Dual-computing-engine scheduling method and apparatus, device, and storage medium

Abstract

The application discloses a double-computing-engine scheduling method and device, equipment and a storage medium, relates to the technical field of data processing, and the double-computing-engine scheduling method comprises the following steps: acquiring data source information related to a business scenario, and determining a computing index of the business scenario according to the data source information; selecting a decision module related to computing engine scheduling in an encapsulation library according to the computing index; splicing the decision module to obtain a decision tree model; acquiring to-be-computed data of the business scenario, and calling double computing engines to respectively perform first computing and second computing on the to-be-computed data on the basis of the decision tree model, to obtain first computing results and second computing results. Since the decision logic is stored in the encapsulation library in the form of a decision module, and the decision logic is solidified into a decision tree model independent of the computing engine, the decision logic is decoupled from the computing engine, so that repeated occupation of system resources is reduced.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a dual-computing-engine scheduling method, apparatus, device, and storage medium. Background Technology

[0002] Traditionally, credit approval was handled manually by loan officers. However, with the development of online and smaller-scale business, some credit approvals are now processed automatically by risk control systems, with risk strategy being a key module. Different products have different risk strategies, including common basic strategies as well as personalized strategies. Furthermore, as the environment and management requirements change, risk strategies need to be quickly adjusted to ensure effective risk management.

[0003] In traditional system architectures and collaboration models, decision logic is embedded in the system in the form of code. When different decision tasks are executed simultaneously, this approach requires multiple computing engines to independently parse and construct the execution structure of the decision logic, resulting in the duplication of system resources. Furthermore, when the decision logic changes, it is necessary to ensure that multiple engines update their respective execution structures. It is impossible to achieve consistent synchronization of the execution logic of all engines through the update of a single data source. Summary of the Invention

[0004] The main purpose of this application is to provide a dual computing engine scheduling method, apparatus, device and storage medium, which aims to solve the technical problem that the decision logic of the existing solution is solidified in the system in the form of code, and multiple engines need to be updated at the same time when the logic changes, resulting in repeated occupation of system resources.

[0005] To achieve the above objectives, this application proposes a dual-computing-engine scheduling method, which includes: Obtain data source information related to the business scenario, and determine the calculation indicators of the business scenario based on the data source information; Based on the computational metrics, the decision module involved in the scheduling of the computational engine is selected from the encapsulation library. The encapsulation library contains decision modules corresponding to different rules, strategies, and components. The decision module is used to judge, calculate, and combine the computational metrics. The decision modules are then concatenated to obtain a decision tree model; The system acquires the data to be calculated for the business scenario, and calls a dual computing engine to perform a first calculation and a second calculation on the data to be calculated based on the decision tree model, thereby obtaining a first calculation result and a second calculation result. The first calculation is the calculation of at least one decision path in the decision tree model, and the second calculation is the calculation of all decision modules and decision paths in the decision tree model.

[0006] In one embodiment, the calculation indicators include basic indicators and derived indicators; The step of determining the calculation indicators for the business scenario based on the data source information includes: The original data fields corresponding to the business scenario are determined based on the data source information; The basic metrics for the business scenario are determined based on the original data fields, wherein the basic metrics are used to characterize metrics that can be directly obtained from the data source; Based on the basic indicators and / or the existing derived indicators, new derived indicators for the business scenario are determined.

[0007] In one embodiment, the step of selecting the decision module involved in the computing engine scheduling from the encapsulation library based on the computing metrics includes: Based on the calculated metrics, decision modules are matched in the encapsulation library to determine the candidate decision modules involved in the scheduling of the computing engine; The candidate decision models are displayed through a visual interface, and the decision modules involved in the computing engine scheduling are determined in response to the user's selection operation based on the visual interface.

[0008] In one embodiment, the step of acquiring the data to be calculated for the business scenario, and calling a dual-computation engine to perform a first calculation and a second calculation on the data to be calculated based on the decision tree model, to obtain a first calculation result and a second calculation result, includes: In response to the user's business request trigger signal, determine that the current data source of the business scenario has been reached; The data to be calculated for the business scenario is determined based on the currently arrived data source; The decision tree model is parsed to obtain the dependencies and execution order of the decision nodes in the decision tree model, wherein the decision tree model includes at least one decision path, and the decision path includes at least one decision node; Based on the dependencies and the execution order, the dual computing engines are invoked to perform the first and second calculations on the data to be calculated, respectively, to obtain the first calculation result and the second calculation result.

[0009] In one embodiment, the step of invoking dual computing engines to perform a first calculation and a second calculation on the data to be calculated according to the dependency relationship and the execution order, and obtaining a first calculation result and a second calculation result, includes: When the data to be calculated satisfies the local computation conditions corresponding to the decision path in the decision tree model, the data to be calculated is input to each decision module in the decision path that satisfies the local computation conditions, and the dual computation engine is invoked to perform a first calculation on the decision module that has input the data to be calculated according to the dependency relationship and the execution order, so as to obtain a first calculation result; When the data to be calculated satisfies the global calculation conditions corresponding to the decision tree model, the data to be calculated is input into each of the decision modules in the decision tree model, and the dual calculation engine is invoked to perform a second calculation on the decision modules that have input the data to be calculated according to the dependency relationship and the execution order, so as to obtain a second calculation result.

