CN116225417A - Financial platform decision engine management system and method based on big data - Google Patents
Financial platform decision engine management system and method based on big data Download PDFInfo
- Publication number
- CN116225417A CN116225417A CN202310504999.2A CN202310504999A CN116225417A CN 116225417 A CN116225417 A CN 116225417A CN 202310504999 A CN202310504999 A CN 202310504999A CN 116225417 A CN116225417 A CN 116225417A
- Authority
- CN
- China
- Prior art keywords
- decision
- rule
- execution
- result
- node
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/30—Creation or generation of source code
- G06F8/34—Graphical or visual programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/25—Integrating or interfacing systems involving database management systems
- G06F16/252—Integrating or interfacing systems involving database management systems between a Database Management System and a front-end application
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/25—Integrating or interfacing systems involving database management systems
- G06F16/258—Data format conversion from or to a database
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/30—Creation or generation of source code
- G06F8/31—Programming languages or programming paradigms
- G06F8/315—Object-oriented languages
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/30—Creation or generation of source code
- G06F8/36—Software reuse
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/60—Software deployment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/70—Software maintenance or management
- G06F8/71—Version control; Configuration management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/54—Interprogram communication
- G06F9/547—Remote procedure calls [RPC]; Web services
- G06F9/548—Object oriented; Remote method invocation [RMI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2209/00—Indexing scheme relating to G06F9/00
- G06F2209/54—Indexing scheme relating to G06F9/54
- G06F2209/549—Remote execution
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Computer Security & Cryptography (AREA)
- Computing Systems (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the field of decision engine management, in particular to a financial platform decision engine management system and a financial platform decision engine management method based on big data.
Description
Technical Field
The invention relates to the field of decision engine management, in particular to a financial platform decision engine management system and method based on big data.
Background
For financial credit business, the loan approval is an important link of the loan flow, the efficiency and the correctness of the loan approval are directly related to the loan risk, and the current business approval flow has the following problems:
firstly, manual approval is mostly adopted, and for internet online loans, the service requirements cannot be met in terms of efficiency and accuracy;
secondly, no intuitive system page is used by the air control personnel, and the air control rule is invisible to the business personnel;
thirdly, under a changeable wind control scene, the strategy can be adjusted in time according to service change and risk requirements, service personnel can not adjust the strategy independently, and the strategy needs to be adjusted by contact technicians, so that response is slow;
fourth, the wind control rule should be encoded in the service system, so that the data acquisition, the wind control rule coupling and the service system bring great pressure to the later development and maintenance work.
Disclosure of Invention
The invention aims to provide a financial platform decision engine management system and method based on big data, which are used for solving the problems in the background technology, and the invention provides the following technical scheme:
A financial platform decision engine management method based on big data, the method comprising the steps of:
s1, establishing accounts for wind control personnel of different departments and business lines through a system background and distributing roles and authorities;
s2, enabling air control personnel to enter a system interface, configuring a mode in a rule by adopting a hit mode in an air control model, configuring conditions in a scoring card by adopting a weight mode, configuring a formula editor by adopting an expression evaluator, configuring parameters for a rule set, constructing a tree structure by taking the rule and the rule set as a node, routing according to a value returned by each node to form a decision tree, extracting an element value in the rule set, constructing a matrix, and arranging the execution sequence and execution conditions of the rule, the rule set, the scoring card, the formula editor, the decision tree and the decision matrix to finish decision flow configuration;
s3, acquiring a configuration result in the S2, taking the configuration result of the wind control event as a service system request parameter, initiating a request instruction, and calling an interface in the SDK according to actual conditions;
s4, based on the calling interface in the S3, matching the parameters requested by the service system with the corresponding decision flow to be executed, initializing a predefined decision flow, and circularly executing each node in the decision flow;
S5, finishing execution of the decision flow according to the S4, and feeding back an execution result to the service system.
Further, the method of S2 includes the following steps:
step 1001, configuring a rule in which a rule includes a plurality of conditions, each condition is composed of a left value a, a right value B, and an operator, wherein the operator includes >, < and=, wherein the hit mode indicates whether output is yes or no after one condition is hit, the rule includes four modes, i.e., a first mode is satisfied for all conditions, a second mode is satisfied for any one condition, a third mode is not satisfied for all conditions, a fourth mode is not satisfied for at least one condition,
the execution logic of the rule is: after all conditions are executed, operation is carried out according to four modes, when all conditions are met, the execution result of each condition is judged, if one false exists, the false is finally returned, when any one condition is met, the execution result of each condition is judged, if one true exists, the true is finally returned, when all conditions are not met, the execution result of each condition is judged, if one true exists, the false is returned, when at least one condition is not met, the execution result of each condition is judged, and if one false exists, the true is returned;
Step 1002, configuring the conditions in the scoring cards by adopting a weight mode, wherein one scoring card consists of a plurality of conditions, each condition outputs a score according to a weight, denoted as F, wherein the weight mode indicates that after one condition is hit, a score is output, finally, adding the scores output by each condition as a final output result,
the f=q A *A+Q B * B, wherein Q A Weight value representing left value A, Q B A weight value representing a right value B, wherein the weight value is a preset constant for a database;
step 1003, evaluating four arithmetic operations and Boolean expressions in a formula editor by an expression evaluator, wherein the formula editor adopts an EvalEx-Java expression evaluator, the EvalEx is a Java convenient expression evaluator, mathematics and Boolean expressions can be evaluated, the formula editor is similar to a calculator, an air control person carries out four arithmetic operations and judgment of Boolean conditions on rules, fields, scoring cards and constant components through pages, and the scoring cards can be used as components for decision stream call;
step 1004, configuring parameters for a rule set, and executing after packaging multiple rules, supporting a traversal executing mode and a sequential executing mode, wherein the rule set comprises a plurality of rule sets,
