CN112463840B - Real-time business wind control system and method based on rule engine - Google Patents
Real-time business wind control system and method based on rule engine Download PDFInfo
- Publication number
- CN112463840B CN112463840B CN202110135446.5A CN202110135446A CN112463840B CN 112463840 B CN112463840 B CN 112463840B CN 202110135446 A CN202110135446 A CN 202110135446A CN 112463840 B CN112463840 B CN 112463840B
- Authority
- CN
- China
- Prior art keywords
- wind control
- rule
- feature
- strategy
- group
- 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.)
- Active
Links
Images
Classifications
-
- 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
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2455—Query execution
- G06F16/24564—Applying rules; Deductive queries
-
- 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
- G06F16/22—Indexing; Data structures therefor; Storage structures
- G06F16/2282—Tablespace storage structures; Management thereof
-
- 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
- G06F16/254—Extract, transform and load [ETL] procedures, e.g. ETL data flows in data warehouses
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
Abstract
The invention provides a real-time business wind control system and method based on a rule engine, which are applied to at least one wind control scene, and the system comprises: the configuration layer is provided with a strategy rule table aiming at each wind control scene, and the strategy rule table comprises at least one strategy rule group; the rule engine is used for extracting features of the accessed wind control basic data to obtain extracted features, determining a wind control scene according to the extracted features, determining feature conditions in a policy rule group which accords with the extracted features in a policy rule table corresponding to the wind control scene, and determining an optimal gradient according to a preset policy combination mode and feature attributes corresponding to all feature conditions which accord with the extracted features; and the action engine executes the execution action corresponding to the optimal gradient. Has the advantages that: and performing wind control hit decision aiming at a plurality of wind control scenes, and effectively selecting the optimal gradient of multiple strategies, multiple actions and multiple gradients in a single wind control scene.
Description
Technical Field
The invention relates to the field of wind control systems, in particular to a real-time business wind control system and method based on a rule engine.
Background
With the rapid development of internet finance, online payment and application thereof are gradually popularized, and convenience is brought to the life of people. Meanwhile, the scale of network black-yielding for profit is gradually enlarged, and phenomena of high transaction risk, information leakage, increased fraud events and the like are brought. Enterprises construct a wind control system, set risk threshold values, and perform risk identification, risk rating and risk evasion so as to ensure that the services of normal users are stably developed. But as the service scenes are increasing and complicated, the protection and black product based on the service security field are increasingly determined. With the help of the increasing perfection and development of the black industry chain, the black products are frequently accessed through pseudo base stations and other modes continuously, and the service loopholes in the black products are detected, utilized and changed.
Therefore, a wind control system is needed, however, the wind control system commonly used in the industry generally includes three modules: the data engine can provide general abstract query capability for the model/rule engine, and the general abstract query capability comprises real-time data query capability, offline data query capability and the like; the model engine, namely an algorithm engine, is used for executing a carrier for a business modeling process by converting business data into an abstract business model, supports relatively complex algorithms and function calculation processes, is generally formulated and implemented by research and development personnel, and has the characteristics of certain variability and customization; the rule engine is generally a feature description of the wind control business rules, and the direct implementation of business wind control business requirements. The industry develops and customizes the wind control system by combining the service characteristics of the industry through the combination of the three engines so as to achieve the purpose of avoiding or finding wind control risks.
However, the wind control system has the following defects:
lack of solution to multi-wind control strategy scenario
(1) Under the condition of only a model, data and a rule engine, a wind control hit decision can be usually only carried out for a single scene, and the problem of flexibly configuring and executing a plurality of strategies in the same scene according to sequence, priority, grade, mutual exclusion and the like is difficult to solve.
(2) The lack of a solution for realizing single-policy multi-gradient actions generally integrates execution of the wind control actions by using a rule engine, and the lack of an independent action engine and gradient setting causes limitations of policy action triggering, and often fails to effectively configure hits of single actions or multiple actions.
Disclosure of Invention
Aiming at the problems in the prior art, a real-time business wind control system and a method based on a rule engine are provided.
The specific technical scheme is as follows:
a real-time service wind control system based on a rule engine is applied to at least one wind control scene and comprises the following components:
the configuration layer is provided with a strategy rule table aiming at each wind control scene, the strategy rule table comprises at least one strategy rule group, and the strategy rule group is provided with a rule group attribute, each characteristic condition and a characteristic attribute of the characteristic condition;
the rule engine is used for extracting features of the accessed wind control basic data to obtain extracted features, determining a wind control scene according to the extracted features, determining feature conditions in a policy rule group which accords with the extracted features in a policy rule table corresponding to the wind control scene, and determining an optimal gradient according to a preset policy combination mode and feature attributes corresponding to all feature conditions which accord with the extracted features;
and the action engine executes the execution action corresponding to the optimal gradient.
