A method for implementing risk control rule configuration based on rule engine
By generating risk control model templates on the risk control platform and combining Drools rule engine and Velocity template engine, the configuration of risk control rules is achieved, and the problems of complexity of risk control rules description files and difficult to understand decision-making processes are solved, and risk control services are improved Define efficiency and system flexibility.
Patent Information
- Application Number
- CN202111317264.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-11-09
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2041-11-09
AI Technical Summary
In the prior art, the syntax of the risk control rule description file is complex and difficult to master, and the Excel-format decision process description file is obscure and difficult to understand, making it difficult for risk control personnel to effectively configure and maintain risk control rules.
By providing a risk control business feature library on the risk control platform, matching the risk control business features and generating risk control model templates, combining Drools rules engine and Velocity template engine, the risk control model and decision-making process are configured, and online canvas drawing of risk control is used to draw Service decision flow chart, generate executable risk control model executors.
It improves the efficiency and accuracy of risk control service definition, reduces the complexity of defining models, increases system flexibility and maintenance convenience, and avoids manual writing of rules and execution logic and complex Excel configurations.
Smart Images

Figure CN113888050B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer information processing technology rule processing, and in particular to a method for implementing risk control rule configuration based on a rule engine. Background Art
[0002] Risk control refers to risk control. Risk control rules are uncertain. With the development of risk control business and the growth of business volume, risk control personnel need to constantly adjust and enrich risk control rules to meet the increasingly complex business risk control logic and provide better risk control services. The risk control business is the actual affairs that various industries need to deal with; the risk control service is the service provided by the risk control platform.
[0003] In the existing public technology, the following method steps are usually adopted:
[0004] Step 1: Write the rule description file directly.
[0005] Step 2: Upload the rule description file.
[0006] Step 3: Parse the rule description file and generate the rule executor.
[0007] Step 4: The rule engine loads the rule executor into the knowledge base.
[0008] Step 5: Write a decision process description file in Excel format.
[0009] Step 6: Upload the decision process description file in Excel format.
[0010] Step 7: Parse the decision process description file in Excel format and generate a decision process executor.
[0011] Step 8: The rule engine loads the decision process executor into the knowledge base.
[0012] Step 9: Call the newly created decision process executor.
[0013] The above steps have the following defects:
[0014] (1) The syntax of the rule description file is complex, which is difficult for risk control personnel to master and prone to errors.
[0015] (2) The decision process description file in Excel format is obscure and difficult to understand and maintain.
[0016] In response to the above-mentioned defects, the patent "A risk control decision engine device" (application number: 202110812040.6) provides a method covering a variable library, a component library, a strategy, a decision laboratory and a decision engine. The rules are configured by using the variables in the variable library, and then the decision strategies in multiple scenarios are configured by using the rules of the component library. However, although this solution can flexibly change and adjust the rules according to actual conditions, it still has the problem of difficulty in abstract rules and cannot solve the difficulty of abstract rules.
[0017] Therefore, a more applicable method of risk control rule configuration is needed. Summary of the invention
[0018] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a method for configuring risk control rules based on a rule engine.
[0019] In order to solve the above technical problems, the present invention provides a method for implementing risk control rule configuration based on a rule engine, which is characterized by comprising the following steps:
[0020] Step 1: In the risk control business feature library provided by the risk control platform, matching the same risk control business features in the current risk control business; the risk control business feature library is a collection of risk control business features;
[0021] Step 2: For each matched risk control business feature, the risk control platform generates a risk control model template by abstracting a risk control business feature through a preset risk control model template generator; the risk control model template is a form for users to set risk control indicators and logical operators;
[0022] Step 3: The risk control platform searches the risk control model template library for a risk control model template with the same risk control business characteristics. If so, the risk control model template abstractly generated in step 2 is updated to the template library; if not, the risk control model template abstractly generated in step 2 is added to the template library;
[0023] Step 4: The risk control platform fills the risk control business rules into the risk control model template to generate new risk control model metadata; the risk control business rules refer to the judgment logic of the risk control business;
[0024] Step 5: Submit the risk control model metadata to the server through the risk control platform;
[0025] Step 6: The risk control model metadata parsing program preset on the server generates a risk control model executor based on the risk control model metadata; the risk control model executor is used to execute the risk control model and output the execution result of the risk control model; the risk control model is a program package of the risk control operation logic.