After the traversing execution mode completes the execution of all rules in the rule set, the hit result is returned,
the sequential execution mode indicates that when one rule is hit, execution is stopped, and unexecuted rules are not executed any more;
step 1005, constructing a tree structure by taking the rule and the rule set as a node, and routing according to the value returned by each node to form a decision tree, wherein the value returned by each node is true or false;
step 1006, extracting an element value in the rule set and constructing a matrix, which is denoted as g= [1,0 ]] T Extracting the calculation result of the scoring card, and representing the calculation result in a matrix form, and marking as P F =[F 1 ,F 2 ,F 3 ,...,F n ] T Wherein n represents the number of scoring card data, and the decision tree result is represented in a matrix form and is recorded as J= [1,0] T Each matrix element is combined and calculated, and the output result is taken as a decision matrix result and is recorded as E= [ E ] 1 ,E 2 ,E 3 ,...,E n ] T Wherein E is 1 =1*F 1 *1;
Step 1007, arranging the rule, rule set, scoring card, formula editor, decision tree, decision matrix execution sequence and execution condition to complete decision flow configuration, implementing policy execution in the form of flow stream, executing decision service by service system by calling decision flow, outputting the result by decision engine through decision flow execution, wherein decision flow is the highest function module of decision engine, decision flow is a non-annular flow chart, flow nodes are composed of rule, rule set, scoring card, formula editor, rule tree, rule matrix, executing decision flow takes the output result executed by each node as the condition of route, and executing next node;
Step 1008, randomly acquiring a service scene, and completing configuration of the wind control event by configuring an event number, wherein the wind control event is defined as parameters transmitted by the service system, a return result of the wind control event is defined for the service system, and one wind control event is used as a service scene.
The invention combines each component as a node of a decision flow into a flow chart by configuring rules, rule sets, scoring cards, formula editors, decision books and decision matrixes, defines parameters transmitted by a business system as a parameter input by a wind control event, defines a return result of the wind control event for the business system to use, and provides data reference for selecting and calling interfaces according to the return result of the decision flow.
Further, the method of S3 includes the following steps:
step 2001, selecting an interface according to the real-time requirement of the decision flow return result and the call time between internal systems, wherein the interface call in the SDK includes two kinds, one is a synchronous interface, one is an asynchronous interface,
if t-alpha is more than or equal to 0 and less than or equal to 8, the interface selects a synchronous interface, if t-alpha is more than 8, the interface selects an asynchronous interface, wherein t represents the calling time between internal systems, alpha represents a preset constant of a database,
Wherein the synchronous interface is suitable for returning the place with the real-time requirement greater than the preset value alpha and the fast execution speed of the decision flow, the asynchronous interface is suitable for the place with the real-time requirement less than or equal to the preset value alpha and the slow execution speed of the decision flow and the need of pedestrian credit data, wherein the calling time between internal systems is currently defined as 8 seconds, the time is considered to be overtime when the execution speed exceeds 8 seconds, the decision execution result cannot be acquired when the execution speed exceeds 8 seconds, so the execution is completed in the place with the fast execution speed in less than 8 seconds or less, the slow execution speed of the decision flow represents the calling time between internal systems to exceed 8 seconds,
the execution time of the decision flow is generally consumed in the data acquisition, a large amount of time is consumed in the data acquisition process, when the execution time is longer than 8 seconds, an asynchronous interface is required to be adopted for processing, when the decision flow is abnormal, the scene that mechanisms such as decision flow retry and the like can be restored is supported, the service can still acquire the final result,
in step 2002, according to the judgment result of step 2001, if the request result is obtained in real time for one request, a synchronous interface is adopted, two corresponding results are adopted, the first type indicates that the decision engine executes an expected result which can be correctly returned in real time, the second type indicates that the decision engine executes an expected result which cannot be returned due to error reporting, wherein the second type indicates that the expected result cannot be returned due to overtime calling caused by exceeding a preset value because of internal execution time of the decision engine, and when the second type occurs, an asynchronous interface is adopted.
According to the method and the device, the call interface is selected according to the real-time requirement of the return result of the decision flow and the call time between internal systems, and the data are processed differently according to different interfaces, so that the situation that the expected result cannot be obtained is avoided in order to ensure that the request result is obtained in real time by one request.
Further, the method for adopting the asynchronous interface comprises the following steps:
step 3001, obtaining an application serial number, wherein the serial number represents a decision execution application, and the service system obtains the serial number at the first time without obtaining an execution result of a decision engine;
step 3002, taking the serial number obtained in step 3001 as an entry, inquiring the execution result, wherein the inquiry execution result is divided into two cases,
the method comprises the steps that a first situation is that an expected result is inquired and marked execution is completed, a second situation is that the expected result is not inquired and continuous training inquiry is needed until the result is acquired, inquiry is stopped, for the second situation, when an execution result cannot be acquired, the method is divided into two situations, one situation is that the decision engine internally executes a wrong report and needs to be solved through internal automatic retry, the other situation is that the decision engine internally executes a slower time, the problem of needing to continue waiting for the training result is solved, when the second situation occurs, a service system can solve the problem through a training circulation mode or a timing task mode, the training circulation step is that the service system acquires the result through the training mode, when a decision flow is abnormal, the expected result cannot be acquired all the time, a contact system administrator is needed to solve at the moment, an abnormal processing mechanism can be arranged in the service system, when the training exceeds a preset time, the result cannot still be past, the transaction mark fails or other results are simultaneously, in order to avoid the situation that the service is infinitely training to cause pressure to the decision engine, the timing task can be executed once every five minutes, the inquiry is executed, the second step is performed through the training circulation mode, if the result is still not required to be processed after the preset time, and the abnormal result is still no abnormal when the result is required to be processed.