[0010] In one embodiment, the step of concatenating the decision modules to obtain a decision tree model includes: Obtain the module type and judgment logic of each decision module; The execution order and dependencies between the decision modules are determined based on the module type and the judgment logic. Based on the execution order and the dependencies, the decision modules are assembled into a tree structure to obtain a decision tree model.

[0011] In one embodiment, after the step of splicing the decision modules to obtain the decision tree model, the method further includes: In response to the user-triggered rollback verification operation, a historical data time period selection window is displayed in the visualization interface; In response to the user's selection of a start time and an end time in the historical data period selection window, a verification data period is determined based on the start time and the end time; The time range constraint for generating verification data is generated based on the verification data period, and the full amount of data is extracted from the historical database as the verification dataset based on the time range constraint. Each historical data point in the validation dataset is input into the decision tree model for decision calculation, and the validation result corresponding to the validation dataset is obtained.

[0012] Furthermore, to achieve the above objectives, this application also proposes a dual-computing-engine scheduling device, which includes: The indicator management module is used to obtain data source information involved in the business scenario and determine the calculation indicators of the business scenario based on the data source information. The encapsulation management module is used to select the decision module involved in the scheduling of the computing engine from the encapsulation library according to the computing indicators. The encapsulation library is provided with different rules, strategies and decision modules corresponding to components. The decision module is used to judge, calculate and combine the computing indicators. The splicing module is used to splice the decision modules to obtain a decision tree model; The computation scheduling module is used to obtain the data to be computed in the business scenario, and call the dual computation engines to perform a first computation and a second computation on the data to be computed based on the decision tree model, so as to obtain a first computation result and a second computation result. The first computation is the computation of at least one decision path in the decision tree model, and the second computation is the computation of all decision modules and decision paths in the decision tree model.

[0013] In addition, to achieve the above objectives, this application also proposes a dual computing engine scheduling device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the dual computing engine scheduling method described above.

[0014] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the dual computing engine scheduling method described above.

[0015] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the dual computing engine scheduling method described above.

[0016] One or more technical solutions proposed in this application have at least the following technical effects: This application obtains data source information related to the business scenario and determines the calculation indicators of the business scenario based on the data source information. Based on the calculation indicators, it selects the decision modules involved in the calculation engine scheduling within the encapsulation library. The encapsulation library contains decision modules corresponding to different rules, strategies, and components. These decision modules are used to judge, calculate, and combine the calculation indicators. The decision modules are then concatenated to obtain a decision tree model. The application obtains the data to be calculated for the business scenario and, based on the decision tree model, calls two calculation engines to perform a first calculation and a second calculation on the data to be calculated, obtaining the first calculation result and the second calculation result. The first calculation involves calculating at least one decision path in the decision tree model, and the second calculation involves traversing all decision modules and decision paths in the decision tree model. By centrally storing the decision logic in the form of decision modules in the encapsulation library and solidifying the decision logic into a decision tree model independent of the calculation engine, the decision logic is decoupled from the calculation engine. This solves the technical problem in existing solutions where the decision logic is solidified in code within multiple engines, requiring modification of each engine when logic changes occur. It avoids redundant development, compilation, testing, and deployment when multiple engines undergo logic changes, thereby reducing the redundant occupation of system resources. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating an embodiment of the dual computing engine scheduling method of this application. Figure 2 A schematic diagram illustrating the generation of a decision tree model in one implementation of the dual-computation engine scheduling method of this application; Figure 3 This is a flowchart illustrating Embodiment 2 of the dual computing engine scheduling method of this application. Figure 4 This is a schematic diagram of the processing flow in one implementation of the dual computing engine scheduling method of this application; Figure 5 This is a flowchart illustrating Embodiment 3 of the dual computing engine scheduling method of this application; Figure 6 This is a schematic diagram of the module structure of the dual computing engine scheduling device in an embodiment of this application; Figure 7 This is a schematic diagram of the device structure of the hardware operating environment involved in the dual computing engine scheduling method in the embodiments of this application.

[0020] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0022] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0023] The main solution of this application embodiment is as follows: First, obtain data source information related to the business scenario and determine the calculation indicators of the business scenario based on the data source information. Second, select the decision modules involved in the calculation engine scheduling in the encapsulation library according to the calculation indicators. The encapsulation library contains decision modules corresponding to different rules, strategies, and components. The decision modules are used to judge, calculate, and combine the calculation indicators. Third, concatenate the decision modules to obtain a decision tree model. Fourth, obtain the data to be calculated in the business scenario and, based on the decision tree model, call the dual calculation engines to perform a first calculation and a second calculation on the data to be calculated, respectively, to obtain a first calculation result and a second calculation result. The first calculation is the calculation of at least one decision path in the decision tree model, and the second calculation is the calculation of all decision modules and decision paths in the decision tree model.

[0024] This application provides a solution where, because the decision logic is stored in a packaged library as decision modules, a structured decision tree model is generated by concatenating these modules. When multiple engines exist, each engine reads the same decision tree model data, eliminating the need for each engine to maintain an independent copy of the decision logic, thus reducing system resource consumption. When the decision logic changes, only the decision tree model data needs to be regenerated. Each engine can obtain the changed decision logic by reading the updated data, eliminating the need for individual updates and reducing resource consumption during the update process.

[0025] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a computer or server, or an electronic device or virtual device capable of performing the above functions. The following description uses a dual-computing-engine scheduling device (hereinafter referred to as the scheduling device) as an example to illustrate this embodiment and the subsequent embodiments.