[0026] Step 7: Repeat steps 1 to 6 above to generate a risk control model executor in the full-link risk control process of the risk control business;
[0027] Step 8: Analyze the process of risk control business, summarize the decision-making process of risk control services, and draw a risk control service decision-making flow chart on the risk control platform based on the decision-making process of risk control services;
[0028] Step 9: Submit the risk control service decision flowchart metadata to the server through the risk control platform;
[0029] Step 10: The risk control service decision process diagram metadata parsing program preset on the server generates a risk control service decision process executor according to the risk control service decision process diagram metadata, and the risk control service decision process executor is used to execute the risk control service decision process, control the execution steps of the risk control service decision process, and output the execution result of the risk control service decision process;
[0030] Step 11: The business system initiates a risk control service call; the business system is a system owned by each business team outside the risk control platform that needs to use the risk control service;
[0031] Step 12: The risk control service interface searches for the relevant risk control service decision process executor in the knowledge base of the rule engine according to the service code and service version number; if the relevant risk control service decision process executor is not found, the relevant risk control service decision process diagram metadata is searched from the database, and the risk control service decision process executor is generated through the risk control service decision process diagram metadata and added to the knowledge base of the rule engine; the risk control service interface is the total entry for the risk control platform to provide services, the service code is the unique identification number of the risk control service, and the service version number is used to identify different versions of the service. Only one version of the same service can be enabled at the same time;
[0032] Step 13: Run the risk control service decision process executor, and execute the decision path calculated according to the risk control service decision flow chart until the process ends. The decision path is the path actually executed by the decision process;
[0033] Step 14: Generate the execution results of the risk control service decision process and the risk control service report.
[0034] In step 1, the risk control business feature library includes the following features: anti-fraud business features, credit business features, repayment business features, company qualification business features, login business features, distribution business features, and payment business features.
[0035] In the step 2, the method of abstracting the risk control business features into the risk control model template is: the risk control platform searches for a risk control model template generator that can convert the risk control business features into the risk control model template according to the association between the risk control business features and the risk control model template generator; the risk control model template generator includes: a rule model template generator, a rule set model template generator, a blacklist model template generator, a decision tree model template generator, a scoring card model template generator, an artificial intelligence model template generator, and a third-party model template generator; the association between the business features and the risk control model template generator includes: a login business feature associated rule model template generator, an anti-fraud business feature associated rule set model template generator, a credit business feature associated scoring card model template generator, a distribution business feature associated artificial intelligence template generator, and a payment business feature associated third-party model template generator.
[0036] In step 4, the judgment logic includes: in the login service, the same account cannot be logged in on multiple terminals.
[0037] In step 6, the risk control model executor is the execution file required by the Drools rule engine when running; the risk control model metadata parsing program generates the risk control model executor, including: using the Velocity template engine to run the Velocity template, setting the risk control model metadata into the Velocity template, and generating the execution file required by the Drools rule engine when running.
[0038] In the step 7, a risk control model executor is generated in the full-link risk control process of the risk control business, including: when drawing a risk control service decision flowchart, the risk control model set on the model node generates a risk control model executor.
[0039] In step 8, the risk control service decision flowchart includes: a process start node, a risk control model node, a risk control decision node, and a process end node; the process start node is the starting point of the risk control service decision process, which is used to mark the start of the risk control service decision process; the risk control model node is used to execute the risk control model on the risk control model node when the process flows to such a node; the types of the risk control model nodes include: rule model nodes, rule set model nodes, decision tree model nodes, scorecard model nodes, blacklist model nodes, artificial intelligence model nodes, and third-party model nodes; the decision logic on the risk control decision node will calculate the next node of the current process according to the process data in the current process, which is used to control the flow direction of the process; the process end node is the end point of the risk control service decision process, which is used to mark the end of the risk control service decision process; the method of drawing the risk control service decision flowchart is: on the online canvas of the risk control service decision flowchart provided by the risk control platform, drag and connect each process node, then set the model on the risk control model node, and then describe the process flow logic on the risk control decision node to form a risk control service decision flowchart.
