a format recognizable by the engine and passed in. The execution engine recursively calculates the indicator result corresponding to each feature value node according to the calculation order of the target feature value dependency tree. First, it calculates the dependency indicator result corresponding to the underlying dependency feature value based on the risk control event data. Then, it uses the underlying calculation result as input, substituting it into the expression corresponding to the upper-level feature value for calculation, until the calculation of the target risk control indicator corresponding to the top-level initial feature value is completed, ultimately generating the target indicator value for that target risk control indicator. If a custom function, such as `sum` or `if`, needs to be called during the calculation process, the engine will automatically retrieve the corresponding function from the preset function library to complete the calculation.

[0092] It is evident that by following the hierarchical order of the target feature value dependency tree and combining the expression execution engine with risk control event data to carry out recursive calculations, the calculation process of the target indicator value is ensured to be orderly, accurate, and fully aligned with the actual business data of the target risk control scenario.

[0093] Optionally, the above step of determining the target indicator value of the target risk control indicator based on the risk control event data, the target expression, and the target feature value dependency tree specifically includes the following steps: A601. Based on the risk control event data, recursively calculate the target feature value dependency tree to determine the index value corresponding to each feature value in the target feature value dependency tree, and obtain the index value set. A602. The target indicator value is determined by the preset expression execution engine based on the target expression and the indicator value set.

[0094] In this embodiment of the application, the index value set refers to the set that integrates all index calculation results after the index corresponding to all feature values ​​in the target feature value dependency tree has been calculated. This set covers the quantitative data corresponding to the target feature value dependency tree from the bottom dependent feature value to the top initial feature value.

[0095] The preset expression execution engine refers to a pre-configured and integrated dedicated expression calculation module. In this embodiment, it is specifically the AviatorScript expression executor, which has the ability to compile expressions, call custom functions, and process data, thus ensuring the efficiency and accuracy of indicator calculation.

[0096] In a specific embodiment, the hierarchical structure of the target feature value dependency tree is used as the basis for the calculation order, and recursive calculation is initiated from the dependent feature value at the bottom level of the dependency tree. For bottom-level feature values ​​without lower-level dependencies, their corresponding indicator values ​​can be directly determined by combining risk control event data. For feature values ​​with lower-level dependencies, the indicator values ​​corresponding to all their lower-level feature values ​​can be calculated first, and then the lower-level indicator values ​​can be used as input to calculate the indicator value corresponding to the current feature value by combining risk control event data.

[0097] The recursive calculation completes the calculation of the indicator values ​​corresponding to all feature values ​​in the target feature value dependency tree, and then integrates all indicator values ​​to form an indicator value set. Next, a preset expression execution engine is called. First, the target expression is passed to the engine for compilation, generating executable calculation logic. Then, the relevant indicator values ​​required by the target expression are extracted from the indicator value set as input parameters and passed to the compiled calculation logic. If the target expression involves a custom function, the engine will automatically call the corresponding function from the preset function library to complete the calculation. Finally, the target indicator value of the target risk control indicator is obtained through the engine's execution operation.

[0098] Please see Figure 6 , Figure 6 This is an application scenario diagram of a risk control indicator calculation method provided in this application embodiment. The system includes: a tax declaration system (providing real-time tax declaration data for enterprises); an invoice database (storing all enterprise invoice data); a business registration information database (storing enterprise registration, changes, and other information); a resource manager (a core module responsible for integrating multi-source data (tax, invoice, and business registration information) and standardizing it); a feature value management module (assigning feature values ​​to risk control indicators and managing feature value relationships); a dynamic dependency tree construction module (building and updating feature value dependency trees based on feature value dependencies); an expression execution engine (a dedicated calculation module for parsing target expressions and calculating target indicator values); real-time streaming (data accessing the resource manager in real-time); batch import (importing batch data from the invoice database into the resource manager at once); and API calls (interface calls by the resource manager to retrieve data from the business registration information database).