Further, the method of S4 includes the following steps:
step 4001, according to the event number of step 1008, obtaining a decision flow to be executed correspondingly;
step 4002, initializing a predefined decision flow, wherein each node of the decision flow serves as an executable object, each object comprises a node id, a node type and a routing condition of the node, wherein the node type comprises a rule, a rule set, a grading card, a formula editor, a decision tree and a decision matrix, the routing condition of the node represents that one node outputs a result after execution, and when the output result is true, the id of the next node is recorded;
step 4003, binding a next node according to current node information after finishing initialization operation based on each node, and forming a tree structure;
step 4004, starting execution from the root node, obtaining a data source relied by the node, and judging whether overdue rule contents exist according to rules preset by the node;
step 4005, after the current node is executed, searching for the next executable node according to the node route, repeating steps 4001-4004, and circularly executing until all nodes are traversed, and outputting the execution result of the whole decision flow.
A financial platform decision engine management system based on big data, the system comprising the following modules:
and a data cleaning module: the data cleaning module is used for rechecking and checking the data, deleting repeated information and correcting errors;
rule management module: the rule management module is used for configuring a scoring card, rules, a rule set and a formula editor;
decision management module: the decision management module is used for forming a decision tree by constructing a tree structure, routing according to the result returned by the rule management module, and generating a decision matrix according to components in the rule management module;
decision stream management module: the decision flow management module realizes the execution of the strategy in a flow circulation mode by arranging the rule, the rule set, the grading card, the formula editor, the decision tree, the execution sequence and the execution condition of the decision matrix;
and a system management module: the system management module is used for logging in the system through the account number distributed by the wind control personnel, and setting account number authority in the system.
Further, the rule management module comprises a scoring card unit, a rule configuration unit, a rule set configuration unit and a formula editor unit:
The scoring card unit is used as a component for the decision flow to call;
the rule configuration unit is used for configuring rules by wind control personnel;
the rule set configuration unit packs and executes multiple rules and supports a traversing execution mode and a sequential execution mode;
the formula editor unit is used for carrying out four-rule operation and judgment of Boolean conditions on rules, fields, scoring cards and constant components.
Further, the decision management module comprises a decision tree unit and a decision matrix unit:
the decision tree unit is used for constructing a tree structure by taking the rule and the rule set as a node, and automatically producing a decision tree by routing according to the feedback result of each node;
the decision matrix unit is used for taking each rule or grading card component as a one-dimensional array, generating a multi-dimensional array through a plurality of arrays, and outputting a final matrix result through the result of combination of the plurality of arrays.
Further, the decision flow management module includes a decision flow configuration unit and a wind control event unit:
the decision flow configuration unit is used for executing the output result executed by each node and executing the next node as a routing condition;
the wind control event unit is used for configuring an event number aiming at a service scene and binding an executable decision flow.
Further, the system management module comprises an account management unit, a right management unit and a log management unit:
the account management unit is used for managing accounts of wind control personnel;
the authority management unit is used for authorizing the account number of the wind control personnel;
the log management unit is used for recording the operation of the wind control personnel in the system.
The invention aims to provide a set of model configuration management system for air control personnel, a decision data platform for cleaning air control indexes is provided, sdk packets are provided for a service system, interface call is carried out by a mode of acquiring service addresses through a registry, and decision flow execution service based on air control events is provided for the service system, the method can improve the efficiency of air control personnel model deployment and configuration, simultaneously strip out the execution of the air control model from the service system, can obviously reduce the difficulty of service logic realization, reduce the complexity of realizing complex service logic components, reduce the maintenance and expandability cost of application programs,
compared with the prior art, the invention realizes the stripped business rules by using the decision engine, can change the changeable business rules into maintainable and easy-to-maintain business rules, can quickly edit complex business rules without coding by using a good business rule designer provided by the decision engine, provides the on-line edited wind control model configuration functions of rules, scoring cards, decision trees, matrixes and the like by using the decision engine through visual pages, realizes real-time deployment, can define complex business rules by using the decision engine even if a business person who does not understand programming at all, provides a unified interface for an upper business system by using a micro-service mode by using the decision engine, can remarkably reduce the difficulty of realizing the business logic and reduce the maintenance and expandability cost of an application program.