[0040] In step 10, the risk control service decision process executor is an execution file in XML format that complies with the BPMN protocol and is required when the Drools rule engine decision flow module is running; the risk control service decision flowchart metadata parsing program converts the risk control service decision flowchart metadata into an executable file in XML format that complies with the BPMN protocol according to the standard BPMN protocol.
[0041] The step 13 comprises:
[0042] Step 13-1: Run the risk control service decision process executor and execute the process start node;
[0043] Step 13-2: Execute the risk control model node, and search for the relevant risk control model executor in the knowledge base of the rule engine according to the risk control model set on the node; if the relevant risk control model executor is not found, search for the relevant model execution file from the database, generate the model executor through the model execution file, and add it to the knowledge base of the rule engine;
[0044] Step 13-3: Run the risk control model executor and add the running results to the process session; the process session is a cache provided by the Drools rule engine to temporarily store process data in the risk control decision process execution. Each time the risk control decision execution process runs, a new process session will be generated;
[0045] Step 13-4: Execute the risk control decision node and calculate the next risk control model node of the risk control service decision process according to the decision logic;
[0046] Step 13-5: Repeat steps 13-1 to 13-4 until the decision-making process is executed to the end node of the process.
[0047] The beneficial effects achieved by the present invention are:
[0048] (1) The present invention realizes the function of defining the risk control service decision flow by dragging and drawing on the online canvas provided by the risk control platform. By integrating the standard BPMN specification, users can avoid defining the risk control service decision flow by complex Excel configuration, thereby improving the efficiency and accuracy of risk control service definition;
[0049] (2) The present invention realizes the conversion of non-executable risk control model metadata into a risk control model executor, thereby avoiding users from manually writing rule execution logic, reducing the complexity of defining the model, and improving the efficiency of defining the model by integrating the Velocity template engine.
[0050] (3) The present invention realizes the dynamic loading and execution of risk control decision flows and risk control models, avoiding manual adjustment of the loading and execution logic in the system when adding decision processes and models. By integrating the Drools rule engine, the flexibility of the system can be increased and the difficulty of maintenance can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] Figure 1 A simplified flowchart of the method steps of an exemplary embodiment of the present invention;
[0052] Figure 2 A flow chart of a method for configuring risk control rules according to an exemplary embodiment of the present invention;
[0053] Figure 3 A flow chart of a configuration method for a risk control service decision process according to an exemplary embodiment of the present invention;
[0054] Figure 4 A flowchart of a micro-loan risk control service calling process according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION
[0055] The present invention will be further described below in conjunction with the accompanying drawings and exemplary embodiments:
[0056] like Figure 1 The main method steps of the exemplary embodiment of the present invention shown in the flowchart include the following steps:
[0057] Step 1: In the risk control business feature library provided by the risk control platform, matching the same risk control business features in the current risk control business; the risk control business feature library is a collection of risk control business features;
[0058] Step 2: For each matched risk control business feature, the risk control platform generates a risk control model template by abstracting a risk control business feature through a preset risk control model template generator; the risk control model template is a form for users to set risk control indicators and logical operators;
[0059] Step 3: The risk control platform searches the risk control model template library for a risk control model template with the same risk control business characteristics. If so, the risk control model template abstractly generated in step 2 is updated to the template library; if not, the risk control model template abstractly generated in step 2 is added to the template library;
[0060] Step 4: The risk control platform fills the risk control business rules into the risk control model template to generate new risk control model metadata; the risk control business rules refer to the judgment logic of the risk control business;
[0061] Step 5: Submit the risk control model metadata to the server through the risk control platform;
[0062] Step 6: The risk control model metadata parsing program preset on the server generates a risk control model executor based on the risk control model metadata; the risk control model executor is used to execute the risk control model and output the execution result of the risk control model; the risk control model is a program package of the risk control operation logic.