[0099] As shown in the diagram, the tax declaration system transmits real-time enterprise tax declaration data to the resource manager via real-time streaming, while the invoice database transmits enterprise invoice data to the resource manager via batch import. The business registration information database transmits enterprise business registration information data to the resource manager via API calls. The resource manager integrates and standardizes the multi-source data before outputting it to the feature value management module. This module assigns feature values ​​to each risk control indicator and establishes dependencies, then passes the relevant information to the dynamic dependency tree construction module. The dynamic dependency tree construction module builds a dynamic feature value dependency tree based on the feature value dependencies and then passes it to the expression execution engine. The expression execution engine combines the risk control event data processed by the resource manager, the target expression, and the dynamic feature value dependency tree to calculate the risk control indicators, ultimately outputting the target indicator values ​​to support tax risk decision-making.

[0100] As can be seen, by first recursively calculating to obtain a complete set of index values, the comprehensiveness and accuracy of the data required for the execution of the target expression are ensured. Then, the target expression is processed by the execution engine according to the preset expression, ensuring the efficiency and standardization of the calculation process.

[0101] S107. Determine the target risk control identification result based on the target indicator value and the target rule, and display the target risk control identification result through a preset visualization system.

[0102] In this embodiment of the application, the target risk control identification result refers to the conclusion generated after matching and judging the target indicator value and target rule based on the target risk control indicator. This conclusion may include the risk level (such as low risk, medium risk, high risk), whether an early warning is triggered, and corresponding handling suggestions (such as normal release, account freezing, manual review).

[0103] The pre-defined visualization system refers to the interactive display system pre-built in this embodiment, which adopts a front-end and back-end separation architecture. The front-end can be built with a web interface based on Vue or React, while the back-end provides API support through Spring Boot. The visualization system has functions such as data visualization, dependency graph display, and real-time result push, used to intuitively present risk control-related information to business personnel.

[0104] In a specific embodiment, the target indicator value is compared with the preset judgment conditions in the target rules. Based on the comparison results, the risk status corresponding to the business behavior, whether an early warning is triggered, and the handling suggestions are determined, and integrated to form the target risk control identification result. For example, if the target indicator value is >10000, it is judged as high risk and an alarm is triggered; if the target indicator value is between 5000 and 10000, it is judged as medium risk and a manual review is prompted.

[0105] Next, the target risk control identification results are transmitted to a pre-set visualization system. The visualization system performs data format conversion and visualization rendering of the results, presenting risk levels using color labels (e.g., green for low risk, yellow for medium risk, and red for high risk). It also displays the triggered rule conditions and their corresponding target indicator values, and can associate them with auxiliary information such as a target feature value dependency tree and key fields of risk control event data for intuitive viewing by business personnel. Simultaneously, if the target risk control identification result is of a high-risk type requiring urgent attention, the visualization system proactively alerts business personnel via pop-up notifications and push notifications to ensure timely delivery of risk information.

[0106] As can be seen, by accurately matching target indicator values ​​with target rules, target risk control identification results are generated, ensuring the standardization and accuracy of risk assessment. A pre-set visualization system intuitively displays the identification results and related information, reducing the understanding cost of risk control data for business personnel, facilitating rapid response to risk events, and simultaneously achieving transparent presentation of risk control results, providing a clear basis for subsequent audit retrospectives.

[0107] In summary, by implementing the embodiments of this application, the following steps are taken: First, target risk control indicators corresponding to the target risk control scenario, as well as target expressions and target rules corresponding to the target risk control indicators; second, dependent indicators in the dependency relationships of the target risk control indicators are identified based on the target expressions, resulting in n dependent indicators, and the dependency relationships corresponding to the n dependent indicators are determined, resulting in n dependency relationships; third, initial feature values ​​corresponding to the target risk control indicators are determined; fourth, n dependent feature values ​​corresponding to the n dependent indicators are determined; fifth, a target feature value dependency tree is constructed based on the initial feature values, the n dependent feature values, and the n dependency relationships; sixth, risk control event data corresponding to the target risk control scenario is obtained; seventh, the target indicator value of the target risk control indicator is determined based on the risk control event data, the target expression, and the target feature value dependency tree; and finally, the target risk control identification result is determined based on the target indicator value and the target rules, and the target risk control identification result is displayed through a preset visualization system. It is evident that by constructing a feature value dependency tree to calculate the risk control indicators, accurate calculation of the risk control indicators is achieved.

