A decision management platform and method

Abstract

The present invention discloses a decision management platform and method. The method includes the following steps: initializing a pre-created decision and configuring basic attributes for the decision; configuring an operation mode for the decision according to the basic attributes of the decision; executing the decision according to the operation mode and saving the generated log data; managing the decision according to the log data. This kind of decision management platform and method provide full life cycle management of decisions, perfect unattended and analysis capabilities, facilitate the configuration of decisions, decision variables and versions, and maintain the operation of decisions; realize the unified management, maintenance and query of decision execution and decisions; solve the heterogeneous problem of decision engines, and at the same time are deeply integrated with decision engines, providing a set of standard decisions.

Description

Technical Field

[0001] The present invention relates to the technical field of decision management, and particularly to a decision management platform and method. Background Art

[0002] The essence of user growth is the growth of the value delivery chain, that is, in the process of delivering value to users, various chains of delivering information to users are recombined to add more effective value delivery chains. Enterprises urgently need enterprise-level product solutions with one-stop services to empower users to grow efficiently and agilely.

[0003] Therefore, a digital intelligent decision-making platform based on user data, facing the business department, and managing the full life cycle of management strategies is needed. Summary of the Invention

[0004] The purpose of the present invention is to provide a decision management platform and method to solve the defect in the prior art that there is no decision management platform based on user data and facing the business department when facing user growth.

[0005] The content of the present invention is further elaborated through the following two aspects:

[0006] In the first aspect, a decision management method is provided, and the method includes the following steps:

[0007] Initialize a pre-created decision and configure basic attributes for the decision;

[0008] Configure an operation mode for the decision according to the basic attributes of the decision;

[0009] Execute the decision according to the operation mode and save the generated log data;

[0010] Manage the decision according to the log data.

[0011] Combined with the first aspect, the method for initializing a pre-created decision and configuring basic attributes for the decision includes the following steps;

[0012] Obtain user requirements and create corresponding decisions according to the user requirements;

[0013] Configure basic attributes for the created decision, and at the same time, configure decision variables and decision versions.

[0014] Combined with the first aspect, the method for configuring an operation mode for the decision according to the basic attributes of the decision includes the following steps:

[0015] Configure skipping or not skipping dry-run operation for the decision according to the name, business scenario, and dry-run status in the basic attributes of the decision;

[0016] Configure skipping or not skipping gray-scale operation for the decision according to the name, business scenario, and gray-scale operation status in the basic attributes of the decision;

[0017] Configure skipping or not skipping formal operation for the decision according to the name and business scenario in the basic attributes of the decision.

[0018] Combined with the first aspect, the method for executing the decision according to the operation mode and saving the generated log data includes the following steps:

[0019] Execute the decision according to the operation mode of the decision, and record the log data generated during the operation of the decision;

[0020] The log data includes the decision empty-running status, empty-running time and duration, gray-scale operation status, gray-scale operation time and duration, and formal operation time and duration.

[0021] Combined with the first aspect, the method for managing the decision according to the log data includes:

[0022] Update the basic attributes and operation mode of the decision according to the obtained log data.

[0023] In the second aspect, a decision management platform is provided, including: a configuration mechanism and an execution mechanism;

[0024] The configuration mechanism includes:

[0025] A decision configuration module: used to configure the basic attributes related to the decision;

[0026] A variable configuration module: used to configure the variables used by the decision;

[0027] A version configuration module: used to configure the version of the decision and update the decision according to the version of the decision;

[0028] An operation mode configuration module: configure the corresponding operation mode for the decision according to the basic attributes of the decision, and manage the life cycle during the operation of the decision;

[0029] The execution mechanism includes:

[0030] A decision execution module: execute the decision according to the configured operation mode of the decision;

[0031] An execution log module: collect log data, and manage, maintain, and query the decision according to the log data.

[0032] Combined with the second aspect, the basic attributes related to the decision include one or more of the name, business scenario, empty-running operation status, gray-scale operation status, decision responsible person, decision notifier, decision engine, decision description, and input item acquisition timeout time of the decision.

[0033] In combination with the second aspect, the variables used in the decision-making include one or more of input variables, intermediate variables, and output variables.

[0034] In combination with the second aspect, the relevant attributes of the variables include one or more of the category, name, count, data type, whether it is in JSON format, and whether it is required to be transmitted of the variables.

[0035] In combination with the second aspect, the operation modes include one or more of dry-run operation, gray-scale operation, and formal operation.