[0026] Based on this, embodiments of this application provide a dual-computing-engine scheduling method, referring to... Figure 1 , Figure 1This is a flowchart illustrating an embodiment of the dual computing engine scheduling method of this application.

[0027] In this embodiment, the dual computing engine scheduling method includes steps S10 to S40: Step S10: Obtain the data source information involved in the business scenario, and determine the calculation indicators of the business scenario based on the data source information.

[0028] It is understood that the business scenarios in this application embodiment are those that require data analysis and decision-making, such as credit approval scenarios, risk control scenarios, and anti-fraud detection scenarios, and this application embodiment does not limit them.

[0029] It should be noted that the above data source information can be descriptive information about the data source that the computing engine needs to use when scheduling calculations in a business scenario. Specifically, it may include, but is not limited to, the identifier of the data source, the storage location of the data source, the data fields contained in the data source, and the data type of each data field.

[0030] For example, in a credit approval scenario, the data source may include credit data source, income data source, asset and liability data source, historical repayment record data source, etc. The data source information may further include specific field information from these data sources, such as credit score and number of overdue payments in the credit data source, and monthly income amount and annual income amount in the income data source.

[0031] It should be noted that the scheduling device in this application embodiment may include an indicator module. This indicator module can further abstract and encapsulate the data source information to form calculated indicators. The aforementioned calculated indicators may be indicators used to characterize specific business characteristics in a business scenario, such as the mean, standard deviation, and variance of monthly income, or the number of credit delinquencies and the income-to-debt ratio, etc. This application embodiment does not impose any limitations on these.

[0032] It should be understood that, in this embodiment, the corresponding data source information can be determined according to the business scenario requiring calculation, and the calculation indicators corresponding to the business scenario can be determined based on the data source information. In this way, the decision tree model can be pre-built, so that when the corresponding data source arrives (i.e., the data source corresponding to the business scenario is obtained), decision calculation can be performed based on the decision tree model.

[0033] In some embodiments of this application, the calculated indicators can be directly derived from the original value of a data field in the data source information, or they can be calculated based on one or more data fields, or they can be derived based on these original values ​​or calculated values ​​through further calculation. Specifically, the calculated indicators include basic indicators and derived indicators; the step of determining the calculated indicators of the business scenario based on the data source information includes: determining the original data field corresponding to the business scenario based on the data source information; determining the basic indicators of the business scenario based on the original data field, wherein the basic indicators are used to characterize indicators that can be directly obtained from the data source; and determining new derived indicators of the business scenario based on the basic indicators and / or the existing derived indicators.

[0034] It is understood that the aforementioned raw data fields are data fields that can be directly obtained from the data source, such as identity information, income information, etc., and this application embodiment does not impose any restrictions on them.

[0035] It should be noted that the aforementioned basic indicators are those obtained by directly mapping the original data fields, such as annual income, monthly income, gender, and credit rating. This application embodiment does not impose any limitations on these. The aforementioned derived indicators can be composite indicators obtained by processing basic indicators or existing derived indicators, such as average monthly income and debt ratio. This application embodiment does not impose any limitations on these. By supporting the nesting and reuse of multi-level indicators (basic indicators and derived indicators), the flexibility and maintainability of the indicator logic are improved.

[0036] Step S20: Select the decision module involved in the scheduling of the computing engine from the encapsulation library according to the computing metrics; The encapsulation library includes different rules, strategies, and decision modules corresponding to components. These decision modules are used to judge, calculate, and combine the calculation indicators.

[0037] like Figure 2 As shown, Figure 2 This diagram illustrates the generation of a decision tree model in one implementation of the dual-computation engine scheduling method of this application. In this implementation, the aforementioned encapsulation library is a pre-built functional component library. Different types of decision modules can be pre-configured in the encapsulation library, which enables the storage and management of these decision modules.

[0038] It should be noted that the aforementioned decision module encapsulates specific decision logic and can be used to judge, calculate, or combine computational indicators. In this embodiment, the decision module may include a rule module corresponding to a rule, a strategy module corresponding to a strategy, and a component module corresponding to a component.

[0039] It should be explained that the aforementioned rule module is the decision module determined based on risk control rules. It is used to perform threshold judgments or logical expression judgments on the calculated indicators and output the judgment results. The rule module in this embodiment can be configured with risk control rules as the smallest unit. Each rule module can independently define judgment logic, such as threshold judgments and logical expressions, to implement refined risk control strategies.

[0040] For example, a rule module for "whether the enterprise's tax credit rating is A" has the judgment logic of "enterprise tax credit rating == A". After inputting the calculation indicator "enterprise tax credit rating", the corresponding judgment result is output.

[0041] It should be noted that the aforementioned strategy module can be a decision-making unit composed of multiple rule modules or other strategy modules, used to aggregate multiple fine-grained rule modules or other strategy modules according to business semantics to obtain a business strategy. By combining and arranging rule modules or other strategy modules, a structured, reusable, and complete risk control strategy is formed, improving the flexibility and maintainability of strategy configuration. Compared to rule modules for fine-grained risk control review items, strategy modules correspond to the risk control review domain. For example, if risk control requires reviewing business risks, the entire business risk can form a strategy. This business risk can protect many review rules, such as business registration amount risk, business registration location risk, etc.