[0108] Please see Figure 7 , Figure 7 This is a schematic diagram of the structure of a risk control indicator calculation system provided in an embodiment of this application. The risk control indicator calculation system 700 includes: a risk control indicator determination unit 701, a dependent indicator determination unit 702, a feature value calculation unit 703, a feature value dependency tree construction unit 704, a risk control event data acquisition unit 705, an indicator value calculation unit 706, and a risk control identification result display unit 707. The risk control index determination unit 701 is used to obtain the target risk control index corresponding to the target risk control scenario, as well as the target expression and target rule corresponding to the target risk control index; The dependency indicator determination unit 702 is used to identify the dependency indicators in the dependency relationship of the target risk control indicator according to the target expression, to obtain n dependency indicators, and to determine the dependency relationship corresponding to the n dependency indicators, to obtain n dependency relationships; n is a natural number; The feature value calculation unit 703 is used to determine the initial feature value corresponding to the target risk control indicator; and to determine the n dependent feature values ​​corresponding to the n dependent indicators; The feature value dependency tree construction unit 704 is used to construct a target feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependency relationships; The risk control event data acquisition unit 705 is used to acquire risk control event data corresponding to the target risk control scenario; The indicator value calculation unit 706 is used to determine the target indicator value of the target risk control indicator based on the risk control event data, the target expression, and the target feature value dependency tree; The risk control identification result display unit 707 is used to determine the target risk control identification result based on the target indicator value and the target rule, and to display the target risk control identification result through a preset visualization system.

[0109] Optionally, in obtaining the target risk control indicators corresponding to the target risk control scenario, as well as the target expressions and target rules corresponding to the target risk control indicators, the risk control indicator determination unit 701 is further specifically used for: The target risk control indicator bound to the target risk control scenario, as well as the target expression and target rule corresponding to the target risk control indicator, are obtained from the preset resource management table. The resource management table includes the binding relationship between scenarios and resources. Each binding relationship includes: scenario ID, resource type, and resource ID. The resource type includes one of the following: indicator, expression, and rule.

[0110] Optionally, in constructing the target feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependencies, the feature value dependency tree construction unit 704 is further specifically used for: Construct an initial feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependencies; Cyclic dependency detection is performed on the initial feature value dependency tree to obtain the target detection result; If the target detection result indicates the existence of a circular dependency, then the circular dependency node is determined based on the target detection result; The initial feature value dependency tree is corrected based on the cyclic dependency nodes to obtain the target feature value dependency tree; If the target detection result indicates that there is no circular dependency, then the initial feature value dependency tree is used as the target feature value dependency tree.

[0111] Optionally, in the step of performing cyclic dependency detection on the initial feature value dependency tree to obtain the target detection result, the feature value dependency tree construction unit 704 is further specifically used for: Traverse each node in the initial feature value dependency tree, determine the ancestor node list and child node list of each node, and obtain n ancestor node lists and n child node lists; The target detection result is determined based on the list of n ancestor nodes and the list of n child nodes.

[0112] Optionally, in determining the target detection result based on the list of n ancestor nodes and the list of n child nodes, the feature value dependency tree construction unit 704 is further specifically used for: Obtain the target ancestor node list and the target child node list; the target ancestor node list is the list of ancestor nodes corresponding to the target node in the n ancestor node list; the target child node list is the list of child nodes corresponding to the target node in the n child node list; If any child node in the target child node list exists in the target ancestor node list, then the target detection result is determined to have a circular dependency. If any child node in the target child node list does not exist in the target ancestor node list, then the target detection result is determined to be that there is no circular dependency.

[0113] Optionally, in the step of correcting the initial feature value dependency tree based on the cyclic dependency nodes to obtain the target feature value dependency tree, the feature value dependency tree construction unit 704 is further specifically used for: Obtain the circular dependency chain containing the circularly dependent node; the circular dependency chain includes m nodes; m is an integer greater than 1; Identify m dependencies in the circular dependency chain; Determine the association weights corresponding to the m dependencies to obtain m association weights; the association weights are used to characterize the importance of the corresponding dependencies in the business. If any of the m associated weights is lower than a preset weight threshold, then a first associated weight is obtained; the first associated weight is the associated weight among the m associated weights that is lower than the preset weight threshold. Obtain the first dependency relationship corresponding to the first association weight; Remove the first dependency from the initial feature value dependency tree to obtain the target feature value dependency tree; If none of the m associated weights is lower than the preset weight threshold, then a second associated weight is obtained; the second associated weight is the smallest of the m associated weights. Obtain the second dependency relationship corresponding to the second association weight; Determine the dependent nodes of the second dependency relationship; Obtain the historical snapshot data corresponding to the dependent node; the historical snapshot data is the valid historical snapshot result when the dependent node has no circular dependencies in the history. The feature values ​​of the dependent nodes in the second dependency relationship are replaced based on the historical snapshot data to obtain the third dependency relationship; The initial feature value dependency tree is updated based on the third dependency relationship to obtain the target feature value dependency tree.