Drawings
FIG. 1 is a flow chart of a financial platform decision engine management method based on big data according to the present invention;
FIG. 2 is a schematic diagram of a financial platform decision engine management system based on big data according to the present invention;
FIG. 3 is a flow chart of the execution of the big data based financial platform decision engine management system of the present invention;
FIG. 4 is a schematic diagram of a big data based decision tree configuration of a financial platform decision engine management system according to the present invention;
FIG. 5 is a schematic diagram of a decision flow of a financial platform decision engine management method based on big data according to the present invention;
FIG. 6 is a flow chart of a big data based financial platform decision engine execution of the present invention;
FIG. 7 is a diagram of a financial platform decision engine management system interface based on big data in accordance with the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1: referring to fig. 1, in this embodiment:
the financial platform decision engine management method based on big data is realized, and comprises the following steps:
s1, establishing accounts for wind control personnel of different departments and business lines through a system background and distributing roles and authorities;
s2, enabling air control personnel to enter a system interface, configuring a mode in a rule by adopting a hit mode in an air control model, configuring conditions in a scoring card by adopting a weight mode, configuring a formula editor by adopting an expression evaluator, configuring parameters for a rule set, constructing a tree structure by taking the rule and the rule set as a node, routing according to a value returned by each node to form a decision tree, extracting an element value in the rule set, constructing a matrix, and arranging the execution sequence and execution conditions of the rule, the rule set, the scoring card, the formula editor, the decision tree and the decision matrix to finish decision flow configuration;
the method of S2 comprises the following steps:
step 1001, configuring a pattern in a rule by adopting a hit pattern, wherein one rule comprises a plurality of conditions, each condition is composed of a left value A, a right value B and an operation symbol, wherein the operation symbol comprises >, < and=;
Step 1002, configuring conditions in the scoring cards by adopting a weight mode, wherein one scoring card consists of a plurality of conditions, each condition outputs a score according to one weight, and the score is marked as F,
the f=q A *A+Q B * B, wherein Q A Weight value representing left value A, Q B A weight value representing a right value B, wherein the weight value is a preset constant for a database;
step 1003, evaluating four arithmetic operations and a Boolean expression in the formula editor by an expression evaluator;
step 1004, configuring parameters for a rule set, and executing after packaging multiple rules, supporting a traversal executing mode and a sequential executing mode, wherein the rule set comprises a plurality of rule sets,
after the traversing execution mode completes the execution of all rules in the rule set, the hit result is returned,
the sequential execution mode indicates that when one rule is hit, execution is stopped, and unexecuted rules are not executed any more;
step 1005, constructing a tree structure by taking the rule and the rule set as a node, and performing routing according to the value returned by each node to form a decision tree;
step 1006, extracting an element value in the rule set and constructing a matrix, which is denoted as g= [1,0 ]] T Extracting the calculation result of the scoring card, and representing the calculation result in a matrix form, and marking as P F =[F 1 ,F 2 ,F 3 ,...,F n ] T Wherein n represents the number of scoring card data, and the decision tree result is represented in a matrix form and is recorded as J= [1,0] T Combining each matrix element and taking the output result as decision matrix resultDenoted as E= [ E 1 ,E 2 ,E 3 ,...,E n ] T Wherein E is 1 =1*F 1 *1;
Step 1007, arranging the rule, rule set, scoring card, formula editor, decision tree, decision matrix execution sequence and execution condition to complete decision flow configuration, implementing policy execution in flow form, executing decision service by service system by calling decision flow, and outputting the result by decision engine through decision flow execution;
step 1008, randomly acquiring a service scene, and completing configuration of the wind control event by configuring an event number, wherein the wind control event is defined as parameters transmitted by the service system, and a return result of the wind control event is defined for the service system to use.
S3, acquiring a configuration result in the S2, taking the configuration result of the wind control event as a service system request parameter, initiating a request instruction, and calling an interface in the SDK according to actual conditions;
the method of S3 comprises the following steps:
step 2001, selecting an interface according to real-time requirements of the decision flow return result and the call time between internal systems,
If t-alpha is more than or equal to 0 and less than or equal to 8, the interface selects a synchronous interface, if t-alpha is more than 8, the interface selects an asynchronous interface, wherein t represents the calling time between internal systems, alpha represents a preset constant of a database,
the synchronous interface is suitable for returning to the place where the real-time requirement is larger than the preset value alpha and the decision flow execution speed is high, and the asynchronous interface is suitable for returning to the place where the real-time requirement is smaller than or equal to the preset value alpha and the decision flow execution speed is low and pedestrian credit information data is needed;
step 2002, according to the judgment result of step 2001, if the request result is obtained in real time for one request, adopting a synchronous interface, wherein two corresponding results are adopted, the first type represents that the decision engine executes an expected result which can be correctly returned in real time, the second type represents that the decision engine executes an expected result which cannot be returned when reporting errors, the second type represents that the expected result cannot be returned when calling overtime is caused due to the fact that the internal execution time of the decision engine exceeds a preset value, and an asynchronous interface is adopted when the second condition occurs;
the method for adopting the asynchronous interface comprises the following steps:
step 3001, obtaining an application serial number, wherein the serial number represents a decision execution application, and the service system obtains the serial number at the first time without obtaining an execution result of a decision engine;
Step 3002, taking the serial number obtained in step 3001 as an entry, inquiring the execution result, wherein the inquiry execution result is divided into two cases,
the first case is that the expected result is inquired and the label execution is completed, and the second case is that the expected result is not inquired and the inquiry needs to be continuously trained until the result is obtained, and the inquiry is stopped.
S4, based on the calling interface in the S3, matching the parameters requested by the service system with the corresponding decision flow to be executed, initializing a predefined decision flow, and circularly executing each node in the decision flow;
the method of S4 comprises the following steps:
step 4001, according to the event number of step 1008, obtaining a decision flow to be executed correspondingly;
step 4002, initializing a predefined decision flow, wherein each node of the decision flow serves as an executable object, and each object comprises a node id, a node type and a routing condition of the node;
step 4003, binding a next node according to current node information after finishing initialization operation based on each node, and forming a tree structure;
step 4004, starting execution from the root node, obtaining a data source relied by the node, and judging whether overdue rule contents exist according to rules preset by the node;
Step 4005, after the current node is executed, searching for the next executable node according to the node route, repeating steps 4001-4004, and circularly executing until all nodes are traversed, and outputting the execution result of the whole decision flow.
And S5, finishing execution of the decision flow according to the S4, and feeding back an execution result to the service system.
In this embodiment:
a big data based financial platform decision engine management system (as shown in fig. 2) is disclosed for implementing the specific scheme content of the method.