[0042] For example, a "Business Risk Strategy" module can be composed of a "Business Registration Amount Risk" rule module, a "Business Registration Location Risk" rule module, and a "Business Abnormality Risk" rule module, which can be used to comprehensively assess the business risk status of an enterprise.

[0043] It should be noted that the aforementioned component modules encapsulate general judgment or calculation logic, and may specifically include segmented components, grouped components, etc., which are not limited in this application embodiment. The component modules of this application embodiment can provide the ability to encapsulate general logic, thereby realizing the modular construction and reuse of complex business logic.

[0044] For example, in a risk control scenario, a "tax amount risk level segmentation component" can be pre-configured. The segmentation rules of this component are as follows: tax amount less than 100,000 yuan outputs "low risk", tax amount greater than or equal to 100,000 yuan and less than 500,000 yuan outputs "medium risk", tax amount greater than or equal to 500,000 yuan and less than 1,000,000 yuan outputs "high risk", and tax amount greater than or equal to 1,000,000 yuan outputs "extremely high risk".

[0045] It is understood that the decision-making module selected in this application embodiment may be based on the calculation indicators of different business scenarios and different business types, for example, it may be pushed in the form of a mapping table. This application embodiment does not limit this.

[0046] Step S30: The decision modules are spliced ​​together to obtain a decision tree model; Step S40: Obtain the data to be calculated for the business scenario, and call the dual calculation engine to perform a first calculation and a second calculation on the data to be calculated based on the decision tree model, to obtain a first calculation result and a second calculation result. The first calculation is the calculation of at least one decision path in the decision tree model, and the second calculation is the calculation of all decision modules and decision paths in the decision tree model.

[0047] It should be noted that the data to be calculated mentioned above is the data actually obtained from the business scenario, which can correspond to the data carried in a specific business request within that scenario. For example, in a corporate credit approval scenario, a specific loan application request might include the company name, unified social credit code, loan amount requested, corporate tax credit rating, years of establishment, credit score, and debt-to-asset ratio.

[0048] It is understood that the aforementioned dual computing engines are two independent computing engines that can be called separately. The dual computing engines in this application embodiment may include a first computing engine and a second computing engine. The first computing engine is used to calculate the local paths in the decision tree model to achieve real-time response to the business. The second computing engine is used to calculate the global paths and nodes in the decision tree model to achieve coverage calculation of the business.

[0049] It should be understood that the first calculation mentioned above is the calculation performed through the first calculation engine, and the second calculation is the calculation performed through the second calculation engine. During the calculation, each decision module in the decision tree model can be considered a decision node.

[0050] It should be noted that during the initial calculation, each decision node in the decision path can be calculated sequentially. If the requirements of a decision node are met, the calculation can continue to the next decision node; if the requirements are not met, the calculation for that decision path ends, and the initial calculation result is considered a failure. If all decision nodes on the decision path are satisfied, the initial calculation result is considered a success. This method enables real-time response to user commands.

[0051] For example, the scheduling device in this application embodiment can obtain the data to be calculated for a loan application request, which may include data such as the enterprise's tax credit rating being A, the establishment period being 4 years, and the credit rating being B. By calling the first calculation engine, path one (tax credit rating being A, establishment period being 3 years, and credit rating meeting the standard) is matched, and calculation is performed only on path one. It sequentially passes through the judgment node of "Is the tax credit rating A?" (result is yes), the judgment node of "Is the establishment period being 3 years?" (result is yes), and the judgment node of "Is the credit rating meeting the standard?" (result is yes), and finally reaches the "Pass" termination node, outputting the first calculation result as "Pass". If the judgment result of a certain node is "No", the process ends directly, and the first calculation result is output as "Fail".

[0052] It should be noted that during the second calculation, all decision paths and decision nodes in the decision tree model can be traversed. In this second calculation, the next decision node will be calculated regardless of whether its requirements are met. This approach enables global processing of the business scenario.

[0053] This application embodiment obtains data source information related to the business scenario and determines the calculation indicators of the business scenario based on the data source information. Based on the calculation indicators, it selects the decision modules involved in the calculation engine scheduling in the encapsulation library. The encapsulation library contains decision modules corresponding to different rules, strategies, and components. The decision modules are used to judge, calculate, and combine the calculation indicators. The decision modules are then concatenated to obtain a decision tree model. The data to be calculated in the business scenario is obtained, and based on the decision tree model, dual calculation engines are called to perform a first calculation and a second calculation on the data to be calculated, respectively, to obtain a first calculation result and a second calculation result. The first calculation is the calculation of at least one decision path in the decision tree model, and the second calculation is the calculation of all decision modules and decision paths in the decision tree model. By centrally storing the decision logic in the form of decision modules in the encapsulation library and solidifying the decision logic into a decision tree model independent of the calculation engine, the decision logic is decoupled from the calculation engine. This solves the technical problem in existing solutions where the decision logic is solidified in code within multiple engines, requiring modification of each engine when logic changes occur. It avoids redundant development, compilation, testing, and deployment when multiple engines undergo logic changes, thereby reducing the redundant occupation of system resources.

[0054] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in Embodiment 1 above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 3 , Figure 3 This is a flowchart illustrating Embodiment 2 of the dual computing engine scheduling method of this application.