[0114] Optionally, in determining the target indicator value of the target risk control indicator based on the risk control event data, the target expression, and the target feature value dependency tree, the risk control identification result display unit 707 is further specifically used for: Based on the risk control event data, the target feature value dependency tree is recursively calculated to determine the index value corresponding to each feature value in the target feature value dependency tree, thereby obtaining the index value set. The target indicator value is determined by the preset expression execution engine based on the target expression and the indicator value set.

[0115] The risk control indicator calculation system 700 described in this application can acquire target risk control indicators corresponding to a target risk control scenario, as well as target expressions and target rules corresponding to the target risk control indicators; identify dependent indicators in the dependency relationships of the target risk control indicators according to the target expressions, obtaining n dependent indicators, and determine the dependency relationships corresponding to the n dependent indicators, obtaining n dependency relationships; determine the initial feature value corresponding to the target risk control indicator; determine the n dependent feature values ​​corresponding to the n dependent indicators; construct a target feature value dependency tree according to the initial feature value, the n dependent feature values, and the n dependency relationships; acquire risk control event data corresponding to the target risk control scenario; determine the target indicator value of the target risk control indicator according to the risk control event data, the target expression, and the target feature value dependency tree; determine the target risk control identification result according to the target indicator value and the target rule, and display the target risk control identification result through a preset visualization system. It can be seen that by constructing a feature value dependency tree to calculate risk control indicators, accurate calculation of risk control indicators is achieved.

[0116] Please see Figure 8 , Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include a processor, a memory, a communication interface, and one or more programs. The processor, memory, and communication interface can be interconnected via a bus. The one or more programs are stored in the memory and configured to be executed by the processor. In this embodiment, the programs include instructions for performing the following steps: Obtain the target risk control indicators corresponding to the target risk control scenario, as well as the target expressions and target rules corresponding to the target risk control indicators; Based on the target expression, identify the dependent indicators in the dependency relationship of the target risk control indicator to obtain n dependent indicators, and determine the dependency relationship corresponding to the n dependent indicators to obtain n dependency relationships; n is a natural number; Determine the initial feature value corresponding to the target risk control indicator; determine the n dependent feature values ​​corresponding to the n dependent indicators; Construct a target feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependency relationships; Obtain the risk control event data corresponding to the target risk control scenario; The target indicator value of the target risk control indicator is determined based on the risk control event data, the target expression, and the target feature value dependency tree. The target risk control identification result is determined based on the target indicator value and the target rule, and the target risk control identification result is displayed through a preset visualization system.

[0117] The electronic device described in this application can acquire target risk control indicators corresponding to a target risk control scenario, as well as target expressions and target rules corresponding to the target risk control indicators; identify dependent indicators in the dependency relationships of the target risk control indicators according to the target expressions, obtaining n dependent indicators, and determine the dependency relationships corresponding to the n dependent indicators, obtaining n dependency relationships; determine the initial feature value corresponding to the target risk control indicator; determine the n dependent feature values ​​corresponding to the n dependent indicators; construct a target feature value dependency tree according to the initial feature value, the n dependent feature values, and the n dependency relationships; acquire risk control event data corresponding to the target risk control scenario; determine the target indicator value of the target risk control indicator according to the risk control event data, the target expression, and the target feature value dependency tree; determine the target risk control identification result according to the target indicator value and the target rule, and display the target risk control identification result through a preset visualization system. It can be seen that by constructing a feature value dependency tree to calculate the risk control indicator, accurate calculation of the risk control indicator is achieved.

[0118] This application also provides a computer-readable storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the methods described in the above method embodiments, wherein the computer includes an electronic device.

[0119] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of any of the methods described in the above method embodiments. The computer program product may be a software installation package, and the computer may include an electronic device.