Preferably, the real-time business wind control system based on the rule engine, wherein the rule engine includes:
the receiving module is used for receiving the wind control basic data;
the characteristic extraction module is connected with the receiving module and used for extracting the characteristics of the accessed wind control basic data to obtain extracted characteristics;
the scene confirmation module is connected with the feature extraction module and used for determining the wind control scene where the extracted features are located and the strategy rule table in the wind control scene;
the characteristic determining module is connected with the scene confirming module, reads the characteristic conditions in the strategy rule group according to the first priority reading sequence of the strategy rule table and the second priority reading sequence of the strategy rule group, obtains the characteristic conditions met by the extracted characteristics and the characteristic attributes of the characteristic conditions, and determines the in-group gradient corresponding to the strategy rule group according to the sum of the characteristic attributes corresponding to the characteristic conditions met by all the extracted characteristics in the strategy rule group;
and the selection module is connected with the characteristic determination module and is used for selecting the intra-group gradient which accords with the strategy combination mode from the intra-group gradients in all the strategy rule groups as the optimal gradient.
Preferably, the real-time service wind control system based on the rule engine, wherein the characteristic determining module reads the characteristic conditions in the policy rule set according to the characteristic reading mode;
the feature reading mode specifically includes:
and setting a reading rule, and reading the characteristic conditions in the strategy rule group according to the second priority reading sequence of the strategy rule group when the extracted characteristics meet the reading rule.
Preferably, the real-time service wind control system based on the rule engine, wherein the policy combination mode specifically includes: and selecting the intra-group gradient in the strategy rule group corresponding to the highest priority in the first priority reading sequence as the optimal gradient.
Preferably, the real-time business wind control system based on the rule engine, wherein the rule engine further includes:
and the rule group reading module is connected with the scene confirmation module and determines a first priority reading sequence of the strategy rule groups in the strategy rule table according to the rule group attribute of each strategy rule group in the strategy rule table.
Preferably, the real-time service wind control system based on the rule engine, wherein the characteristic conditions are correspondingly provided with condition numbers, and the size of the condition numbers is set as a second priority reading sequence.
Preferably, the real-time business wind control system based on the rule engine is characterized in that the first condition atom of each characteristic condition in the strategy rule group is the same, and the second condition atom is different.
Preferably, the real-time service wind control system based on the rule engine, wherein the configuration layer includes a scene configuration module, and when the real-time service wind control system is applied to a plurality of wind control scenes, the scene configuration module is used for setting an execution sequence of the wind control scenes.
Preferably, the real-time business wind control system based on the rule engine further comprises an access layer and a data engine, wherein the access layer is connected with the data engine;
the access layer is connected with the configuration layer, and receives and distributes the strategy rule table under each wind control scene to the data engine;
the data engine preprocesses the strategy rule table under each wind control scene, and stores the preprocessed strategy rule table under each wind control scene for the rule engine to call.
The method also comprises a real-time business wind control method based on the rule engine, wherein the method comprises the following steps:
a strategy rule table is arranged aiming at each wind control scene, the strategy rule table comprises at least one strategy rule group, and the strategy rule group is provided with a rule group attribute, each characteristic condition and a characteristic attribute of the characteristic condition;
carrying out feature extraction on the accessed wind control basic data to obtain extracted features, determining a wind control scene according to the extracted features, setting a strategy combination mode, determining feature conditions in a strategy rule group which accords with the extracted features in a strategy rule table corresponding to the wind control scene, and determining an optimal gradient according to the strategy combination mode and all the feature conditions which accord with the extracted features;
and executing the executing action corresponding to the optimal gradient.
The technical scheme has the following advantages or beneficial effects:
a corresponding strategy rule table is set for each wind control scene in a configuration layer, so that wind control hit decision is carried out for a plurality of wind control scenes;
the rule engine is adopted to accord with the optimal gradient in the strategy rule group of the extracted characteristic and strategy combination mode in the strategy rule table corresponding to the wind control scene according to the extracted characteristic and strategy combination mode, so that the optimal gradient of multiple strategies, multiple actions and multiple gradients in a single wind control scene is effectively selected, and the action is determined to be executed according to the optimal gradient;
flexible matching among strategy rule groups can be realized through the rule engine, and the requirement of universal configuration of the wind control rule is further met;
and by adding an action engine, expanding the responsibility boundary of the decoupling rule engine and executing the execution action corresponding to the optimal gradient selected by the rule engine.
Drawings
Embodiments of the present invention will now be described more fully hereinafter with reference to the accompanying drawings. The drawings are, however, to be regarded as illustrative and explanatory only and are not restrictive of the scope of the invention.
FIG. 1 is a first schematic block diagram of an embodiment of a rules engine based real-time business wind control system of the present invention;
FIG. 2 is a schematic block diagram of a second embodiment of a rules engine based real-time business wind control system of the present invention;
FIG. 3 is a schematic block diagram of a rules engine of an embodiment of a rules engine based real-time business wind control system of the present invention;
FIG. 4 is a schematic block diagram of a policy rules table in a configuration layer of an embodiment of a rules engine based real-time traffic wind control system of the present invention;
fig. 5 is a flowchart of an embodiment of a method for real-time traffic wind control based on a rule engine according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that the embodiments and features of the embodiments may be combined with each other without conflict.
The invention is further described with reference to the following drawings and specific examples, which are not intended to be limiting.