[0036] The advantages of the present invention are as follows: This kind of decision management platform and method provides full life cycle management of decisions, perfect unattended and analysis capabilities, facilitates the configuration of decisions, decision variables and versions, and maintains the operation of decisions; realizes the unified management, maintenance, and query of decision execution and decisions; solves the heterogeneous problem of the decision engine, and at the same time is deeply integrated with the decision engine, providing a set of standard decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the life cycle process during the operation of the decision in the present invention.

[0038] Figure 2 It is a schematic diagram of the execution process of the decision in the present invention.

[0039] Figure 3 It is a schematic diagram of the decision management method in the present invention.

[0040] Among them: 1 - application system, 2 - decision scheduling module, 3 - decision engine, 4 - configuration center, 5 - cache center, 6 - token management, 7 - configuration management, 8 - decision log processing platform, 9 - decision log query platform. DETAILED DESCRIPTION OF THE INVENTION

[0041] In order to make the technical means, creative features, achieved purposes, and functions of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0042] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "center", "longitudinal", "transverse", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation to the present invention. In addition, the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first", "second", etc. may explicitly or implicitly include one or more of such features. In the description of the present invention, unless otherwise specified, the meaning of "a plurality of" is two or more.

[0043] In the description of the present invention, it should be noted that unless otherwise clearly specified and limited, the terms "mounted", "connected", "coupled" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood through specific situations.

[0044] The decision referred to in the present invention refers to the business strategy of the enterprise facing the users;

[0045] A decision management method, the method comprising the following steps:

[0046] S1: Initialize the pre-created decision and configure the basic attributes for the decision;

[0047] The specific method comprises the following steps;

[0048] Obtain user requirements and create corresponding decisions according to the user requirements;

[0049] Configure the basic attributes for the created decision, and at the same time, it also includes configuring decision variables and decision versions;

[0050] Among them, the basic attributes of the decision include the name of the decision, business scenario, dry-run status, gray-scale run status, decision responsible person, decision notifier, decision engine, decision description, and input item acquisition timeout;

[0051] The decision variables include input variables, intermediate variables, and output variables;

[0052] Among them, the input variables include the category, name, count, data type, whether it is in JSON format, and whether it is required;

[0053] The output variable is a variable table;

[0054] The version of the decision is a new decision created according to user requirements or for decision maintenance. When the new version of the decision runs, the old version of the decision needs to be taken offline to prevent conflicts during the operation of the two decisions;

[0055] S2: Configure the running mode for the decision according to the basic attributes of the decision;

[0056] The specific method includes the following steps:

[0057] Configure skip or no skip for dry run according to the name, business scenario, and dry run status in the basic attributes of the decision;

[0058] Configure skip or no skip for gray run according to the name, business scenario, and gray run status in the basic attributes of the decision;

[0059] Configure skip or no skip for formal run according to the name and business scenario in the basic attributes of the decision;

[0060] The dry run status includes two states: execute dry run and skip dry run. Dry run is a running state that only generates output values for data analysis during decision operation, and does not cause any impact on user data during the running process. The dry run also includes the form of dry run and the duration of dry run. The form of dry run includes three forms: by a certain proportion, random proportion, or by conditions. The strategy personnel will select the corresponding form and the time of dry run according to the running situation of the decision; skip dry run means not to execute the current dry run and directly jump to the next link;

[0061] The gray run status includes two states: execute gray run and skip gray run. Gray run is a running state that only executes part of the data during decision operation. The strategy personnel allocate part of the traffic to the new decision, and the remaining traffic runs in the original decision. Gray run includes the gray form and the gray priority. The gray form includes three forms: by a certain proportion, random proportion, or by conditions. The strategy personnel will select the corresponding form and the gray priority according to the running situation of the decision;

[0062] The formal run is the normal running step of the decision, but in some cases, the decision does not enter the normal running, but skips the formal run and directly ends the running;

[0063] S3: Execute the decision according to the running mode and save the generated log data;

[0064] The specific method includes the following steps:

[0065] Execute the decision according to the operation mode of the decision, and record the log data generated during the operation of the decision;

[0066] The log data includes the decision idle running state, idle running time and duration, gray-scale running state, gray-scale running time and duration, and formal running time and duration;

[0067] S4: Manage the decision according to the log data;

[0068] Update the basic attributes and operation mode of the decision according to the obtained log data;

[0069] The operation process in the technical solution of the present invention sequentially includes idle running, gray-scale running and formal running;