[0120] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

[0121] The steps of the methods or algorithms described in the embodiments of this application can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in RAM, flash memory, ROM, EPROM, electrically erasable programmable read-only memory (EEPROM), registers, hard disk, portable hard disk, read-only optical disk (CD-ROM), or any other form of storage medium well known in the art. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and storage medium can reside in an ASIC. Furthermore, the ASIC can reside in a terminal device or management device. Alternatively, the processor and storage medium can exist as discrete components in the terminal device or management device.

[0122] Those skilled in the art will recognize that, in one or more of the examples above, the functions described in the embodiments of this application can be implemented, in whole or in part, by software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., digital video discs (DVDs)), or semiconductor media (e.g., solid-state disks (SSDs)).

[0123] The modules / units included in the various devices and products described in the above embodiments can be software modules / units, hardware modules / units, or a combination of both. For example, for devices and products applied to or integrated into a chip, all modules / units can be implemented using hardware methods such as circuits, or at least some modules / units can be implemented using software programs that run on a processor integrated within the chip, while the remaining (if any) modules / units can be implemented using hardware methods such as circuits. For devices and products applied to or integrated into a chip module, all modules / units can be implemented using hardware methods such as circuits. Different modules / units can be located in the same component (e.g., chip, circuit module, etc.) or different components of the chip module, or at least some modules / units can be implemented using hardware methods such as circuits. The implementation is achieved through a software program that runs on the processor integrated within the chip module. The remaining modules / units (if any) can be implemented using hardware methods such as circuits. For various devices and products applied to or integrated into terminal equipment, each of their modules / units can be implemented using hardware methods such as circuits. Different modules / units can be located in the same component (e.g., chip, circuit module, etc.) or different components within the terminal equipment. Alternatively, at least some modules / units can be implemented through a software program that runs on the processor integrated within the terminal equipment, while the remaining modules / units (if any) can be implemented using hardware methods such as circuits.

[0124] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the embodiments of this application. It should be understood that the above descriptions are merely specific embodiments of the embodiments of this application and are not intended to limit the protection scope of the embodiments of this application. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solutions of the embodiments of this application should be included within the protection scope of the embodiments of this application.

Claims

1. A method for calculating risk control indicators, characterized in that, The method includes: Obtain the target risk control indicators corresponding to the target risk control scenario, as well as the target expressions and target rules corresponding to the target risk control indicators; Based on the target expression, identify the dependent indicators in the dependency relationship of the target risk control indicator to obtain n dependent indicators, and determine the dependency relationship corresponding to the n dependent indicators to obtain n dependency relationships; n is a natural number; Determine the initial feature value corresponding to the target risk control indicator; determine the n dependent feature values ​​corresponding to the n dependent indicators; Construct a target feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependency relationships; Obtain the risk control event data corresponding to the target risk control scenario; The target indicator value of the target risk control indicator is determined based on the risk control event data, the target expression, and the target feature value dependency tree. The target risk control identification result is determined based on the target indicator value and the target rule, and the target risk control identification result is displayed through a preset visualization system.

2. The method as described in claim 1, characterized in that, The acquisition of the target risk control indicators corresponding to the target risk control scenario, as well as the target expressions and target rules corresponding to the target risk control indicators, includes: The target risk control indicator bound to the target risk control scenario, as well as the target expression and target rule corresponding to the target risk control indicator, are obtained from the preset resource management table. The resource management table includes the binding relationship between scenarios and resources. Each binding relationship includes: scenario ID, resource type, and resource ID. The resource type includes one of the following: indicator, expression, and rule.

3. The method as described in claim 1, characterized in that, The step of constructing a target feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependency relationships includes: Construct an initial feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependencies; Cyclic dependency detection is performed on the initial feature value dependency tree to obtain the target detection result; If the target detection result indicates the existence of a circular dependency, then the circular dependency node is determined based on the target detection result; The initial feature value dependency tree is corrected based on the cyclic dependency nodes to obtain the target feature value dependency tree; If the target detection result indicates that there is no circular dependency, then the initial feature value dependency tree is used as the target feature value dependency tree.