[0063] Step 7: Repeat steps 1 to 6 above to generate a risk control model executor in the full-link risk control process of the risk control business;
[0064] Step 8: Analyze the process of risk control business, summarize the decision-making process of risk control services, and draw a risk control service decision-making flow chart on the risk control platform based on the decision-making process of risk control services;
[0065] Step 9: Submit the risk control service decision flowchart metadata to the server through the risk control platform;
[0066] Step 10: The risk control service decision process diagram metadata parsing program preset on the server generates a risk control service decision process executor according to the risk control service decision process diagram metadata, and the risk control service decision process executor is used to execute the risk control service decision process, control the execution steps of the risk control service decision process, and output the execution result of the risk control service decision process;
[0067] Step 11: The business system initiates a risk control service call; the business system is a system owned by each business team outside the risk control platform that needs to use the risk control service;
[0068] Step 12: The risk control service interface searches for the relevant risk control service decision process executor in the knowledge base of the rule engine according to the service code and service version number; if the relevant risk control service decision process executor is not found, the relevant risk control service decision process diagram metadata is searched from the database, and the risk control service decision process executor is generated through the risk control service decision process diagram metadata and added to the knowledge base of the rule engine; the risk control service interface is the total entry for the risk control platform to provide services, the service code is the unique identification number of the risk control service, and the service version number is used to identify different versions of the service. Only one version of the same service can be enabled at the same time;
[0069] Step 13: Run the risk control service decision process executor, and execute the decision path calculated according to the risk control service decision flow chart until the process ends. The decision path is the path actually executed by the decision process;
[0070] Step 14: Generate the execution results of the risk control service decision process and the risk control service report.
[0071] More specifically, in an exemplary embodiment of the present invention, the following steps are included:
[0072] Step S1: The risk control business personnel analyze the characteristics of the risk control business and summarize the characteristics of the risk control business. The method for the risk control business personnel to analyze the characteristics of the risk control business is: in the risk control business feature library provided by the risk control platform, match the characteristics similar to the current business risk control process. The characteristics of the risk control business are the risk control processes used for risk control judgment. The risk control business feature library is a collection of risk control business features. The features included in the risk control business feature library include: anti-fraud business features, credit business features, repayment business features, company qualification business features, login business features, distribution business features, payment business features, etc.
[0073] Step S2: The risk control platform abstracts each risk control business feature into a risk control model template. The risk control model template is a form that allows users to set risk control indicators and logical operators. The method of abstracting the risk control business feature into a risk control model template is: the risk control platform searches for a risk control model template generator that can convert the business feature into a risk control model template based on the association between the business feature and the risk control model template generator. The risk control model template generators include: rule model template generator, rule set model template generator, blacklist model template generator, decision tree model template generator, score card model template generator, artificial intelligence model template generator, third-party model template generator, etc. The association between the business feature and the risk control model template generator includes: login business feature association rule model template generator, anti-fraud business feature association rule set model template generator, credit business feature association score card model template generator, distribution business feature association artificial intelligence template generator, payment business feature association third-party model template generator, etc.
[0074] Step S3: The risk control platform searches for and matches the same risk control model template in the risk control model template library according to the risk control business characteristics. The matching method between the risk control business characteristics and the risk control model template is to determine whether the risk control business characteristics used to generate the risk control model template are the same as the risk control business characteristics currently being matched. If yes, the risk control model template abstractly generated in step S2 is updated to the template library; if no, the risk control model template abstractly generated in step S2 is added to the template library.
[0075] Step S4: The risk control platform fills the risk control business rules into the risk control model template to generate new risk control model metadata. The risk control business rules refer to the actual judgment logic of the risk control business. For example, the judgment logic includes: in the login business, there is a rule that the same account cannot log in on multiple terminals.
[0076] Step S5: The risk control platform operation and maintenance submits the risk control model metadata to the server through the risk control platform.
[0077] Step S6: The risk control model metadata parsing program on the risk control platform server generates a risk control model executor based on the risk control model metadata. The method by which the risk control model metadata parsing program generates the risk control model executor is: using the Velocity template engine, running the Velocity template, and then setting the risk control model metadata into the Velocity template to generate the execution file required for the Drools rule engine to run. The risk control model executor refers to the execution file required for the Drools rule engine to run.