The invention includes a real-time service wind control system based on a rule engine, which is applied to at least one wind control scene, as shown in fig. 1, the real-time service wind control system includes:
the configuration layer 1 is provided with a strategy rule table aiming at each wind control scene, the strategy rule table comprises at least one strategy rule group, and the strategy rule group is provided with a rule group attribute, each characteristic condition and a characteristic attribute of the characteristic condition;
in the above embodiment, the configuration layer 1 represents a visual representation of the entire real-time service wind control system, and the configuration layer 1 supports entry of a policy rule table corresponding to each wind control scene of the wind control system, where the policy rule table includes at least one policy rule group, each policy rule group is provided with a rule group attribute, each characteristic condition and a characteristic attribute of the characteristic condition, and each policy rule group supports a user to perform configuration modification.
In the above embodiment, as shown in fig. 4, in the configuration layer 1, one wind control scenario corresponds to one policy rule table, one policy rule table may be provided with one or more policy rule groups, each policy rule group is provided with a corresponding rule group attribute, each policy rule group is further provided with one or more feature conditions, and each feature condition is provided with a corresponding feature attribute;
that is, the wind control scenario and the policy rule table are in a one-to-one relationship, the policy rule table and the policy rule group may be in a one-to-one relationship or in a one-to-many relationship, and the policy rule group and the characteristic condition may be in a one-to-one relationship or in a one-to-many relationship.
In the above embodiment, a corresponding policy rule table is set for each wind control scene in the configuration layer 1, so that a wind control hit decision is performed for a plurality of wind control scenes.
Further, the real-time service wind control system further includes: the rule engine 2 is used for extracting features of the accessed wind control basic data to obtain extracted features, determining a wind control scene according to the extracted features, determining feature conditions in a policy rule group which accords with the extracted features in a policy rule table corresponding to the wind control scene, and determining an optimal gradient according to a preset policy combination mode and feature attributes corresponding to all feature conditions which accord with the extracted features;
in the above embodiment, the rule engine 2 is adopted to determine the optimal gradient according to the feature attributes corresponding to all the feature conditions, and determine the optimal gradient of multiple strategies, multiple actions and multiple gradients in a single wind control scene, according to all the feature conditions in each policy rule group corresponding to the extracted feature and policy combination mode in the policy rule table corresponding to the wind control scene.
In the above embodiment, the rule engine 2 may implement flexible matching between policy rule sets, and further meet the requirement of generic configuration of the wind control rules.
Further, the real-time service wind control system further includes: the action engine 3 executes an execution action corresponding to the optimal gradient.
In the above embodiment, by adding the action engine 3, the responsibility boundary of the decoupling rule engine 2 is expanded, and the execution action corresponding to the optimal gradient selected by the rule engine 2 is executed.
In the above embodiment, the action engine 3 returns a set of policy hit results after executing an execution action, and the set is accompanied by key information in the whole execution process of the wind control rule, including wind control basic data, feature conditions corresponding to a hit policy rule group, feature attributes of the feature conditions, and the like, which is used as an action execution basis for the subsequent action engine 3, because the set is a structure, the execution of multi-policy and multi-gradient actions can be triggered by linking the action engine 3, and the action engine 3 of the present system has functions of notification, marking, and the like.
As a preferred embodiment, one gradient corresponds to one performed action.
As a preferred embodiment, the action engine 3 can be packaged separately to decouple the responsibility boundaries of the rules engine 2, providing more windage action behavior that can be extended.
In the above embodiment, the real-time service wind control system provides multiple wind control scenes through the configuration layer 1, performs selection of optimal gradients of multiple strategies, multiple actions and multiple gradients on a single wind control scene through the rule engine 2, and extends and decouples the responsibility boundary of the rule engine 2 through the action engine 3, thereby implementing execution of multiple strategies, multiple actions and multiple gradients in a single scene.
Further, in the above-described embodiment, as shown in fig. 3, the rule engine 2 includes:
the receiving module 21 is used for receiving the wind control basic data;
the characteristic extraction module 22 is connected with the receiving module 21 and is used for extracting the characteristics of the accessed wind control basic data to obtain extracted characteristics;
the scene confirmation module 23 is connected with the feature extraction module 22 and is used for determining the wind control scene where the extracted features are located and the strategy rule table in the wind control scene;
the characteristic determining module 24 is connected with the scene confirming module 23, reads the characteristic conditions in the policy rule group according to the first priority reading sequence of the policy rule table and the second priority reading sequence of the policy rule group, obtains the characteristic conditions met by the extracted characteristics and the characteristic attributes of the characteristic conditions, and determines the intra-group gradient corresponding to the policy rule group according to the sum of the characteristic attributes corresponding to the characteristic conditions met by all the extracted characteristics in the policy rule group;
and the selecting module 25 is connected with the characteristic determining module 24 and is used for selecting the intra-group gradient which accords with the strategy combination mode from the intra-group gradients in all the strategy rule groups as the optimal gradient.