4. The method as described in claim 3, characterized in that, The step of performing cyclic dependency detection on the initial feature value dependency tree to obtain the target detection result includes: Traverse each node in the initial feature value dependency tree, determine the ancestor node list and child node list of each node, and obtain n ancestor node lists and n child node lists; The target detection result is determined based on the list of n ancestor nodes and the list of n child nodes.

5. The method as described in claim 4, characterized in that, The step of determining the target detection result based on the list of n ancestor nodes and the list of n child nodes includes: Obtain the target ancestor node list and the target child node list; the target ancestor node list is the list of ancestor nodes corresponding to the target node in the n ancestor node list; the target child node list is the list of child nodes corresponding to the target node in the n child node list; If any child node in the target child node list exists in the target ancestor node list, then the target detection result is determined to have a circular dependency. If any child node in the target child node list does not exist in the target ancestor node list, then the target detection result is determined to be that there is no circular dependency.

6. The method as described in claim 5, characterized in that, The step of correcting the initial feature value dependency tree based on the cyclic dependency nodes to obtain the target feature value dependency tree includes: Obtain the circular dependency chain containing the circularly dependent node; the circular dependency chain includes m nodes; m is an integer greater than 1; Identify m dependencies in the circular dependency chain; Determine the association weights corresponding to the m dependencies to obtain m association weights; the association weights are used to characterize the importance of the corresponding dependencies in the business. If any of the m associated weights is lower than a preset weight threshold, then a first associated weight is obtained; the first associated weight is the associated weight among the m associated weights that is lower than the preset weight threshold. Obtain the first dependency relationship corresponding to the first association weight; Remove the first dependency from the initial feature value dependency tree to obtain the target feature value dependency tree; If none of the m associated weights is lower than the preset weight threshold, then a second associated weight is obtained; the second associated weight is the smallest of the m associated weights. Obtain the second dependency relationship corresponding to the second association weight; Determine the dependent nodes of the second dependency relationship; Obtain the historical snapshot data corresponding to the dependent node; the historical snapshot data is the valid historical snapshot result when the dependent node has no circular dependencies in the history. The feature values ​​of the dependent nodes in the second dependency relationship are replaced based on the historical snapshot data to obtain the third dependency relationship; The initial feature value dependency tree is updated based on the third dependency relationship to obtain the target feature value dependency tree.

7. The method according to any one of claims 1-6, characterized in that, The step of determining the target indicator value of the target risk control indicator based on the risk control event data, the target expression, and the target feature value dependency tree includes: Based on the risk control event data, the target feature value dependency tree is recursively calculated to determine the index value corresponding to each feature value in the target feature value dependency tree, thereby obtaining the index value set. The target indicator value is determined by the preset expression execution engine based on the target expression and the indicator value set.

8. A risk control indicator calculation system, characterized in that, The risk control indicator calculation system includes: a risk control indicator determination unit, a dependent indicator determination unit, a feature value calculation unit, a feature value dependency tree construction unit, a risk control event data acquisition unit, an indicator value calculation unit, and a risk control identification result display unit. The risk control indicator determination unit is used to obtain the target risk control indicator corresponding to the target risk control scenario, as well as the target expression and target rule corresponding to the target risk control indicator; The dependency indicator determination unit is used to identify the dependency indicators in the dependency relationship of the target risk control indicator according to the target expression, to obtain n dependency indicators, and to determine the dependency relationship corresponding to the n dependency indicators, to obtain n dependency relationships; n is a natural number; The feature value calculation unit is used to determine the initial feature value corresponding to the target risk control indicator; and to determine the n dependent feature values ​​corresponding to the n dependent indicators; The feature value dependency tree construction unit is used to construct a target feature value dependency tree based on the initial feature value, the n dependent feature values, and the n dependency relationships; The risk control event data acquisition unit is used to acquire risk control event data corresponding to the target risk control scenario; The indicator value calculation unit is used to determine the target indicator value of the target risk control indicator based on the risk control event data, the target expression, and the target feature value dependency tree. The risk control identification result display unit is used to determine the target risk control identification result based on the target indicator value and the target rule, and to display the target risk control identification result through a preset visualization system.

9. An electronic device, characterized in that, include: Processor, memory, communication interface, and one or more programs; The one or more programs are stored in the memory and configured to be executed by the processor, the programs including instructions for performing the steps of the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method as described in any one of claims 1-7.

Images