[0078] Step S7: Repeat the above steps S1 to S6 to generate the risk control model executor required for use in the full-link risk control process of the business. The required risk control model executor refers to the risk control model executor generated by the risk control model set on the model node when drawing the risk control service decision flowchart.
[0079] Step S8: The risk control business personnel analyze the process of the risk control business and summarize the decision-making process of the risk control service. The risk control platform operation and maintenance personnel draw a risk control service decision flow chart on the risk control platform according to the risk control service decision process. The risk control service decision flow chart is a diagram containing the process start node, the risk control model node, the risk control decision node, and the process end node. The process start node is the starting point of the risk control service decision process, marking the start of the risk control service decision process, and has no actual logic. The risk control model node is a node containing a risk control model. When the process flows to such a node, the risk control model on the risk control model node will be executed. The types of the risk control model node include: rule model node, rule set model node, decision tree model node, scorecard model node, blacklist model node, artificial intelligence model node, third-party model node, etc. The risk control decision node is a node used to control the direction of the process flow. The decision logic on the node will calculate the next node of the current process based on the process data in the current process. The process end node is the end point of the risk control service decision process, marking the end of the risk control service decision process, and has no actual logic. The method for drawing the risk control service decision flowchart is: on the risk control service decision flowchart online canvas provided by the risk control platform, drag and connect each process node, then set the model on the risk control model node, and then describe the process flow logic on the risk control decision node, and finally form a risk control service decision flowchart.
[0080] Step S9: The risk control platform operation and maintenance personnel submit the risk control service decision flowchart metadata to the server through the risk control platform.
[0081] Step S10: The risk control service decision flowchart metadata parsing program on the risk control platform server generates a risk control service decision process executor based on the risk control service decision flowchart metadata. The risk control service decision process executor refers to an XML-formatted executable file that complies with the BPMN protocol when the Drools rule engine decision flow module is running. The method by which the risk control service decision flowchart metadata parsing program generates the risk control service decision process executor is to convert the risk control service decision flowchart metadata into an executable file in XML format that complies with the BPMN protocol according to the standard BPMN protocol.
[0082] Step S11: The business system initiates a risk control service call. The business system refers to a system owned by each business team outside the risk control platform that needs to use the risk control service.
[0083] Step S12: The risk control service interface searches for the relevant risk control service decision process executor in the knowledge base of the rule engine according to the service code and service version number. If the relevant risk control service decision process executor is not found, the relevant risk control service decision process diagram metadata is searched from the database, and the risk control service decision process executor is generated through the risk control service decision process diagram metadata and added to the knowledge base of the rule engine. The risk control service interface is the main entrance for the services provided by the risk control platform. All business systems connected to the risk control platform use this entrance to initiate risk control service requests. The service code is the unique identification number of the risk control service. Services with the same service code are the same service. The service version number is used to identify different versions of the service. At the same time, only one version of the same service can be enabled.
[0084] The introduction of service version numbers is to support the functions of service upgrade and service version rollback. The service upgrade means that in the process of business development, the original risk control rules are no longer applicable, and the risk control platform upgrades the original risk control service to generate a new version of the risk control service; after the service upgrade is completed, the traffic on the old version of the service will be routed to the new version. The service version rollback refers to the operation of switching to the old version when there is a problem with the current version of the risk control service during operation; after the service version rollback is successful, the traffic of other versions of the current service will be routed to the newly switched version.
[0085] Step S13: running the risk control service decision process executor, and executing the decision path calculated according to the risk control service decision flow chart until the process ends. The decision path is the path actually executed by the decision process;
[0086] The step S13 comprises:
[0087] Step S13-1: Run the risk control service decision process executor and execute the process start node;
[0088] Step S13-2: Execute the risk control model node, and search for the relevant risk control model executor in the knowledge base of the rule engine according to the risk control model set on the node; if the relevant risk control model executor is not found, search for the relevant model execution file from the database, generate the model executor through the model execution file, and add it to the knowledge base of the rule engine;
[0089] Step S13-3: Run the risk control model executor and add the running results to the process session; the process session is a cache provided by the Drools rule engine to temporarily store process data in the execution of the risk control decision process. Each time the risk control decision execution process is run, a new process session will be generated;
[0090] Step S13-4: Execute the risk control decision node and calculate the next risk control model node of the risk control service decision process according to the decision logic;
[0091] Step S13-5: Repeat steps S13-1 to S13-4 until the decision-making process is executed to the end node of the process.