In the above embodiment, the receiving module 21 receives the wind control basic data, and may preprocess the wind control basic data, so as to facilitate the subsequent feature extraction module 22 to perform feature extraction on the preprocessed wind control basic data;
as a preferred embodiment, the receiving module 21 may include a script sub-engine and a query sub-engine;
the script sub-engine adopts script engine (script engine is a Java dynamic execution code) as the top interface class, the script engine provides an initialization method (may be initial, initial is a computer language) and an operation method (may be run, run is a code), other script languages can be expanded by the script engine, for example, osscript engine can obtain script file loading operation through OSS (Open Storage Service) system, the script file may be a dynamic script similar to Groovy, and can be pre-written and uploaded, and loaded into system memory when the container is started, so as to prevent frequent loading, and if the script rule changes, the script can be refreshed through load method;
the most important role of the script sub-engine is to convert the wind control basic data into an fact object of easyRules, wherein the number of the obtained data depends on the query sub-engine, a query engine method in the query sub-engine can be similar to an impala or mysql related query language and can be expanded, an execute method is run after the script object converts the wind control basic data to perform rule matching, a rule matching algorithm adopted by the subsequent characteristic determining module 24 can be realized by the rule engine, and as a preferred implementation mode, the rule engine can adopt a ScoreCard algorithm (scorscorescard).
In the above embodiment, the feature extraction module 22 performs feature extraction on the wind control basic data to obtain at least one extracted feature;
determining the wind control scene where each extracted feature is located through the scene confirmation module 23, wherein the number of the wind control scenes at this time can be multiple, and determining a policy rule table in each wind control scene;
determining a feature condition which is satisfied by the extraction features in the policy rule groups in the policy rule table in the single wind control scene by using a feature determination module 24, wherein the feature condition which is satisfied by the extraction features does not necessarily exist in each policy rule group;
it should be noted that, the scheme of multi-policy parallel execution may use ConditionRuleGroup as the implementation construction idea, but, due to the configurability and variability of the actions that need to be supported, it is not possible to implement a specific action in a rule action, this appears to be too customized, and in theory, a policy rule table in a single wind-controlled scenario could solve all the above problems if it were made sufficiently complex, but a great deal of customization needs are required, so the application distinguishes the relationship between the policy rule sets by reading the first priority of the policy rule sets in the policy rule table, if the policy rule groups have the same priority, the policy rule groups are considered to have an equal and independent relationship, if different priorities are set, the priorities are used as the order of policy execution or mutual exclusion basis, so that different policy rule sets can be concatenated by means of similar intermediate variables.
In the above embodiment, the feature determining module 24 sequentially reads the policy rule sets according to the first priority reading order of the policy rule table, sequentially reads the feature conditions in the currently read policy rule set according to the second priority reading order, and obtains the feature conditions meeting the extracted features according to the extracted features; at this time, there may be a plurality of feature conditions that meet the feature extraction in the policy rule group that is currently read, and therefore, it is possible to acquire the feature attribute of each feature condition that meets the feature extraction from the policy rule group that is currently read, and calculate the sum of the feature attributes of each feature condition that meets the feature extraction, and then acquire the intra-group gradient from the sum.
In the above-described embodiment, the feature attribute of the feature condition is the feature score, the sum of the feature attributes of the policy rule sets that meet the feature condition for extracting the feature is the gradient score, the gradient score corresponds to one gradient (for example, a gradient score of 10 to 19 corresponds to a gradient of 10, and a gradient score of 20 to 29 corresponds to a gradient of 10) and the introduction of the gradient is to solve the problem of the level of execution action corresponding to different feature conditions in a single policy rule set, so that the present embodiment can select the intra-group gradient that meets the policy combination pattern as the optimal gradient from among the intra-group gradients in all policy rule sets.
As a preferred embodiment, the policy combination mode specifically includes: selecting an intra-group gradient in a policy rule group corresponding to the highest priority in the first priority reading order as an optimal gradient; for example, the policy combination mode may adopt an activantrulegroup structure, that is, an intra-group gradient in a policy rule group corresponding to the highest priority in the first priority reading order is selected as an optimal gradient, and other policy rule groups in the policy rule table are ignored, and since the condition of the gradient is judged to be relatively single, only the gradient is used for comparing the sum of the characteristic attributes of each characteristic condition conforming to the extracted characteristic with the gradient score, and the mode of code generation is more suitable, the policy combination mode may select a construction method of inheriting defaultrule carried by easyles, so that the result attribute of the gradient comparison is filled for the subsequent action engine 3 to make action trigger reference and risk hit record; wherein, the strategy combination mode can be designed into a characteristic formula which can be identified by the MVEL.
It should be noted that, the gradient with higher priority using the ActivationRuleGroup is determined preferentially, and if hit, execution of other gradients is skipped, so that selection of the optimal gradient is only guaranteed;
in the above embodiment, the policy combination mode adopts an ActivationRuleGroup structure, where the ActivationRuleGroup structure is used to represent the optimal gradient of the policy rule group with the highest priority in the trigger policy rule table, and ignore other policy rule groups (XOR logic) in the policy rule table;
for example, XOR logic: a ^ b = (-a ^ b) (V-b);
wherein, a is used for representing the policy with the highest priority, and can be recorded as policy a;
b is used for representing a strategy with medium priority and can be marked as a strategy B;
executing the strategy a and the strategy b from high to low according to the priority, and if the fact needing to be judged hits the strategy a (reaches the gradient), not hitting the strategy b; and vice versa.
And if policy b is hit, it represents a miss for policy a.
-a represents other than a;
-b represents non-b.