Example 2: in this embodiment, the overall execution process of the decision engine is mainly described, including the configuration and principle of the wind control personnel on each wind control model, the service system calls the decision engine through sdk, the execution process of the decision execution engine and the process of calling the decision data platform to obtain the wind control index, as shown in figure 3,
step 1: the wind control personnel logs in the system according to the allocated account number, and after logging in the system, the wind control personnel can access corresponding menus and buttons according to the authority of the account number to perform wind control strategy configuration;
step 2: configuring an air control strategy, which mainly comprises rule configuration, rule set configuration, grading card configuration, formula editor configuration, decision tree configuration, decision matrix configuration, decision flow configuration and air control event configuration;
Rule configuration: the rule module is designed based on the principle that a rule consists of a plurality of conditions, wherein the condition with the age of more than 20 is a condition, the condition comprises three elements, the left value is age, the sign is >, and the right value is 20;
the rule adopts a hit mode, namely, the calculation logic according to the condition returns yes or no, wherein the execution of the rule is divided into four modes:
mode 1: all of the conditions are met and,
mode 2: any one of the conditions is satisfied,
mode 3: all of the conditions are not met,
mode 4: at least one of the conditions is not met,
setting a condition as condition, obtaining condition nA (return true), condition B (return true), condition C (return false) and condition D (return false) by hit mode,
the four calculation logics are respectively:
in mode 1, when all conditions are satisfied, that is, the result of the and operation performed by conditions conditionA, conditionB and conditions c is equal to false, the and operation performed by conditions conditional na and conditions b is equal to true,
in mode 2, when any one of the conditions is satisfied, that is, the result of the exclusive OR operation between the condition conditionA, conditionB and the condition C is equal to true, the exclusive OR operation between the condition C and the condition D is equal to false,
In mode 3, when all conditions are not satisfied, that is, the result of the and operation performed by the conditions b and c is equal to false, the result of the and operation performed by the conditions c and d is equal to true,
in mode 4, when at least one condition is not satisfied, that is, the result of the and operation performed by the conditions conditionA, conditionB and the conditions c is equal to true, the result of the and operation performed by the conditions conditional na and the conditions b is equal to false,
rule set configuration: the principle of the rule set module is that n rules are packaged, each rule is circularly executed, wherein the rule set supports a sequential execution mode and a traversing execution mode, the sequential execution mode is used for executing each rule of the rule package, and if one rule set has three rules rule A, rule B and rule C, the execution formula of one rule set is as follows:
rule result = rule a & rule b & rule c,
after the traversing execution mode is that the execution rule meets the condition, stopping execution, assuming that one rule set comprises three rules rule A, rule B and rule C, wherein the passing number or refusing number n of rule execution is rule set configuration bar conditionna, the conditionna is n and is more than 1, starting the rule set, after the rule A is executed, judging whether n is more than 1, if yes, ending the execution of rule B and rule C, returning a result, if not, continuing to execute rule B, and so on, completing the execution of the whole rule set,
Score card configuration: the principle of the scoring card module is that one scoring card consists of three elements, namely: the condition, the weight and the score are that two conditions of condition na and condition b exist, each condition is configured with a weight, after each condition is executed, a score is output, and then a scoring card executes the following process: when the condition is executed, if true, the output result resulta=socre, when the condition is executed, if true, the output result resulta=socre, and when the condition is executed, if true, the output result resulta=socre, weigth, the output result of the final scoring card indicates that the scores obtained after each condition is executed are added, namely, result=resulta+resultab,
formula editor configuration: the algorithm or function of the page configuration is predefined as an EvalEx-Java expression, the execution of the mathematical and Boolean expressions is completed, the final output result is the execution result of the formula editor,
decision tree configuration: the decision tree is mainly used for meeting complex business scenes, rules, rule sets and grading cards can be used as a decision tree node, each node is combined to form a tree structure, as shown in figure 4,
decision matrix configuration: the design principle of the decision matrix is that a plurality of nodes such as rules, grading cards and rule sets are combined to form a multidimensional array, each node is provided with a one-dimensional array, and the structure is as follows:
ruleA={ruleA1-ture,ruleA2-false},
socreA={socreA1-value1,socreA2-value2,socreA3-value3,…},
ruleSetsA={ruleSetsA1-true,ruleSetsA2-false,…},
The final output result is a value of a multi-dimensional array configuration, taking a three-dimensional array result [2] [3] [2] as an example, as follows:
reulst[0][0][0]=value1,
reulst[0][1][0]=value2,
reulst[0][1][1]=value3,
reulst[1][1][1]=value4,
the final output result of the decision matrix is result,
decision stream configuration: the method realizes the execution of the strategy by arranging the execution sequence and the execution condition of rules, rule sets, scoring cards, formula editors, decision trees and decision matrixes in a flow circulation mode, the business system executes the decision service by calling the decision flow, the decision engine outputs the result through the execution of the decision flow,
the decision flow is a non-annular flow chart, the flow nodes are composed of rules, rule sets, scoring cards, formula editors, rule trees and rule matrixes, the execution of the decision flow takes the output result of each node as a routing condition to execute the next node, if the start node is rule A, the execution result of rule A is true or false, when rule A is executed as true, the execution result is the node connected when true, when rule A is executed as false, the execution result is the node connected when false, as shown in figure 5,
wind control event configuration: an air control event can be used as a business scene, and the air control event configuration mainly comprises:
1: defining an event number as an entry called by the service system sdk;
2: defining the entry of the event interface, and defining according to the data source relied on by the wind control model, if the identity card, the name, the mobile phone number, the loan business serial number and the like need to be uploaded;
3: and defining a return result of the event, wherein if the event is trusted, a credit limit, a user grade and the like need to be returned.