[0055] like Figure 3 As shown in this embodiment, the step of selecting the decision module involved in the computing engine scheduling in the encapsulation library based on the computing metrics includes: Step S21: Based on the calculation indicators, perform decision module matching in the encapsulation library to determine the candidate decision modules involved in the scheduling of the computing engine; Step S22: Display the candidate decision model through a visual display interface, and respond to the user's selection operation based on the visual display interface to determine the decision module involved in the scheduling of the computing engine.

[0056] It should be noted that the aforementioned candidate decision modules are those selected from the encapsulation library through a matching operation, and these candidate decision modules have a correlation with the calculated indicators. Specifically, this correlation can be between the indicator types of the decision modules and the calculated indicators, or between the decision modules and the business scenarios corresponding to the calculated indicators, etc., and this application embodiment does not impose any limitations on this.

[0057] It is understandable that the aforementioned visualization interface can be an interactive interface that presents information to users in a graphical way. Through this visualization interface, candidate decision-making modules can be presented to users in a visual form.

[0058] In some embodiments of this application, the visual display interface of this application may include at least one of the following areas: a display area, a drag operation area, and a connection operation area.

[0059] It should be noted that the above-mentioned selection operation refers to the user's selection behavior on the candidate decision modules through the visual display interface, which may include clicking, dragging, checking, etc., and this application embodiment does not limit this. Through the selection operation, the module to be used in the end can be determined from the candidate decision modules, and the connection between the modules can be realized.

[0060] In some embodiments of this application, users can drag and drop candidate decision modules to the connection operation area, connecting the decision modules in the area. The connection can be made by user intervention or by connecting the decision modules sequentially from left to right and top to bottom according to their position in the connection operation area; this application does not limit this approach.

[0061] In some embodiments of this application, an architecture combining a highly efficient funnel-shaped computing mode and a coverage-based computing mode is adopted. The funnel-shaped computing mode corresponds to the first computing engine. Through this computing mode, the system can quickly process and generate decision results in a short time, ensuring the efficient operation of business processes. Simultaneously, the coverage-based computing mode of this application corresponds to the second computing engine, enabling the system to traverse all strategies and rules across the entire chain, providing comprehensive data support for business analysis. Specifically, it can be as follows... Figure 4 As shown, Figure 4 This is a schematic diagram of the processing flow in one implementation of the dual computing engine scheduling method of this application.

[0062] Reference Figure 4 In this embodiment of the application, users can realize the front-end decision tree arrangement through the drag and drop and connection functions of the decision module. When arranging the front-end decision tree, users can also perform visual attribute configuration, online debugging and historical verification.

[0063] In this embodiment of the application, during backend processing, the debugging device can perform graph parsing and pre-storage. Graph parsing may include decision tree dependency module parsing, reverse tracing-routing branch parsing-routing branch dependency data source parsing, calculation order parsing, etc., and this embodiment of the application does not limit this.

[0064] Understandably, decision tree dependency resolution involves resolving the dependencies between decision modules within the decision tree. For example, if rule B requires the judgment result of rule A before it can be executed, then rule B depends on rule A. By analyzing the connections in the graph and the input-output relationships between modules, the dependencies between each module can be determined.

[0065] It should be understood that, based on the determination of dependencies, the execution order between decision modules can be determined through computational order resolution. For example, when the output of module A is the input of module B, module A must be executed before module B.

[0066] It is understood that a decision tree model typically includes multiple routing branch nodes, which can divide the decision tree model into multiple different branch paths. This application does not impose restrictions on the judgment and identification methods of routing branch nodes; they can be selected according to the needs of actual applications.

[0067] It should be understood that reverse tracing – route branch parsing – starts from the terminal node of the decision tree model and traces back the route branches and decision modules. In this way, the complete sequence of nodes traversed by each decision path and the combination of data source conditions corresponding to each decision path can be determined. Thus, when the data source conditions corresponding to a certain decision path are met (i.e., the data source arrival status), the first computing engine is invoked by calling the service to calculate the decision path.

[0068] It is understood that through the above parsing operations, structured decision tree model data can be obtained, which can be pre-stored in a database or cache. The pre-stored content may include the configuration information of each node, the dependencies and execution order between nodes, the data source conditions corresponding to each branch path, and the path information obtained by reverse tracing, etc., which are not limited in this embodiment. At the same time, the first calculation result of the funnel-shaped scenario can be read, and a mapping relationship between the first calculation result and the decision tree model can be established.

[0069] It should be noted that, through the timed service, the scheduling device can periodically query whether the global data source conditions corresponding to the decision tree model are met, and when they are met, call the second computing engine to perform coverage scenario calculations to obtain the second calculation result.

[0070] In some embodiments of this application, the step of obtaining the data to be calculated for the business scenario and, based on the decision tree model, calling a dual computing engine to perform a first calculation and a second calculation on the data to be calculated to obtain a first calculation result and a second calculation result includes: responding to the user's business request trigger signal to determine the currently arrived data source of the business scenario; determining the data to be calculated for the business scenario based on the currently arrived data source; parsing the decision tree model to obtain the dependency relationship and execution order of the decision nodes in the decision tree model, wherein the decision tree model includes at least one decision path, and the decision path includes at least one decision node; calling the dual computing engine to perform a first calculation and a second calculation on the data to be calculated based on the dependency relationship and the execution order to obtain a first calculation result and a second calculation result.

[0071] It is understandable that when a specific business request arrives, a corresponding business request trigger signal can be generated. The first computing engine responds to the signal and starts processing the business request, thereby obtaining the first calculation result and realizing a real-time response to the business request.