[0070] During the operation, the decision can choose to skip a certain operation process. For example, it can directly skip the idle running and enter the gray-scale running, then enter the formal running, and finally end the operation, or skip the idle running and enter the formal running, then end the operation, or skip the idle running and gray-scale running and directly enter the formal running, then end the operation, or directly skip the idle running, gray-scale running and formal running and directly end the operation;

[0071] Similarly, after the idle running is completed first, it can choose to skip the gray-scale running and enter the formal running, then end the operation, or skip the gray-scale running and then skip the formal running and directly end the operation, or sequentially enter the gray-scale running and formal running, then end the operation;

[0072] Adjust the idle running according to the state, idle running time and duration of the decision during the idle running, so that the idle running reaches the ideal operation effect, including adjusting whether the decision enters the idle running or skips the idle running. If it enters the idle running, then adjust the time node when the idle running enters and adjust the duration of the idle running. If it skips the idle running, then adjust the time node of skipping and the next operation mode entered after skipping;

[0073] Adjust the gray-scale running according to the state, gray-scale running time and duration of the decision during the gray-scale running, so that the gray-scale running reaches the ideal operation effect, including adjusting whether the decision enters the gray-scale running or skips the gray-scale running. If it enters the gray-scale running, then adjust the time node when the gray-scale running enters and adjust the duration of the gray-scale running. If it is necessary to skip the gray-scale running, then adjust the time node of skipping and the next operation mode entered after skipping.

[0074] A decision management platform, for product managers or strategists, includes: a configuration mechanism and an execution mechanism;

[0075] The configuration mechanism: used to configure decisions, variables and versions of decisions and maintain the operation of decisions;

[0076] The execution mechanism: used for the execution of decisions and the unified management, maintenance and query of decisions.

[0077] In this embodiment, the decision management platform is based on an assembly factory. The assembly factory is oriented to product managers and is used to configure the operation platform for operation personnel. It includes trigger objects, trigger methods, gateways, events, activities, processes, policies, actions, approvals that can be used under the operation platform, encapsulates the differences of each operation center of the enterprise, and manages the object configurations used by each operation center. It can be foreseen that the managed objects will also be continuously enriched and improved with the diversified operation of the enterprise. Based on the product positioning of the assembly factory, the technical implementation focuses on high scalability and high security, and ensures non-intrusiveness in business. The assembly factory provides the underlying technical support for the decision management platform.

[0078] The data storage of the assembly factory adopts a vertical table structure to ensure the scalability of the system; uses multi-tenant technology to logically isolate the data of each operation platform, ensuring data security while making full use of storage resources. The server uses the responsibility chain pattern to dynamically concatenate the factory configurations and flexibly delete and modify them; uses aspect technology to achieve zero intrusion of business logic. The client uses a dynamic form engine to customize and display the differentiated page elements and function buttons of the platform according to the requirements of each operation platform;

[0079] The inputs of the assembly factory are: platform name, first-level menu, business person in charge, R & D person in charge, test person in charge, trigger object, event, action, list page retrieval field, list page field

[0080] The output is: xx operation platform;

[0081] In this embodiment, the configuration mechanism includes a decision configuration module, a variable configuration module, a version configuration module and a running mode configuration module;

[0082] The decision configuration module: used to configure the basic attributes related to decisions;

[0083] The decision configuration module is oriented to policy personnel or product managers. Policy personnel or product managers configure the basic attributes related to decisions according to user requirements;

[0084] The basic attributes related to decisions include the name of the decision, business scenario, whether it can skip empty running, whether it can skip gray scale, decision person in charge, decision notifier, decision engine, decision description, and input item acquisition timeout. The decision code is generated using a distributed ID, and strict restrictions are placed on the access to decisions to improve the quality of decisions and provide data support for subsequent unattended operation and analysis and troubleshooting;

[0085] The inputs of the decision configuration module are: decision name, business scenario, whether it is possible to skip dry running, whether it is possible to skip gray scale, decision responsible person, decision notifier, decision engine, decision description, input item acquisition timeout;

[0086] The output is: decision;

[0087] The variable configuration module: is used to configure the variables used in the decision;

[0088] The variable configuration module is for policy personnel and configures the variables used in the decision;

[0089] The variables used in the decision include input variables, intermediate variables, and output variables;

[0090] Among them, the relevant attributes of the variable include the category, name, count, data type, whether it is in JSON format, and whether it is required;

[0091] The inputs of the variable configuration module are: input variables, intermediate variables, output variables, variable category, name, count, data type, whether it is in JSON format, whether it is required;