In the above embodiment, the feature determining module reads the feature condition in the policy rule set according to the feature reading mode;
the feature reading mode specifically includes:
and setting a reading rule, and reading the characteristic conditions in the strategy rule group according to the second priority reading sequence of the strategy rule group when the extracted characteristics meet the reading rule.
In the above embodiment, the characteristic reading mode adopts a conditional rulegroup structure, where the conditional rulegroup structure is used to indicate that the reading rule with the highest priority evaluates to true, and then triggers a characteristic condition in the policy rule set;
for example, the policy rule table includes a first policy rule group, a second policy rule group, and a third policy rule group, where the first policy rule group, the second policy rule group, and the third policy rule group have high, medium, and low priorities, respectively;
at this time, when the extracted features only accord with the reading rules corresponding to the first strategy rule group and the third strategy rule group, the feature conditions in the first strategy rule group are preferentially selected for reading because the extracted features accord with the reading rules corresponding to the first strategy rule group with the highest priority;
at this time, when the extracted features only accord with the reading rules corresponding to the second policy rule group and the third policy rule group, the feature conditions in the second policy rule group are preferentially selected for reading because the extracted features do not accord with the reading rules corresponding to the first policy rule group with the highest priority.
As a preferred embodiment, the multi-strategy multi-gradient wind control system can be implemented by using the above strategy combination mode in combination with the complex rule of easy rules, as follows:
when a program is initialized, an easy rules object is constructed according to a policy rule table set in a configuration layer 1, wherein policy rule groups in the policy rule table are in parallel interaction, and feature conditions and feature attributes in each policy rule group are in a parallel structure;
and then, the policy rule groups are sorted according to a first priority reading order, the policy rule groups with higher priority are executed preferentially, if the policy rule groups with higher priority are the same, all the policy rule groups are executed, and the logic is judged by the result of the first rule executed by the policy with higher priority and is represented by a pseudo code:
MVELRule().when((currentStrategy.priority==lastStratege.priority()).then(continue)；
MVELRule().when((currentStrategy.priority!=lastStratege.priority()).then(return)。
further, in the above embodiment, as shown in fig. 3, the rule engine 2 further includes:
and the rule group reading module 26 is connected to the scene confirmation module 23, and determines a first priority reading order of the policy rule groups in the policy rule table according to the rule group attribute of each policy rule group in the policy rule table.
In the above embodiment, the rule group attribute is a priority score of the policy rule group.
Further, in the above-described embodiment, the characteristic condition is set with the condition number in correspondence, and the size of the condition number is set to the second priority reading order.
Further, in the above-described embodiment, the first condition atom of each characteristic condition in the policy rule set is the same, and the second condition atom is different.
As a preferred embodiment, one policy rule group in the policy rule table under the order scenario includes a first characteristic condition and a second characteristic condition:
when the first characteristic condition is: the amount of the order is more than or equal to 100;
when the second characteristic condition is: the amount of the order is more than or equal to 1000;
then the first conditional atom at this time: the amount of the order; second condition atom: "≧" the parameter behind.
Further, in the above embodiment, the configuration layer 1 includes the scene configuration module 11, which is used for setting the execution sequence of the wind control scenes when the real-time service wind control system is applied to a plurality of wind control scenes.
In the above embodiment, the scene configuration module 11 may perform a wind control hit decision for a plurality of wind control scenes.
Further, in the above embodiment, as shown in fig. 2, the access layer 4 and the data engine 5 are further included, and the access layer 4 is connected to the data engine 5;
the access layer 4 is connected with the configuration layer 1, receives and distributes the strategy rule table under each wind control scene to the data engine 5;
the data engine 5 preprocesses the policy rule table in each wind control scene, and stores the preprocessed policy rule table in each wind control scene for the rule engine 2 to call.