[0092] Step S14: Generate execution results and risk control service report.
[0093] In the decision-making process of step S13, compared with the prior art, the present invention adds a dedicated risk control decision node, directly calculates the next risk control model node of the current risk control model node on the risk control decision node, and creatively uses the online canvas provided by the risk control platform to draw the risk control service decision flowchart, thereby greatly improving the performance.
[0094] like Figure 2 The flow chart of the risk control model configuration method in the embodiment of the present invention shown in the figure has the following specific steps:
[0095] Step 101: Risk control personnel analyze the micro-loan business and conclude that it is necessary to determine whether a user is likely to commit fraud based on the user's historical behavior and to determine the user's credit limit based on the user's credit rating.
[0096] Step 102: Based on the above characteristics, the risk control platform abstracts the anti-fraud rule model template and the credit rating scorecard model template.
[0097] Step 103: The risk control platform searches for templates related to the anti-fraud rule model in the risk control model template library. If there are related templates in the risk control model template library, a new anti-fraud rule model template is not generated.
[0098] Step 104: The risk control platform searches for a credit rating scorecard model template in the risk control model template library. If there is no relevant template in the risk control model template library, a credit rating scorecard model template is created and added to the risk control model template library.
[0099] Step 105: The risk control platform generates anti-fraud rule model metadata according to the anti-fraud business rules and the anti-fraud rule model template, and submits it to the risk control platform server parsing program.
[0100] Step 106: The risk control platform server-side parsing program generates an anti-fraud rule model executor based on the anti-fraud rule model metadata and adds it to the risk control model knowledge base.
[0101] Step 107: The risk control platform operation and maintenance generates the credit rating scorecard model metadata according to the credit rating business rules and the credit rating scorecard model template, and submits it to the risk control platform server parsing program.
[0102] Step 108: The risk control platform server-side parsing program generates a credit rating scorecard model executor based on the credit rating scorecard model metadata and adds it to the risk control model knowledge base.
[0103] like Figure 3 The flowchart of the configuration method of the risk control service decision process in the embodiment of the present invention shown in the figure has the following specific steps:
[0104] Step 201: Risk control personnel analyze the process of small loan business and summarize the risk control decision-making process of small loan business. In the process of small loan business, users with fraud potential are identified through anti-fraud rule model. If the user has fraud potential, the loan is rejected; if the user has no fraud potential, the credit rating of the user is calculated through the credit rating scorecard model, and the credit limit of the user is determined according to the user's credit rating.
[0105] Step 202: The risk control platform operation and maintenance draws a small loan business risk control decision flow chart online on the risk control platform according to the small loan business risk control decision process, and submits it to the server-side decision flow chart parser for analysis.
[0106] Step 203: The risk control platform server-side decision flowchart parsing program generates an executable micro-loan business risk control decision process executor based on the micro-loan business risk control decision flowchart, and adds it to the risk control decision process knowledge base.
[0107] like Figure 4 This is a flowchart of the small loan risk control service calling process in an embodiment of the present invention, and the specific steps are:
[0108] Step 301: The micro-loan business system initiates a micro-loan risk control service call.
[0109] Step 302: The risk control platform searches for the micro-loan business risk control decision process executor in the risk control decision process knowledge base according to the service code and service version number of the micro-loan risk control service. If no relevant executor is found, the micro-loan business risk control decision process flow chart is searched from the database, and an executor is generated according to the flow chart and added to the risk control decision process knowledge base.
[0110] Step 303: Run the micro-loan business risk control decision process executor to execute the anti-fraud rule model.
[0111] Step 304: The anti-fraud rule model shows that the user does not have any fraud risk, the current node risk control passes, and the process is smoothly transferred to the credit rating scorecard model node.
[0112] Step 305: Execute the credit rating scorecard model to display that the user's credit rating is average. The risk control of the current node is completed, and the process is smoothly transferred to the next node.