Step 3: the service system calls a decision engine through sdk, and the decision engine provides a synchronous interface and an asynchronous interface, wherein the synchronous interface refers to a real-time return result after the service system calls the decision engine;
step 4: execution of the decision execution engine,
the execution engine requests the wind control event number of the participation through the service system, acquires the decision flow to be executed, and loads and deploys each node component of the decision flow into the memory after the execution engine is preprogrammed, and starts to circularly execute each node according to the flow, as shown in fig. 6;
step 5: and after the loop execution finishes the decision flow executable node, outputting a final decision result.
Example 3: this embodiment describes that the service system initiates the invocation of the decision engine through sdk, and the decision engine completes the execution of the decision flow by constructing the decision flow node and acquiring the data through the decision data platform, as shown in fig. 7:
Step 1: the business system builds an entry for calling the decision engine according to the business scene, and calls the decision engine through sdk to execute the decision flow;
step 2: after receiving the request, the decision engine inquires the corresponding decision flow through the event number in the request parameter;
step 3: starting to construct a decision flow, wherein one decision flow is composed of a plurality of wind control components, and is used as a node of the decision flow, and mainly comprises the following steps: rules, rule sets, scoring cards, formula editors, decision trees, decision matrices,
each node is constructed as follows:
1: current node id
2: current node type (rule, rule set, scoring card, formula editor, decision tree, and decision matrix)
3: route information executed by the current node mainly comprises:
if the result after the execution of one node is true or false, the condition is that when the execution result is true, the next decision stream node id is recorded, and the next executable node next Id associated with the current node is saved by analogy;
after each node is constructed, starting each node from a root node, and constructing the whole decision stream into a tree structure through the hierarchical relationship of id and nextId;
Step 4: executing a decision flow, namely starting execution from a root node, acquiring the next node from the route information of the current node after each node is executed, and circularly executing until the execution reaches an end node;
step 5: when executing each node, if the node depends on the support of a data source, calling a decision data platform in an http mode at the moment to acquire data;
step 6: the process of the data platform obtaining data is as follows:
step 6.1, classifying the data depending on the decision by the data platform, for example: list class data, credit class data and tax class data;
step 6.2, after classification is completed, inquiring a corresponding data source interface according to the data type, and acquiring each type of data in a multi-thread mode to provide data acquisition efficiency, wherein the data is realized by using a thread pool in the embodiment;
step 6.3, after the original data is obtained, processing and cleaning the data, and converting the data into a data structure required by a decision engine;
step 6.4, returning a final result;
and 7, circularly executing the processes of the step 5 and the step 6, and outputting the decision result to the service system after finishing the execution of the decision flow.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Finally, it should be noted that: the foregoing description is only a preferred embodiment of the present invention, and the present invention is not limited thereto, but it is to be understood that modifications and equivalents of some of the technical features described in the foregoing embodiments may be made by those skilled in the art, although the present invention has been described in detail with reference to the foregoing embodiments. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A financial platform decision engine management method based on big data, the method comprising the steps of:
s1, establishing accounts for wind control personnel of different departments and business lines through a system background and distributing roles and authorities;
s2, enabling air control personnel to enter a system interface, configuring a mode in a rule by adopting a hit mode in an air control model, configuring conditions in a scoring card by adopting a weight mode, configuring a formula editor by adopting an expression evaluator, configuring parameters for a rule set, constructing a tree structure by taking the rule and the rule set as a node, routing according to a value returned by each node to form a decision tree, extracting an element value in the rule set, constructing a matrix, and arranging the execution sequence and execution conditions of the rule, the rule set, the scoring card, the formula editor, the decision tree and the decision matrix to finish decision flow configuration;
s3, acquiring a configuration result in the S2, taking the configuration result of the wind control event as a service system request parameter, initiating a request instruction, and calling an interface in the SDK according to actual conditions;
s4, based on the calling interface in the S3, matching the parameters requested by the service system with the corresponding decision flow to be executed, initializing a predefined decision flow, and circularly executing each node in the decision flow;
S5, finishing execution of the decision flow according to the S4, and feeding back an execution result to the service system.
2. The financial platform decision engine management method based on big data according to claim 1, wherein the method of S2 comprises the steps of:
step 1001, configuring a pattern in a rule by adopting a hit pattern, wherein one rule comprises a plurality of conditions, each condition is composed of a left value A, a right value B and an operation symbol, wherein the operation symbol comprises >, < and=;
step 1002, configuring conditions in the scoring cards by adopting a weight mode, wherein one scoring card consists of a plurality of conditions, each condition outputs a score according to one weight, and the score is marked as F,
the f=q A *A+Q B * B, wherein Q A Weight value representing left value A, Q B A weight value representing a right value B, wherein the weight value is a preset constant for a database;
step 1003, evaluating four arithmetic operations and a Boolean expression in the formula editor by an expression evaluator;
step 1004, configuring parameters for a rule set, and executing after packaging multiple rules, supporting a traversal executing mode and a sequential executing mode, wherein the rule set comprises a plurality of rule sets,
after the traversing execution mode completes the execution of all rules in the rule set, the hit result is returned,
The sequential execution mode indicates that when one rule is hit, execution is stopped, and unexecuted rules are not executed any more;
step 1005, constructing a tree structure by taking the rule and the rule set as a node, and performing routing according to the value returned by each node to form a decision tree;
step 1006, extracting an element value in the rule set and constructing a matrix, which is denoted as g= [1,0 ]] T Extracting the calculation result of the scoring card, and representing the calculation result in a matrix form, and marking as P F =[F 1 ,F 2 ,F 3 ,...,F n ] T Wherein n represents the number of scoring card data, and the decision tree result is represented in a matrix form and is recorded as J= [1,0] T Each matrix element is combined and calculated, and the output result is taken as a decision matrix result and is recorded as E= [ E ] 1 ,E 2 ,E 3 ,...,E n ] T Wherein E is 1 =1*F 1 *1;
Step 1007, arranging the rule, rule set, scoring card, formula editor, decision tree, decision matrix execution sequence and execution condition to complete decision flow configuration, implementing policy execution in flow form, executing decision service by service system by calling decision flow, and outputting the result by decision engine through decision flow execution;
step 1008, randomly acquiring a service scene, and completing configuration of the wind control event by configuring an event number, wherein the wind control event is defined as parameters transmitted by the service system, and a return result of the wind control event is defined for the service system to use.