[0072] In some embodiments of this application, the step of calling the dual computing engine to perform a first calculation and a second calculation on the data to be calculated according to the dependency relationship and the execution order, and obtaining a first calculation result and a second calculation result, includes: when the data to be calculated satisfies the local calculation conditions corresponding to the decision path in the decision tree model, inputting the data to be calculated to each decision module in the decision path that satisfies the local calculation conditions, and calling the dual computing engine to perform a first calculation on the decision module that has input the data to be calculated according to the dependency relationship and the execution order, to obtain a first calculation result; when the data to be calculated satisfies the global calculation conditions corresponding to the decision tree model, inputting the data to be calculated to each decision module in the decision tree model, and calling the dual computing engine to perform a second calculation on the decision module that has input the data to be calculated according to the dependency relationship and the execution order, to obtain a second calculation result.

[0073] It is understandable that the aforementioned local computation conditions, i.e., the data source conditions corresponding to the decision path, can be considered satisfied when the arrived data source meets the data source conditions required for the decision path computation. In this case, the decision path can be computed using the first computation engine. Similarly, the global computation conditions, i.e., the data source conditions corresponding to the decision tree model, can be considered satisfied when the arrived data source meets the data source conditions required for the computation of each decision node and decision path in the decision tree model. In this case, the decision tree model can be traversed and computed using the second computation engine.

[0074] In some embodiments of this application, the decision tree model can also be verified and backtracked. That is, after the step of assembling the decision modules to obtain the decision tree model, the method further includes: in response to a user-triggered backtracking verification operation, displaying a historical data period selection window in the visualization interface; in response to the user's selection of a start time and end time in the historical data period selection window, determining a verification data period based on the start time and end time; generating a time range constraint for the verification data based on the verification data period, and extracting all data from the historical database as a verification dataset based on the time range constraint; inputting each historical data item in the verification dataset into the decision tree model for decision calculation, and obtaining the verification result corresponding to the verification dataset.

[0075] It should be noted that the embodiments of this application can utilize historical real-world case data to quickly validate the newly configured decision tree model, assess whether its output results meet the expected business objectives, thereby ensuring the accuracy and practicality of the model logic. Specifically, rapid validation can be achieved through online debugging and backtracking validation. When clicking online debugging, input data is constructed for debugging, allowing the decision tree model to perform calculations and output model results. When performing backtracking validation, a full set of historical real-world data for a specific time period can be clicked in the pop-up window, and the system can trigger batch backtracking validation to verify the accuracy of the model using real data.

[0076] Understandably, the backtesting operation can be a feature used to batch validate generated decision tree models using historical data. By performing backtesting, the accuracy and stability of the decision tree model's decisions when processing real historical data can be verified.

[0077] It should be noted that the aforementioned historical data selection window can be displayed through a visual interface, allowing users to specify the time range of historical data to be backtracked for verification. This historical data selection window may include start time and end time selection controls, which are not limited in this embodiment.

[0078] Understandably, by determining the start and end times selected by the user, the end interval between the start and end times can be determined, i.e., the validation data period can be defined, thus generating a time range constraint for data extraction from the historical database. By using this time range constraint in the historical data, the validation dataset can be extracted.

[0079] This application embodiment determines candidate decision modules involved in computing engine scheduling by matching decision modules in the encapsulated library based on computational metrics. The candidate decision models are displayed through a visual interface, and the user's selection operation based on the visual interface is responded to, thus determining the decision modules involved in computing engine scheduling. In this application embodiment, the user interacts through the visual interface, re-executing the selection operation at different times according to changes in business needs, selecting different combinations of decision modules from the candidate decision modules. When the business scenario changes or the computational metrics are adjusted, decision module matching is re-performed, generating a new set of candidate decision modules, and then the decision modules are re-selected based on the new candidate set. This dynamic adjustment mechanism allows the decision modules involved in decision engine scheduling to be flexibly adjusted according to changes in business needs without redeveloping code or modifying underlying configuration files, improving the flexibility and response speed of the decision engine configuration.

[0080] Based on the first and / or second embodiments of this application, in the third embodiment of this application, the content that is the same as or similar to the first and / or second embodiments described above can be referred to the above description and will not be repeated hereafter. Based on this, please refer to... Figure 5 , Figure 5 This is a flowchart illustrating Embodiment 3 of the dual computing engine scheduling method of this application.

[0081] In this embodiment of the application, the step of splicing the decision modules to obtain a decision tree model includes: Step S31: Obtain the module type and judgment logic of each decision module; Step S32: Determine the execution order and dependencies between the decision modules based on the module type and the judgment logic; Step S33: According to the execution order and the dependency relationship, the decision modules are spliced ​​into a tree structure to obtain a decision tree model.

[0082] It is understood that the above module types can be used to characterize the type of a decision module, specifically including at least one of rule type, strategy type, and component type. The above judgment rules can be used to describe what judgment or calculation operation the decision module performs when it receives input data.

[0083] In a specific implementation, the scheduling device in this embodiment can read the configuration information of each decision module from the encapsulation library when performing splicing. This configuration information can record the module type and judgment logic description of various decision modules. Through this configuration information, the scheduling device can determine the module type and judgment logic of the decision module, thereby realizing the splicing of decision modules.