[0113] Step 306: The process is transferred to the end node, and the micro-loan user risk control report is output. The user is not likely to commit fraud and the credit rating is average.
[0114] The present invention is mainly used to provide a method for implementing risk control rule configuration based on a rule engine, and its beneficial effects are:
[0115] (1) The present invention realizes the function of defining the risk control service decision flow by dragging and dropping on the online canvas provided by the risk control platform. By integrating the standard BPMN specification, users can avoid defining the risk control service decision flow through complex Excel configuration, thereby improving the efficiency and accuracy of risk control service definition.
[0116] (2) The present invention realizes the conversion of non-executable risk control model metadata into a risk control model executor, thereby avoiding users from manually writing rule execution logic, reducing the complexity of defining the model, and improving the efficiency of defining the model by integrating the Velocity template engine.
[0117] (3) The present invention realizes the dynamic loading and execution of risk control decision flows and risk control models, avoiding manual adjustment of the loading and execution logic in the system when adding decision processes and models. By integrating the Drools rule engine, the flexibility of the system can be increased and the difficulty of maintenance can be reduced.
[0118] The above embodiments do not limit the present invention in any way. Any other improvements and applications made to the above embodiments in an equivalent transformation manner belong to the protection scope of the present invention.
Claims
1. A method for implementing risk control rule configuration based on a rule engine, characterized in that: The steps include: Step 1: In the risk control business feature library provided by the risk control platform, matching the same risk control business features in the current risk control business; the risk control business feature library is a collection of risk control business features; Step 2: For each matched risk control business feature, the risk control platform generates a risk control model template by abstracting a risk control business feature through a preset risk control model template generator; the risk control model template is a form for users to set risk control indicators and logical operators; Step 3: The risk control platform searches the risk control model template library for a risk control model template with the same risk control business characteristics. If so, the risk control model template abstractly generated in step 2 is updated to the template library; if not, the risk control model template abstractly generated in step 2 is added to the template library; Step 4: The risk control platform fills the risk control business rules into the risk control model template to generate new risk control model metadata; the risk control business rules refer to the judgment logic of the risk control business; Step 5: Submit the risk control model metadata to the server through the risk control platform; Step 6: The risk control model metadata parsing program preset by the server generates a risk control model executor according to the risk control model metadata; the risk control model executor is used to execute the risk control model and output the execution result of the risk control model; the risk control model is a program package of the risk control operation logic; Step 7: Repeat steps 1 to 6 above to generate a risk control model executor in the full-link risk control process of the risk control business; Step 8: Analyze the process of risk control business, summarize the decision-making process of risk control services, and draw a risk control service decision-making flow chart on the risk control platform based on the decision-making process of risk control services; Step 9: Submit the risk control service decision flowchart metadata to the server through the risk control platform; Step 10: The risk control service decision process diagram metadata parsing program preset on the server generates a risk control service decision process executor according to the risk control service decision process diagram metadata, and the risk control service decision process executor is used to execute the risk control service decision process, control the execution steps of the risk control service decision process, and output the execution result of the risk control service decision process; Step 11: The business system initiates a risk control service call; the business system is a system owned by each business team outside the risk control platform that needs to use the risk control service; Step 12: The risk control service interface searches for the relevant risk control service decision process executor in the knowledge base of the rule engine according to the service code and service version number; if the relevant risk control service decision process executor is not found, the relevant risk control service decision process diagram metadata is searched from the database, and the risk control service decision process executor is generated through the risk control service decision process diagram metadata and added to the knowledge base of the rule engine; the risk control service interface is the total entry for the risk control platform to provide services, the service code is the unique identification number of the risk control service, and the service version number is used to identify different versions of the service. Only one version of the same service can be enabled at the same time; Step 13: Run the risk control service decision process executor, and execute the decision path calculated according to the risk control service decision flow chart until the process ends. The decision path is the path actually executed by the decision process; Step 14: Generate the execution results of the risk control service decision process and the risk control service report.