[0092] The output is: variable table;

[0093] The version configuration module: is used to configure the version of the decision and update the decision according to the version of the decision;

[0094] The version configuration module is for policy personnel or product managers. According to user requirements and the usage of the decision, it modifies or creates a new decision version, and completes the version update of the decision. It uses the distributed lock technology to ensure version increment, is deeply integrated with the decision engine, provides a system automatic version creation channel, improves configuration efficiency, and optimizes the user experience;

[0095] The inputs of the version configuration module are: decision flow, decision step, rule set, rule, decision type, project, decision version, policy;

[0096] The output is: decision version;

[0097] The operation mode configuration module: configures the corresponding operation mode for the decision according to the basic attributes of the decision and manages the life cycle during the decision operation;

[0098] The inputs of the operation mode configuration module are: dry running (dry running time, by proportion, random proportion, by condition), gray scale operation (gray scale priority, by proportion, random proportion, by condition), formal operation;

[0099] The output is: decision version operation;

[0100] The operation modes of the decision include dry running state, gray scale operation state, and formal operation state;

[0101] The life cycle of a decision includes the creation of the decision, the initialization of the decision, the dry-run state of the decision, the gray-scale running state of the decision, the formal running state of the decision, and the end of the decision running.

[0102] As Figure 1 shown, it is a schematic diagram of the decision running mode process in the present invention. The running mode includes initialization, dry-run, gray-scale running, and formal running.

[0103] The life cycle of a decision includes the following situations:

[0104] 1) After the decision is created, configure the running mode of the decision. First, initialize the decision. After initialization, the form and time of the dry-run can be set for the decision. After the dry-run settings are completed, then set the form and time of the gray-scale running. Then the decision enters the formal running. After the running ends, take the decision offline, that is, the running ends.

[0105] 2) After the decision is created, configure the running mode of the decision. First, initialize the decision. After initialization, the form and time of the dry-run can be set for the decision. After the dry-run settings are completed, skip the gray-scale and directly enter the formal running. After the running ends, take the decision offline, that is, the running ends.

[0106] 3) After the decision is created, configure the running mode of the decision. First, initialize the decision. After initialization, the form and time of the dry-run can be set for the decision. After the dry-run settings are completed, directly take the decision offline, that is, the running ends.

[0107] 4) After the decision is created, configure the running mode of the decision. First, initialize the decision. After initialization, the form and time of the dry-run can be set for the decision. After the dry-run settings are completed, then set the form and time of the gray-scale running. Then skip the formal running and directly take the decision offline, that is, the running ends.

[0108] 5) After the decision is created, configure the running mode of the decision. First, initialize the decision. Then the decision skips the dry-run and enters the gray-scale running. Then set the form and time of the gray-scale running. Then the decision enters the formal running. After the running ends, take the decision offline, that is, the running ends.

[0109] 6) After the decision is created, configure the running mode of the decision. First, initialize the decision. Then the decision skips the dry-run and enters the gray-scale running. Then set the form and time of the gray-scale running. Then skip the formal running. After the running ends, take the decision offline, that is, the running ends.

[0110] 7) After the decision is created, run the decision configuration mode. First, initialize the decision, then the decision skips dry-run and gray-scale run and directly enters the formal run. After the run ends, take the decision offline, that is, the run ends.

[0111] 8) After the decision is created, run the decision configuration mode. First, initialize the decision. After initialization, skip dry-run, gray-scale run, and formal run, and directly take the decision offline to end the run.

[0112] In the above running mode, the dry-run state includes two states: executing dry-run and skipping dry-run. Dry-run is a running state where only output values for data analysis are generated during decision running, and no impact is caused to user data during the running process. The dry-run also includes the form of dry-run and the duration of dry-run. The form of dry-run includes three forms: by a certain ratio, random ratio, or by conditions. The strategy personnel will select the corresponding form and the dry-run time according to the running situation of the decision; skipping dry-run means not executing the current dry-run and directly jumping to the next link.

[0113] The gray-scale run state includes two states: executing gray-scale run and skipping gray-scale run. Gray-scale run is a running state where only part of the data is executed during decision running. The strategy personnel allocate part of the traffic to the new decision, and the remaining traffic runs in the original decision. Gray-scale run includes the gray-scale form and the gray-scale priority. The gray-scale form includes three forms: by a certain ratio, random ratio, or by conditions. The strategy personnel will select the corresponding form and the gray-scale priority according to the running situation of the decision.