As a preferred embodiment, the real-time service wind control system is applied to a first wind control scene and a second wind control scene, where policy rule tables corresponding to the first wind control scene and the second wind control scene are respectively as follows:
TABLE 1
TABLE 2
Table 1 is a policy rule table corresponding to the first wind control scenario, and the policy rule table is denoted as a first policy rule table, where the first policy rule table includes two policy rule groups: the system comprises a first policy rule group and a second policy rule group, wherein the priorities of the first policy rule group and the second policy rule group are both 10, namely the priorities of the first policy rule group and the second policy rule group are consistent;
the first policy rule set includes two characteristic conditions: a first characteristic condition and a second characteristic condition, wherein the characteristic scores of the first characteristic condition and the second characteristic condition are both 10, and the gradient of the first policy rule set comprises 10 and 20;
the second policy rule set includes two characteristic conditions: a third characteristic condition and a fourth characteristic condition, wherein the characteristic scores of the third characteristic condition and the fourth characteristic condition are both 10, and the gradient of the second policy rule set comprises 10 and 20;
for example, when the extracted features both satisfy the first feature condition and the second feature condition of the first policy rule group, the feature determination module 24 in the rule engine 2 may obtain a feature attribute of the first feature condition (here, the feature score is 10), a feature attribute of the second feature condition (here, the feature score is 10), that is, the sum of the feature attributes at this time is 20 (10 +10= 20), and thus the intra-group gradient of the first policy rule group is 20;
when the extracted features only satisfy the third feature condition of the second policy rule group, the feature determining module 24 in the rule engine 2 obtains the feature attribute (here, the feature score is 10) of the third feature condition, that is, the sum of the feature attributes at this time is 10, so that the intra-group gradient of the second policy rule group is 10;
when the policy combination mode is adopted for the first wind control scenario, since the priorities of the first policy rule group and the second policy rule group are both 10, and the first policy rule group and the second policy rule group determine the intra-group gradient at this time, the intra-group gradient of the first policy rule group that is 20 and the intra-group gradient of the second policy rule group that is 10 are hit, and the hit result is shown in table 3 below:
TABLE 3
For example, when the extracted features both satisfy the first feature condition and the second feature condition of the first policy rule group, the feature determination module 24 in the rule engine 2 may obtain a feature attribute of the first feature condition (here, the feature score is 10), a feature attribute of the second feature condition (here, the feature score is 10), that is, the sum of the feature attributes at this time is 20 (10 +10= 20), and thus the intra-group gradient of the first policy rule group is 20;
when the extracted features do not satisfy the third feature condition and the fourth feature condition of the second policy rule group, the intra-group gradient of the second policy rule group is therefore 0;
when the policy combination mode is adopted for the first wind control scenario, since the priorities of the first policy rule group and the second policy rule group are both 10, and the first policy rule group and the second policy rule group determine the intra-group gradient at this time, the intra-group gradient of 20 in the first policy rule group and the intra-group gradient of 0 in the second policy rule group are hit (that is, no gradient is hit), and the hit result is shown in table 4 below:
TABLE 4
For example, table 2 is a policy rule table corresponding to the second wind control scenario, and is denoted as a second policy rule table, where the second policy rule table includes two policy rule groups: the priority of the third strategy rule group is 10, the priority of the fourth strategy rule group is 20, namely the priority of the third strategy rule group is lower than that of the fourth strategy rule group;
the third policy rule set includes two characteristic conditions: a fifth characteristic condition and a sixth characteristic condition, wherein the characteristic scores of the fifth characteristic condition and the sixth characteristic condition are respectively 10 and 20, and the gradient of the third policy rule set comprises 10 and 20;
the fourth policy rule set includes two characteristic conditions: a seventh characteristic condition and an eighth characteristic condition, wherein the characteristic scores of the seventh characteristic condition and the eighth characteristic condition are 10 and 20 respectively, and the gradient of the fourth policy rule set comprises 10 and 20;
for example, when the extracted features both satisfy the fifth feature condition and the sixth feature condition of the third policy rule group, the feature determination module 24 in the rule engine 2 may acquire the feature attribute of the fifth feature condition (here, the feature score is 10) and the feature attribute of the sixth feature condition (here, the feature score is 10), that is, the sum of the feature attributes at this time is 20 (10 +10= 20), so that the intra-group gradient of the third policy rule group is 20;
when the extracted features only satisfy the seventh feature condition of the fourth policy rule group, the feature determining module 24 in the rule engine 2 obtains the feature attribute (here, the feature score is 10) of the seventh feature condition, that is, the sum of the feature attributes at this time is 10, so that the intra-group gradient of the fourth policy rule group is 10;
when the policy combination mode is adopted for the second wind control scenario, since the priority of the third policy rule group is lower than the priority of the fourth policy rule group, the intra-group gradient of 10 of the fourth policy rule group is hit, and the hit result is shown in table 5 below:
TABLE 5
For example, when the extracted features both satisfy the fifth feature condition and the sixth feature condition of the third policy rule group, the feature determination module 24 in the rule engine 2 may acquire the feature attribute of the fifth feature condition (here, the feature score is 10) and the feature attribute of the sixth feature condition (here, the feature score is 10), that is, the sum of the feature attributes at this time is 20 (10 +10= 20), so that the intra-group gradient of the third policy rule group is 20;
when the extracted features do not satisfy the seventh feature condition and the eighth feature condition of the fourth policy rule group, there is no intra-group gradient of the fourth policy rule group;
when the policy combination pattern adopted for the second wind control scenario, and therefore the intra-group gradient of 20 for the third policy rule group, is hit, the hit result is as shown in table 6 below:
TABLE 6
The method for real-time business wind control based on the rule engine is further included, and as shown in fig. 5, the method comprises the following steps:
a strategy rule table is arranged aiming at each wind control scene, the strategy rule table comprises at least one strategy rule group, and the strategy rule group is provided with a rule group attribute, each characteristic condition and a characteristic attribute of the characteristic condition;
carrying out feature extraction on the accessed wind control basic data to obtain extracted features, determining a wind control scene according to the extracted features, setting a strategy combination mode, determining feature conditions in a strategy rule group which accords with the extracted features in a strategy rule table corresponding to the wind control scene, and determining an optimal gradient according to the strategy combination mode and all the feature conditions which accord with the extracted features;
and executing the executing action corresponding to the optimal gradient.
The specific implementation of the real-time service wind control method based on the rule engine is basically the same as that of the above real-time service wind control system based on the rule engine, and is not described herein again.
While the invention has been described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the invention.