2. A method for implementing risk control rule configuration based on a rule engine as claimed in claim 1, characterized in that: In step 1, the risk control business feature library includes the following features: anti-fraud business features, credit business features, repayment business features, company qualification business features, login business features, distribution business features, and payment business features.
3. A method for implementing risk control rule configuration based on a rule engine as claimed in claim 2, characterized in that: In the step 2, the method of abstracting the risk control business features into the risk control model template is: the risk control platform searches for a risk control model template generator that can convert the risk control business features into the risk control model template according to the association between the risk control business features and the risk control model template generator; the risk control model template generator includes: a rule model template generator, a rule set model template generator, a blacklist model template generator, a decision tree model template generator, a scoring card model template generator, an artificial intelligence model template generator, and a third-party model template generator; the association between the business features and the risk control model template generator includes: a login business feature associated rule model template generator, an anti-fraud business feature associated rule set model template generator, a credit business feature associated scoring card model template generator, a distribution business feature associated artificial intelligence template generator, and a payment business feature associated third-party model template generator.
4. A method for implementing risk control rule configuration based on a rule engine as claimed in claim 3, characterized in that: In step 4, the judgment logic includes: in the login service, the same account cannot be logged in on multiple terminals.
5. A method for implementing risk control rule configuration based on a rule engine as claimed in claim 4, characterized in that: In step 6, the risk control model executor is the execution file required by the Drools rule engine when running; the risk control model metadata parsing program generates the risk control model executor, including: using the Velocity template engine to run the Velocity template, setting the risk control model metadata into the Velocity template, and generating the execution file required by the Drools rule engine when running.
6. A method for implementing risk control rule configuration based on a rule engine as claimed in claim 5, characterized in that: In the step 7, a risk control model executor is generated in the full-link risk control process of the risk control business, including: when drawing a risk control service decision flowchart, the risk control model set on the model node generates a risk control model executor.
7. A method for implementing risk control rule configuration based on a rule engine as claimed in claim 6, characterized in that: In step 8, the risk control service decision flowchart includes: a process start node, a risk control model node, a risk control decision node, and a process end node; the process start node is the starting point of the risk control service decision process, which is used to mark the start of the risk control service decision process; the risk control model node is used to execute the risk control model on the risk control model node when the process flows to such a node; the types of the risk control model nodes include: rule model nodes, rule set model nodes, decision tree model nodes, scorecard model nodes, blacklist model nodes, artificial intelligence model nodes, and third-party model nodes; the decision logic on the risk control decision node will calculate the next node of the current process according to the process data in the current process, which is used to control the flow direction of the process; the process end node is the end point of the risk control service decision process, which is used to mark the end of the risk control service decision process; the method of drawing the risk control service decision flowchart is: on the online canvas of the risk control service decision flowchart provided by the risk control platform, drag and connect each process node, set the model on the risk control model node, describe the process flow logic on the risk control decision node, and form a risk control service decision flowchart.
8. A method for implementing risk control rule configuration based on a rule engine as claimed in claim 7, characterized in that: In step 10, the risk control service decision process executor is an execution file in XML format that complies with the BPMN protocol and is required when the Drools rule engine decision flow module is running; the risk control service decision flowchart metadata parsing program converts the risk control service decision flowchart metadata into an executable file in XML format that complies with the BPMN protocol according to the standard BPMN protocol.
9. A method for implementing risk control rule configuration based on a rule engine as claimed in claim 8, characterized in that: The step 13 comprises: Step 13-1: Run the risk control service decision process executor and execute the process start node; Step 13-2: Execute the risk control model node, and search for the relevant risk control model executor in the knowledge base of the rule engine according to the risk control model set on the node; if the relevant risk control model executor is not found, search for the relevant model execution file from the database, generate the model executor through the model execution file, and add it to the knowledge base of the rule engine; Step 13-3: Run the risk control model executor and add the running results to the process session; the process session is a cache provided by the Drools rule engine to temporarily store process data in the risk control decision process execution. Each time the risk control decision execution process runs, a new process session will be generated; Step 13-4: Execute the risk control decision node and calculate the next risk control model node of the risk control service decision process according to the decision logic; Step 13-5: Repeat steps 13-1 to 13-4 until the decision-making process is executed to the end node of the process.