[0114] In this embodiment, the execution mechanism includes a decision execution module and an execution log module.

[0115] The decision execution module: Execute the decision according to the running mode configured by the decision.

[0116] The execution log module: Collect log data and manage, maintain, and query the decision according to the log data.

[0117] The execution of the decision has extremely high requirements for high performance and high availability. The execution logic is encapsulated into executable components such as the decision scheduling module 2 and decentralized and deployed to each used application such as the application system 1 to ensure system fault tolerance and make full use of the hardware resources of each integrated application.

[0118] As Figure 2 shown, it is a schematic diagram of the decision execution process in the present invention. The processing part of the execution log module is centrally deployed and processed, including the decision log processing platform 8 and the decision log query platform 9. The logs are collected centrally to facilitate the unified management and maintenance of the decision by the decision platform, and the decision data stakeholders can use it with confidence and convenience.

[0119] The execution logic of the decision-making includes three parts: obtaining configuration data, fetching variable values, and invoking the decision-making. The decision-making configuration data is cached at multiple levels. One copy of the data is cached in the configuration center, such as Configuration Center 4, and another copy is cached in the local memory of the application. The data is preferentially fetched from the memory with the highest acquisition efficiency. The acquisition of the credential token is triggered by a scheduled task, and the obtained token is cached in the Redis middleware cluster, such as Cache Center 5, taking full advantage of the high availability and high performance characteristics of the Redis cluster itself. The variable value fetching adopts a batch fetching method, and technologies such as circuit breaker, degradation, and caching are used to improve the execution efficiency and availability.

[0120] The execution results and traces of the decision-making and variables are collected through a data collection method that combines streaming and batch processing. The detailed decision-making execution collection uses the gauge and counter combined metric types of the time series database Promethus to ensure the real-time, integrity, and accuracy of the data, laying a solid foundation for real-time decision-making analysis.

[0121] The specific implementation process is as follows:

[0122] The decision-making configuration system provided by the business in the decision-making management platform includes token management 6 and configuration management 7. The configurable decision-making basic information includes: the business scenario to which the decision belongs, the decision-making person in charge, the alarm notifier, the decision-making engine type, as well as the variable timeout time and the default value for fallback after timeout.

[0123] Based on these configurations, the underlying implementation of the decision-making management platform performs decision-making permission control, precise push of monitoring alarm information, and customization specifications and constraints for the decision-making execution state. After configuring the decision-making basic information, the business can choose to generate the corresponding decision-making version and variable table for version management. The decision-making version is presented in the form of a form, and the business can operate, delete, and view it; the variable table is displayed in a tree form, and the hierarchical structure can better reflect the association relationship between variables. The business can select attributes such as whether the variable is required to be transmitted and whether it is used for JSON serialization based on the business status to ensure that the decision-making call meets the actual business needs. After the version is generated, the business sets the running strategy of the version as needed. If it is desired that the decision-making does not affect the online environment, the decision-making running strategy can be set to dry run, and the dry run ratio and dry run trigger conditions can be set as needed; similarly, if the business hopes to release the decision-making in a gray manner, the decision-making running strategy can be set to gray, and the gray ratio and gray trigger conditions can be set as needed. After the decision-making is verified to be correct, the decision-making is officially launched to serve the company's business. If problems or anomalies are found only during the official launch stage of the decision-making, the decision-making management platform also provides a one-key rollback function to ensure the smooth operation of the company's business.

[0124] Decision calls have extremely high requirements for high performance and high availability. During decision configuration, the decision management platform will perform a large amount of call preprocessing work to ensure the high performance and high availability of decision calls. The configurations of decision basic configuration, version, variable table, and operation policy of the business will be cached in the distributed configuration center 4. The configuration center 4 is based on system memory operations, with extremely high performance, and is deployed in a multi-cluster distributed manner to achieve the purpose of high availability. When making a call, the business system can directly obtain the configuration metadata from the configuration center. When the business sets the decision version operation policy, the decision management platform obtains the call credential, that is, the token, through the decision engine 3. And the call credential is cached in the cache center 5, and the survival period of the credential is set. When the business system makes a decision call, it does not need to perceive the interaction functions unrelated to the core business and only needs to be responsible for the call, which simplifies the access cost of the business system. The credential is pre-obtained at the configuration end, and a multi-level cache (Redis and system memory secondary cache) storage mechanism is adopted to improve the performance and availability of credential acquisition.

[0125] After completing the decision configuration work, the decision enters the execution stage. The decision call adheres to the idea of decentralization and provides a set of high-performance scheduling components including the decision call module 2, making full use of the resource configurations of each application system. The application system 1 integrates the decision call module 2. The component call module is responsible for obtaining the decision metadata configuration from the configuration center 4, parsing the configuration, and completing the definition and declaration of the entire scheduling module call rules; obtaining the input items for decision calculation from the data fetching platform, using the Trie construction method, trading space for time, and integrating the decision request variables into a request message in Trie structure; obtaining the call credential from the cache center 5; routing the obtained credential and the generated request message to the decision engine 3 of the corresponding service node according to the configuration to obtain the decision calculation result and complete the call work. The scheduling component collection module will collect the entire execution path data to the decision log processing platform through the stream-batch integrated method. The decision log processing platform 8 performs big data cleaning and merging, stores the massive data in a high-performance OLAP database, and supplies it for real-time business analysis and calculation. A unified unattended rule and data export query platform, that is, the decision log query platform 9, is provided to facilitate real-time business analysis and monitoring by the business and form a business closed-loop.

[0126] As is known by technical common sense, the present invention can be implemented by other embodiments that do not deviate from its spiritual essence or essential features. Therefore, the above-disclosed embodiments are illustrative in all aspects and are not exclusive. All changes within the scope of the present invention or within the scope equivalent to the present invention are encompassed by the present invention.

Claims

1. A decision-making management method, characterized in that, The method includes the following steps: Initialize a pre-created decision and configure basic attributes for the decision; According to the basic attributes of the decision, configure the running mode for the decision, including: Configure whether to skip or not skip the dry-run based on the name, business scenario, and dry-run status in the basic attributes of the decision; Configure whether to skip or not skip the gray-scale run based on the name, business scenario, and gray-scale run status in the basic attributes of the decision; Configure whether to skip or not skip the formal run based on the name and business scenario in the basic attributes of the decision; Execute the decision according to the running mode and save the generated log data, where the log data includes the dry-run status of the decision, dry-run time and duration, gray-scale run status, gray-scale run time and duration, and formal run time and duration; Manage the decision according to the log data.

2. The decision management method according to claim 1, characterized in that: The method for initializing a pre-created decision and configuring basic attributes for the decision includes the following steps; Obtain user requirements and create corresponding decisions according to the user requirements; Configure basic attributes for the created decision, and also include configuring decision variables and decision versions.

3. The decision management method according to claim 2, characterized in that: The method for executing the decision according to the running mode and saving the generated log data includes the following steps: Execute the decision according to the running mode of the decision and record the log data generated during the running of the decision.

4. The decision management method according to claim 3, wherein: The method for managing the decision according to the log data includes: Update the basic attributes and running mode of the decision according to the obtained log data.

5. A decision management platform, characterized in that, Include: Configure the organization and the executing agency; The configuration organization includes: Decision configuration module: used to configure basic attributes related to the decision; Variable configuration module: used to configure variables used by the decision; Version configuration module: used to configure the version of the decision and update the decision according to the version of the decision; Running mode configuration module: configure the corresponding running mode for the decision according to the basic attributes of the decision and manage the life cycle during the running of the decision; Configure whether to skip or not skip the dry-run based on the name, business scenario, and dry-run status in the basic attributes of the decision; Configure whether to skip or not skip the gray-scale run based on the name, business scenario, and gray-scale run status in the basic attributes of the decision; Configure whether to skip or not skip the formal run based on the name and business scenario in the basic attributes of the decision; The executing agency includes: Decision execution module: execute the decision according to the configured running mode of the decision; Execution log module: collect log data, manage, maintain, and query the decision according to the log data, where the log data includes the dry-run status of the decision, dry-run time and duration, gray-scale run status, gray-scale run time and duration, and formal run time and duration.

6. The decision management platform according to claim 5, characterized in that: The basic attributes of the decision include one or more of the name, business scenario, dry-run status, gray-scale run status, decision responsible person, decision notifier, decision engine, decision description, and input item acquisition timeout time of the decision.

7. A decision management platform according to claim 6, characterized in that: The variables used by the decision include one or more of input variables, intermediate variables, and output variables.

8. A decision management platform according to claim 7, characterized in that: The relevant attributes of the variable include one or more of the variable's category, name, count, data type, whether it is in JSON format, and whether it is required to be passed.

9. A decision management platform according to claim 8, characterized in that: The operating modes include one or more of idle running, gray-scale running, and formal running.

Images