3. The financial platform decision engine management method based on big data according to claim 2, wherein the method of S3 comprises the steps of:
step 2001, selecting an interface according to real-time requirements of the decision flow return result and the call time between internal systems,
if t-alpha is more than or equal to 0 and less than or equal to 8, the interface selects a synchronous interface, if t-alpha is more than 8, the interface selects an asynchronous interface, wherein t represents the calling time between internal systems, alpha represents a preset constant of a database,
the synchronous interface is suitable for returning to the place where the real-time requirement is larger than the preset value alpha and the decision flow execution speed is high, and the asynchronous interface is suitable for returning to the place where the real-time requirement is smaller than or equal to the preset value alpha and the decision flow execution speed is low and pedestrian credit information data is needed;
in step 2002, according to the judgment result of step 2001, if the request result is obtained in real time for one request, a synchronous interface is adopted, two corresponding results are adopted, the first type indicates that the decision engine executes an expected result which can be correctly returned in real time, the second type indicates that the decision engine executes an expected result which cannot be returned due to error reporting, wherein the second type indicates that the expected result cannot be returned due to overtime calling caused by exceeding a preset value because of internal execution time of the decision engine, and when the second type occurs, an asynchronous interface is adopted.
4. A method of managing a financial platform decision engine based on big data according to claim 3, wherein the method using an asynchronous interface comprises the steps of:
step 3001, obtaining an application serial number, wherein the serial number represents a decision execution application, and the service system obtains the serial number at the first time without obtaining an execution result of a decision engine;
step 3002, taking the serial number obtained in step 3001 as an entry, inquiring the execution result, wherein the inquiry execution result is divided into two cases,
the first case is that the expected result is inquired and the label execution is completed, and the second case is that the expected result is not inquired and the inquiry needs to be continuously trained until the result is obtained, and the inquiry is stopped.
5. The financial platform decision engine management method based on big data according to claim 4, wherein the method of S4 comprises the steps of:
step 4001, according to the event number of step 1008, obtaining a decision flow to be executed correspondingly;
step 4002, initializing a predefined decision flow, wherein each node of the decision flow serves as an executable object, and each object comprises a node id, a node type and a routing condition of the node;
Step 4003, binding a next node according to current node information after finishing initialization operation based on each node, and forming a tree structure;
step 4004, starting execution from the root node, obtaining a data source relied by the node, and judging whether overdue rule contents exist according to rules preset by the node;
step 4005, after the current node is executed, searching for the next executable node according to the node route, repeating steps 4001-4004, and circularly executing until all nodes are traversed, and outputting the execution result of the whole decision flow.
6. A financial platform decision engine management system based on big data, the system comprising the following modules:
and a data cleaning module: the data cleaning module is used for rechecking and checking the data, deleting repeated information and correcting errors;
rule management module: the rule management module is used for configuring a scoring card, rules, a rule set and a formula editor;
decision management module: the decision management module is used for forming a decision tree by constructing a tree structure, routing according to the result returned by the rule management module, and generating a decision matrix according to components in the rule management module;
Decision stream management module: the decision flow management module realizes the execution of the strategy in a flow circulation mode by arranging the rule, the rule set, the grading card, the formula editor, the decision tree, the execution sequence and the execution condition of the decision matrix;
and a system management module: the system management module is used for logging in the system through the account number distributed by the wind control personnel, and setting account number authority in the system.
7. The financial platform decision engine management system based on big data according to claim 6, wherein the rule management module comprises a score card unit, a rule configuration unit, a rule set configuration unit, and a formula editor unit:
the scoring card unit is used as a component for the decision flow to call;
the rule configuration unit is used for configuring rules by wind control personnel;
the rule set configuration unit packs and executes multiple rules and supports a traversing execution mode and a sequential execution mode;
the formula editor unit is used for carrying out four-rule operation and judgment of Boolean conditions on rules, fields, scoring cards and constant components.
8. The financial platform decision engine management system based on big data of claim 7, wherein the decision management module comprises a decision tree unit and a decision matrix unit:
The decision tree unit is used for constructing a tree structure by taking the rule and the rule set as a node, and automatically producing a decision tree by routing according to the feedback result of each node;
the decision matrix unit is used for taking each rule or grading card component as a one-dimensional array, generating a multi-dimensional array through a plurality of arrays, and outputting a final matrix result through the result of combination of the plurality of arrays.
9. The financial platform decision engine management system based on big data according to claim 8, wherein the decision flow management module comprises a decision flow configuration unit and a wind control event unit:
the decision flow configuration unit is used for executing the output result executed by each node and executing the next node as a routing condition;
the wind control event unit is used for configuring an event number aiming at a service scene and binding an executable decision flow.