Claims (9)
1. The utility model provides a real-time service wind control system based on rule engine which characterized in that, is applied to at least one wind control scene, and real-time service wind control system includes:
the configuration layer is provided with a strategy rule table aiming at each wind control scene, the strategy rule table comprises at least one strategy rule group, and the strategy rule group is provided with a rule group attribute, each characteristic condition and a characteristic attribute of the characteristic condition;
the rule engine is used for extracting features of the accessed wind control basic data to obtain extracted features, determining the wind control scene according to the extracted features, determining the feature conditions in the strategy rule group which accord with the extracted features in the strategy rule table corresponding to the wind control scene, and determining the optimal gradient according to a preset strategy combination mode and feature attributes corresponding to all the feature conditions which accord with the extracted features;
the action engine executes the execution action corresponding to the optimal gradient;
wherein the rules engine comprises:
the receiving module is used for receiving the wind control basic data;
the characteristic extraction module is connected with the receiving module and used for extracting the characteristics of the accessed wind control basic data to obtain the extracted characteristics;
the scene confirmation module is connected with the feature extraction module and used for determining the wind control scene where the extracted features are located and the strategy rule table in the wind control scene;
a feature determination module, connected to the scene confirmation module, configured to read the feature conditions in the policy rule set according to a first priority reading order of the policy rule table and a second priority reading order of the policy rule set, obtain the feature conditions that the extracted features satisfy and the feature attributes of the feature conditions, and determine an intra-group gradient corresponding to the policy rule set according to a sum of the feature attributes corresponding to the feature conditions that all the extracted features satisfy in the policy rule set;
and the selection module is connected with the characteristic determination module and used for selecting the intra-group gradient which accords with the strategy combination mode from the intra-group gradients in all the strategy rule groups as the optimal gradient.
2. The real-time business wind control system of a rules engine of claim 1, wherein said feature determination module reads said feature conditions in said set of policy rules according to a feature reading mode;
the feature reading mode specifically includes:
and setting a reading rule, and reading the characteristic conditions in the strategy rule group according to the second priority reading sequence of the strategy rule group when the extracted characteristics meet the reading rule.
3. The real-time business wind control system of a rules engine of claim 1, wherein the policy combination model specifically comprises: selecting the intra-group gradient in the set of policy rules corresponding to a highest priority in the first priority reading order as the optimal gradient.
4. The rules engine based real-time business wind control system of claim 1, wherein the rules engine further comprises:
and the rule group reading module is connected with the scene confirmation module and determines the first priority reading sequence of the strategy rule groups in the strategy rule table according to the rule group attribute of each strategy rule group in the strategy rule table.
5. The rules engine based real-time traffic wind control system according to claim 1, wherein the characteristic condition is set with a condition number, and the condition number is set to the second priority reading order.
6. The rules engine-based real-time traffic wind control system of claim 1, wherein a first condition atom of each of the characteristic conditions in the policy rule set is the same and a second condition atom is different.
7. The rules engine based real-time service wind control system according to claim 1, wherein the configuration layer comprises a scene configuration module for setting an execution sequence of the wind control scenes when the real-time service wind control system is applied to a plurality of the wind control scenes.
8. The rules engine based real-time traffic wind control system of claim 1, further comprising an access stratum and a data engine, said access stratum and said data engine connected;
the access layer is connected with the configuration layer, and receives and distributes the strategy rule table under each wind control scene to the data engine;
and the data engine preprocesses the strategy rule table under each wind control scene, and stores the preprocessed strategy rule table under each wind control scene for the rule engine to call.
9. A real-time business wind control method based on a rule engine is characterized by comprising the following steps:
setting a strategy rule table aiming at each wind control scene, wherein the strategy rule table comprises at least one strategy rule group, and the strategy rule group is provided with a rule group attribute, each characteristic condition and a characteristic attribute of the characteristic condition;
carrying out feature extraction on the accessed wind control basic data to obtain extracted features, determining the wind control scene according to the extracted features, setting a strategy combination mode, determining the feature conditions in the strategy rule group which accord with the extracted features in the strategy rule table corresponding to the wind control scene, and determining the optimal gradient according to the strategy combination mode and all the feature conditions which accord with the extracted features;
executing the executing action corresponding to the optimal gradient;
the method specifically includes the following steps of performing feature extraction on accessed wind control basic data to obtain extracted features, determining a wind control scene according to the extracted features, setting a strategy combination mode, determining feature conditions in a strategy rule group which accord with the extracted features in a strategy rule table corresponding to the wind control scene, and determining an optimal gradient according to the strategy combination mode and all the feature conditions which accord with the extracted features:
receiving the wind control basic data;
performing feature extraction on the accessed wind control basic data to obtain the extracted features;
determining the wind control scene where the extracted features are located and the strategy rule table in the wind control scene;
reading the feature conditions in the policy rule group according to a first priority reading sequence of the policy rule table and a second priority reading sequence of the policy rule group, obtaining the feature conditions met by the extracted features and the feature attributes of the feature conditions, and determining an intra-group gradient corresponding to the policy rule group according to the sum of the feature attributes corresponding to the feature conditions met by all the extracted features in the policy rule group;
selecting the intra-group gradient which accords with the strategy combination mode from the intra-group gradients in all the strategy rule groups as the optimal gradient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110135446.5A CN112463840B (en) | 2021-02-01 | 2021-02-01 | Real-time business wind control system and method based on rule engine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110135446.5A CN112463840B (en) | 2021-02-01 | 2021-02-01 | Real-time business wind control system and method based on rule engine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112463840A CN112463840A (en) | 2021-03-09 |
| CN112463840B true CN112463840B (en) | 2021-06-01 |
Family
ID=74802417
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110135446.5A Active CN112463840B (en) | 2021-02-01 | 2021-02-01 | Real-time business wind control system and method based on rule engine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112463840B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113076348B (en) * | 2021-04-13 | 2022-12-13 | 度小满科技(北京)有限公司 | Policy information management method, device, server and storage medium |
| CN113450067B (en) * | 2021-06-04 | 2022-10-04 | 杭州搜车数据科技有限公司 | Risk control method, device and system based on decision engine and electronic device |
| CN113608796A (en) * | 2021-06-30 | 2021-11-05 | 北京新氧科技有限公司 | Rule engine configuration and operation method and device, electronic equipment 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 |
| CN113409138A (en) * | 2021-07-19 | 2021-09-17 | 浙江孚临科技有限公司 | Risk control decision engine device |
| CN113888181A (en) * | 2021-10-25 | 2022-01-04 | 支付宝(杭州)信息技术有限公司 | Business processing and risk detection strategy system construction method, device and equipment |
| CN116012123B (en) * | 2023-03-27 | 2023-07-04 | 湖南三湘银行股份有限公司 | Wind control rule engine method and system based on Rete algorithm |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110298601A (en) * | 2019-07-05 | 2019-10-01 | 上海观安信息技术股份有限公司 | A kind of real time business air control system of rule-based engine |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10019677B2 (en) * | 2009-11-20 | 2018-07-10 | Alert Enterprise, Inc. | Active policy enforcement |
| CN107067157A (en) * | 2017-03-01 | 2017-08-18 | 北京奇艺世纪科技有限公司 | Business risk appraisal procedure, device and air control system |
| CN107424069B (en) * | 2017-08-17 | 2020-11-17 | 创新先进技术有限公司 | Wind control feature generation method, risk monitoring method and equipment |
| CN110443512A (en) * | 2019-08-09 | 2019-11-12 | 北京思维造物信息科技股份有限公司 | A kind of regulation engine and regulation engine implementation method |
| CN110674174B (en) * | 2019-09-24 | 2020-09-01 | 北京九章云极科技有限公司 | Data real-time processing method and data real-time processing system |
| CN111127196A (en) * | 2019-12-31 | 2020-05-08 | 中信百信银行股份有限公司 | Credit wind control characteristic variable management method and system |
-
2021
- 2021-02-01 CN CN202110135446.5A patent/CN112463840B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110298601A (en) * | 2019-07-05 | 2019-10-01 | 上海观安信息技术股份有限公司 | A kind of real time business air control system of rule-based engine |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112463840A (en) | 2021-03-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112463840B (en) | Real-time business wind control system and method based on rule engine | |
| CN112600810B (en) | Ether house phishing fraud detection method and device based on graph classification | |
| CN107798390B (en) | Training method and device of machine learning model and electronic equipment | |
| CN106453061B (en) | A kind of method and system identifying network fraudulent act | |
| CN110287479A (en) | Name entity recognition method, electronic device and storage medium | |
| CN109214914A (en) | A kind of loan information checking method and device based on communication open platform | |
| CN106933808A (en) | Article title generation method, device, equipment and medium based on artificial intelligence | |
| CN111737654B (en) | Infringement detection method and device based on block chain and electronic equipment | |
| US11256712B2 (en) | Rapid design, development, and reuse of blockchain environment and smart contracts | |
| CN112435137B (en) | Cheating information detection method and system based on community mining | |
| CN112560114B (en) | Method and device for calling intelligent contract | |
| CN111930366B (en) | Rule engine implementation method and system based on JIT real-time compilation | |
| CN109710255B (en) | Special effect processing method, special effect processing device, electronic device and storage medium | |
| CN110443512A (en) | A kind of regulation engine and regulation engine implementation method | |
| CN109918678B (en) | Method and device for identifying field meaning | |
| CN105763420A (en) | Automatic information reply method and device | |
| KR102295799B1 (en) | Metaverse service management server for managing the metaverse service to change interface design of metaverse environment based on payment record, and the operating method thereof | |
| CN110009216A (en) | Credit risk subscribers recognition methods and device | |
| CN107992545A (en) | A kind of searching method, device, terminal and readable storage medium storing program for executing | |
| CN110609908A (en) | Case serial-parallel method and device | |
| CN103902281A (en) | Providing a common virtual item repository in a virtual space | |
| CN104574275B (en) | A kind of method for merging textures during modeling rendering | |
| CN110263817B (en) | Risk grade classification method and device based on user account | |
| CN113362162A (en) | Wind control identification method and device based on network behavior data, electronic equipment and medium | |
| CN107657286A (en) | A kind of advertisement recognition method and computer-readable recording medium |
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 |