[0084] This application embodiment obtains the module type and judgment logic of each decision module; determines the execution order and dependencies between decision modules based on the module type and judgment logic; and concatenates the decision modules into a tree structure according to the execution order and dependencies to obtain a decision tree model. This application embodiment obtains the module type and judgment logic of each decision module, explicitly determines the execution order and dependencies between decision modules based on the module type and judgment logic, and embeds the above execution order and dependencies as structured data in the decision tree model, providing a data foundation for subsequent independent reading and calling by the dual computing engines.

[0085] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the dual computing engine scheduling method of this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0086] This application also provides a dual-computing-engine scheduling device; please refer to... Figure 6 , Figure 6 This is a schematic diagram of the module structure of the dual computing engine scheduling device according to an embodiment of this application. The dual computing engine scheduling device includes: The indicator management module 10 is used to obtain data source information involved in the business scenario and determine the calculation indicators of the business scenario based on the data source information. The encapsulation management module 20 is used to select the decision module involved in the scheduling of the computing engine from the encapsulation library according to the computing indicators. The encapsulation library is provided with different rules, strategies and decision modules corresponding to components. The decision module is used to judge, calculate and combine the computing indicators. The splicing module 30 is used to splice the decision modules to obtain a decision tree model; The calculation scheduling module 40 is used to obtain the data to be calculated in the business scenario, and call the dual calculation engine to perform a first calculation and a second calculation on the data to be calculated based on the decision tree model, so as to obtain a first calculation result and a second calculation result. The first calculation is the calculation of at least one decision path in the decision tree model, and the second calculation is the calculation of all decision modules and decision paths in the decision tree model.

[0087] The dual-computing engine scheduling device provided in this application, employing the dual-computing engine scheduling method in the above embodiments, can solve the technical problem that in existing solutions, the decision logic is fixed in the system in the form of code, requiring multiple engines to be updated simultaneously when the logic changes, leading to duplicate occupation of system resources. Compared with the prior art, the beneficial effects of the dual-computing engine scheduling device provided in this application are the same as those of the dual-computing engine scheduling method provided in the above embodiments, and other technical features in the dual-computing engine scheduling device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0088] This application provides a dual computing engine scheduling device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the dual computing engine scheduling method in the first embodiment described above.

[0089] The following is for reference. Figure 7The diagram illustrates a structural schematic suitable for implementing a dual-computing engine scheduling device according to embodiments of this application. The dual-computing engine scheduling device in embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), tablet computers, PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 7 The dual computing engine scheduling device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0090] like Figure 7 As shown, the dual-computing-engine scheduling device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 1002 or programs loaded from storage device 1003 into random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the dual-computing-engine scheduling device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the dual computing engine scheduling device to communicate wirelessly or wiredly with other devices to exchange data. While the figures show dual computing engine scheduling devices with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0091] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0092] The dual-computing engine scheduling device provided in this application, employing the dual-computing engine scheduling method in the above embodiments, can solve the technical problem that in existing solutions, the decision logic is fixed in the system in the form of code, requiring multiple engines to be updated simultaneously when the logic changes, leading to duplicate occupation of system resources. Compared with the prior art, the beneficial effects of the dual-computing engine scheduling device provided in this application are the same as those of the dual-computing engine scheduling method provided in the above embodiments, and other technical features in this dual-computing engine scheduling device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0093] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0094] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0095] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the dual computing engine scheduling method in the above embodiments.

[0096] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or flash memory, optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0097] The aforementioned computer-readable storage medium may be included in the dual-computing engine scheduling device; or it may exist independently and not be assembled into the dual-computing engine scheduling device.

[0098] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the dual-computing-engine scheduling device, cause the dual-computing-engine scheduling device to: Obtain data source information related to the business scenario, and determine the calculation indicators of the business scenario based on the data source information; Based on the computational metrics, the decision module involved in the scheduling of the computational engine is selected from the encapsulation library. The encapsulation library contains decision modules corresponding to different rules, strategies, and components. The decision module is used to judge, calculate, and combine the computational metrics. The decision modules are then concatenated to obtain a decision tree model; The system acquires the data to be calculated for the business scenario, and calls a dual computing engine to perform a first calculation and a second calculation on the data to be calculated based on the decision tree model, thereby obtaining a first calculation result and a second calculation result. The first calculation is the calculation of at least one decision path in the decision tree model, and the second calculation is the calculation of all decision modules and decision paths in the decision tree model.

[0099] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0100] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation that may be implemented in systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0101] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0102] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., computer programs) for executing the aforementioned dual-computing engine scheduling method. This solves the technical problem in existing solutions where the decision logic is fixed in the system in code form, requiring multiple engines to be updated simultaneously when the logic changes, leading to redundant consumption of system resources. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the dual-computing engine scheduling method provided in the above embodiments, and will not be repeated here.

[0103] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the dual computing engine scheduling method described above.

[0104] The computer program product provided in this application can solve the technical problem that the decision logic of existing solutions is fixed in the system in the form of code, and multiple engines need to be updated simultaneously when the logic changes, resulting in repeated occupation of system resources. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as the beneficial effects of the dual computing engine scheduling method provided in the above embodiments, and will not be repeated here.

[0105] The above description is only a part of the embodiments of this application and does not limit the scope of protection of this application. All equivalent structural transformations made under the technical concept of this application and using the content of this application specification and drawings, or direct / indirect applications in other related technical fields, are included in the scope of protection of this application.

Claims

1. A dual-computing-engine scheduling method, characterized in that, The method includes: Obtain data source information related to the business scenario, and determine the calculation indicators of the business scenario based on the data source information; Based on the computational metrics, the decision module involved in the scheduling of the computational engine is selected from the encapsulation library. The encapsulation library contains decision modules corresponding to different rules, strategies, and components. The decision module is used to judge, calculate, and combine the computational metrics. The decision modules are then concatenated to obtain a decision tree model; The system acquires the data to be calculated for the business scenario, and calls a dual computing engine to perform a first calculation and a second calculation on the data to be calculated based on the decision tree model, thereby obtaining a first calculation result and a second calculation result. The first calculation is the calculation of at least one decision path in the decision tree model, and the second calculation is the calculation of all decision modules and decision paths in the decision tree model.

2. The dual-computing engine scheduling method as described in claim 1, characterized in that, The calculation indicators include basic indicators and derived indicators; The step of determining the calculation indicators for the business scenario based on the data source information includes: The original data fields corresponding to the business scenario are determined based on the data source information; The basic metrics for the business scenario are determined based on the original data fields, wherein the basic metrics are used to characterize metrics that can be directly obtained from the data source; Based on the basic indicators and / or the existing derived indicators, new derived indicators for the business scenario are determined.

3. The dual-computing engine scheduling method as described in claim 1, characterized in that, The step of selecting the decision module involved in the computing engine scheduling in the encapsulation library based on the computing metrics includes: Based on the calculated metrics, decision modules are matched in the encapsulation library to determine the candidate decision modules involved in the scheduling of the computing engine; The candidate decision models are displayed through a visual interface, and the decision modules involved in the computing engine scheduling are determined in response to the user's selection operation based on the visual interface.

4. The dual-computing engine scheduling method as described in claim 3, characterized in that, The steps of obtaining the data to be calculated in the business scenario, and calling the dual computing engine to perform a first calculation and a second calculation on the data to be calculated based on the decision tree model, to obtain the first calculation result and the second calculation result, include: In response to the user's business request trigger signal, determine that the current data source of the business scenario has been reached; The data to be calculated for the business scenario is determined based on the currently arrived data source; The decision tree model is parsed to obtain the dependencies and execution order of the decision nodes in the decision tree model, wherein the decision tree model includes at least one decision path, and the decision path includes at least one decision node; Based on the dependencies and the execution order, the dual computing engines are invoked to perform the first and second calculations on the data to be calculated, respectively, to obtain the first calculation result and the second calculation result.

5. The dual-computing engine scheduling method as described in claim 4, characterized in that, The step of calling the dual computing engines to perform a first calculation and a second calculation on the data to be calculated according to the dependency relationship and the execution order, and obtaining the first calculation result and the second calculation result, includes: When the data to be calculated satisfies the local computation conditions corresponding to the decision path in the decision tree model, the data to be calculated is input to each decision module in the decision path that satisfies the local computation conditions, and the dual computation engine is invoked to perform a first calculation on the decision module that has input the data to be calculated according to the dependency relationship and the execution order, so as to obtain a first calculation result; When the data to be calculated satisfies the global calculation conditions corresponding to the decision tree model, the data to be calculated is input into each of the decision modules in the decision tree model, and the dual calculation engine is invoked to perform a second calculation on the decision modules that have input the data to be calculated according to the dependency relationship and the execution order, so as to obtain a second calculation result.

6. The dual-computing engine scheduling method as described in claim 1, characterized in that, The step of concatenating the decision modules to obtain the decision tree model includes: Obtain the module type and judgment logic of each decision module; The execution order and dependencies between the decision modules are determined based on the module type and the judgment logic. Based on the execution order and the dependencies, the decision modules are assembled into a tree structure to obtain a decision tree model.

7. The dual computing engine scheduling method as described in claim 3, characterized in that, After the step of concatenating the decision modules to obtain the decision tree model, the method further includes: In response to the user-triggered rollback verification operation, a historical data time period selection window is displayed in the visualization interface; In response to the user's selection of a start time and an end time in the historical data period selection window, a verification data period is determined based on the start time and the end time; The time range constraint for generating verification data is generated based on the verification data period, and the full amount of data is extracted from the historical database as the verification dataset based on the time range constraint. Each historical data point in the validation dataset is input into the decision tree model for decision calculation, and the validation result corresponding to the validation dataset is obtained.

8. A dual-computing-engine scheduling device, characterized in that, The dual computing engine scheduling device includes: The indicator management module is used to obtain data source information involved in the business scenario and determine the calculation indicators of the business scenario based on the data source information. The encapsulation management module is used to select the decision module involved in the scheduling of the computing engine from the encapsulation library according to the computing indicators. The encapsulation library is provided with different rules, strategies and decision modules corresponding to components. The decision module is used to judge, calculate and combine the computing indicators. The splicing module is used to splice the decision modules to obtain a decision tree model; The computation scheduling module is used to acquire the data to be computed in the business scenario, and call the dual computation engines to perform a first computation and a second computation on the data to be computed based on the decision tree model, so as to obtain a first computation result and a second computation result. The first computation is the computation of at least one decision path in the decision tree model, and the second computation is the computation of all decision modules and decision paths in the decision tree model.

9. A dual-computing-engine scheduling device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the dual computing engine scheduling method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the dual computing engine scheduling method as described in any one of claims 1 to 7.

Images