10. The financial platform decision engine management system based on big data according to claim 9, wherein the system management module comprises an account management unit, a rights management unit, and a log management unit:
the account management unit is used for managing accounts of wind control personnel;
The authority management unit is used for authorizing the account number of the wind control personnel;
the log management unit is used for recording the operation of the wind control personnel in the system.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310504999.2A CN116225417B (en) | 2023-05-08 | 2023-05-08 | Financial platform decision engine management system and method based on big data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310504999.2A CN116225417B (en) | 2023-05-08 | 2023-05-08 | Financial platform decision engine management system and method based on big data |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116225417A true CN116225417A (en) | 2023-06-06 |
CN116225417B CN116225417B (en) | 2023-07-21 |
Family
ID=86569851
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310504999.2A Active CN116225417B (en) | 2023-05-08 | 2023-05-08 | Financial platform decision engine management system and method based on big data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116225417B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116777606A (en) * | 2023-07-05 | 2023-09-19 | 无锡锡商银行股份有限公司 | Intelligent anti-fraud system and method for bank credit based on relation graph |
CN117785157A (en) * | 2023-12-29 | 2024-03-29 | 北京开科唯识技术股份有限公司 | Decision engine based on financial wind control business rule scene and implementation method |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110609830A (en) * | 2019-08-15 | 2019-12-24 | 上海新颜人工智能科技有限公司 | Decision engine system based on auditing strategy |
CN111489254A (en) * | 2020-04-14 | 2020-08-04 | 上海数喆数据科技有限公司 | Credit risk assessment intelligent engine system based on historical credit big data |
CN113256408A (en) * | 2021-07-06 | 2021-08-13 | 中证信用云科技(深圳)股份有限公司 | Risk control method and system based on consumption finance and computer equipment |
CN113793213A (en) * | 2021-09-27 | 2021-12-14 | 武汉众邦银行股份有限公司 | Method and device for realizing decision mode of asynchronous credit wind control breakpoint continued operation |
WO2022105153A1 (en) * | 2020-11-18 | 2022-05-27 | 平安普惠企业管理有限公司 | Risk control decision-making method and apparatus, device, and storage medium |
CN115756901A (en) * | 2022-11-30 | 2023-03-07 | 沈阳尚源智慧科技有限公司 | Business decision processing method and rule engine system thereof |
-
2023
- 2023-05-08 CN CN202310504999.2A patent/CN116225417B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110609830A (en) * | 2019-08-15 | 2019-12-24 | 上海新颜人工智能科技有限公司 | Decision engine system based on auditing strategy |
CN111489254A (en) * | 2020-04-14 | 2020-08-04 | 上海数喆数据科技有限公司 | Credit risk assessment intelligent engine system based on historical credit big data |
WO2022105153A1 (en) * | 2020-11-18 | 2022-05-27 | 平安普惠企业管理有限公司 | Risk control decision-making method and apparatus, device, and storage medium |
CN113256408A (en) * | 2021-07-06 | 2021-08-13 | 中证信用云科技(深圳)股份有限公司 | Risk control method and system based on consumption finance and computer equipment |
CN113793213A (en) * | 2021-09-27 | 2021-12-14 | 武汉众邦银行股份有限公司 | Method and device for realizing decision mode of asynchronous credit wind control breakpoint continued operation |
CN115756901A (en) * | 2022-11-30 | 2023-03-07 | 沈阳尚源智慧科技有限公司 | Business decision processing method and rule engine system thereof |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116777606A (en) * | 2023-07-05 | 2023-09-19 | 无锡锡商银行股份有限公司 | Intelligent anti-fraud system and method for bank credit based on relation graph |
CN116777606B (en) * | 2023-07-05 | 2024-01-23 | 无锡锡商银行股份有限公司 | Intelligent anti-fraud system and method for bank credit based on relation graph |
CN117785157A (en) * | 2023-12-29 | 2024-03-29 | 北京开科唯识技术股份有限公司 | Decision engine based on financial wind control business rule scene and implementation method |
Also Published As
Publication number | Publication date |
---|---|
CN116225417B (en) | 2023-07-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104216912B (en) | The implementation method and device of a kind of business form work fluidisation without intrusive mood | |
CN103038752B (en) | A kind of method for management software problem report, system and equipment | |
CN1713196B (en) | Product ordering system based on automatic design grid | |
CN1485762A (en) | System for the telemanagement of weaving looms in weaving premises and method of assistance thereof | |
CN108234127A (en) | A kind of Internet of Things method and device based on block chain | |
CN1533542A (en) | Systems and methods for creating financial advice applications | |
CN116225417B (en) | Financial platform decision engine management system and method based on big data | |
WO2002077753A2 (en) | Automated transaction management system and method | |
CN1545674A (en) | Business process policy object | |
CN106358289A (en) | Data acquiring method and device based on crowdsourcing and server | |
CN113254061B (en) | Business decision method, system and storage medium based on rule engine | |
CN114048024A (en) | Task deployment method, device, equipment, storage medium and product | |
CN115170048B (en) | Workflow realization method, system and medium based on model and rule | |
CN110111620A (en) | A kind of virtual reality learning platform | |
CN106897107A (en) | A kind of web teledatas calculating realizes system and method with emulation | |
CN110489142A (en) | Appraisal procedure and device, storage medium, the terminal of equipment software upgrading | |
CN117134986A (en) | Method, system and device for generating external network honey point based on ChatGPT | |
CN108509259A (en) | Obtain the method and air control system in multiparty data source | |
CN115130959B (en) | Method, system, terminal and storage medium for generating spare part BOM | |
Chun | Using AI for e-Government Automatic Assessment of Immigration Application Forms. | |
CN114880387A (en) | Data integration script generation method and device, storage medium and electronic equipment | |
CN108711074A (en) | Business sorting technique, device, server and readable storage medium storing program for executing | |
CN111651262B (en) | Batch task generation method and device | |
Di Lorenzo et al. | Turning web applications into web services by wrapping techniques | |
CN115687054A (en) | Self-adaptive test method and device based on service segmentation and restoration |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |