CN115994744A - Intelligent decision management system, operation task issuing method, equipment and storage medium - Google Patents

Abstract

The application relates to an intelligent decision management system, an operation task issuing method, computer equipment and a storage medium. The system comprises: the user interaction module is used for providing a corresponding configuration page for the user according to the configuration instruction sent by the user, so that the user can upload relevant data of operation business; the rule verification module is used for verifying the rule data; the flow experiment verification module is used for creating a flow experiment and verifying the flow experiment; the strategy verification module is used for creating an algorithm strategy and verifying the algorithm strategy; the model verification module is used for loading an algorithm model corresponding to the model data uploaded by the user and verifying the algorithm model; the task verification module is used for creating an operation task and verifying the operation task; and the issuing module is used for issuing the target operation task to all service instances of the target access party. The method and the device can simplify a long-chain cooperative mode which originally requires the joint participation and repeated communication of multiple parties, and remarkably save the communication cost and the development cost.

Description

Intelligent decision management system, operation task issuing method, equipment and storage medium

Technical Field

The present disclosure relates to the field of intelligent operation technologies, and in particular, to an intelligent decision management system, an operation task publishing method, a computer device, and a storage medium.

Background

At present, the internet industry gradually adopts algorithmic and intelligent to replace part of work (namely intelligent decision management) carried out by traditional manual operation in operation business, so that the intelligent operation is more and more valued, but the support for the intelligent operation to be efficient and stable is not made under the internet technology architecture system at present.

From a technical perspective, currently, developing intelligent operations and intelligent decision services at least faces the following problems:

1. the development of one intelligent operation activity relates to the communication of operators, algorithm engineers, technical engineers and testers, wherein one simple routine business iteration needs a plurality of times;

2. the development period is long, the operation activities of multiple cities in the whole country are related, the development code quantity is thousands of lines, and the whole period takes month as a unit;

3. the related personnel are wide, the workload is large, the period is long, and the stability problem is necessarily caused, namely, bug is easy to occur after online operation, and adverse effects or consequences are caused;

4. compared with the traditional manual operation mode, the intelligent operation method has the advantages of intelligent operation of the algorithm, but is inferior to the traditional operation in timeliness and usability.

Disclosure of Invention

Aiming at the defects or shortcomings, the intelligent decision management system, the operation task issuing method, the computer equipment and the storage medium are provided, and the embodiment of the application can simplify a long-chain cooperation mode which originally needs the joint participation and repeated communication of multiple parties, and remarkably saves the communication cost and the development cost.

The present application provides, according to a first aspect, an intelligent decision management system, which in one embodiment comprises:

the user interaction module is used for providing a corresponding configuration page for the user according to a configuration instruction sent by the user, so that the user can upload operation service related data through the configuration page, and the operation service related data comprises feature data, rule data, strategy data, flow experiment configuration data, model data and task data;

the rule verification module is used for verifying the rule data uploaded by the user;

the flow experiment verification module is used for creating a flow experiment according to the flow experiment configuration data uploaded by the user and verifying the flow experiment;

the strategy verification module is used for creating an algorithm strategy according to the strategy data uploaded by the user and verifying the algorithm strategy;

the model verification module is used for loading an algorithm model corresponding to the model data uploaded by the user and verifying the algorithm model;

the task verification module is used for creating an operation task according to task data uploaded by a user and verifying the operation task;

and the issuing module is used for determining a target operation task and a target access party according to an issuing instruction sent by the user and issuing the target operation task to all service instances of the target access party.

In some embodiments, the model data includes model files; the model verification module is specifically used for: and uploading the model file in the model data uploaded by the user to a designated file server to create a corresponding algorithm model, and verifying the created algorithm model through a simulation environment service provided by an access party corresponding to the model data.

In some embodiments, the task verification module is specifically configured to: and creating an operation task according to task data uploaded by a user, and verifying the created operation task through a simulation environment service provided by an access party corresponding to the operation task.

In some embodiments, the task verification module is specifically configured to, when verifying an operation task: acquiring rule data, a flow experiment, an algorithm model and an algorithm policy associated with an operation task, copying the currently acquired rule data, flow experiment, algorithm model and algorithm policy as a task snapshot associated with the operation task, and verifying the operation task according to the task snapshot;

accordingly, when the publishing module publishes the target operation task to all service instances of the target access party, the publishing module is specifically configured to: and acquiring a target task snapshot associated with the target operation task, and publishing the target operation task to all service instances of the target access party according to the target task snapshot.

In some embodiments, the task verification module is specifically configured to, when verifying an operation task according to a task snapshot:

checking whether rule data, flow experiments, algorithm models and algorithm strategies contained in the task snapshot are valid or not;

when the detection results are valid, acquiring instance data of all valid instances of the access party, which are arranged in the simulation environment, from the service registration and service discovery cluster according to the unique identifier of the access party corresponding to the task snapshot;

the related data of the operation task is sent to each effective instance through the operation task pushing interface;

and when receiving the successful notification of task loading returned by all the effective examples through the callback interface, determining that the operation task passes the verification.

In some embodiments, the task verification module is further configured to return a prompt message to the user when the detection result is not valid, or when a task loading failure notification returned by any valid instance is received.

In some embodiments, the above system further comprises:

the feature module is used for managing all feature data;

the rule module is used for managing all rule data;

the flow experiment module is used for managing all flow experiment configuration data;

the model module is used for managing all model data;

the policy module is used for managing all policy data;

and the task module is used for managing all task data.

The present application provides, according to a second aspect, an intelligent decision management method, in one embodiment, the method comprising:

receiving a configuration instruction sent by a user, and providing a configuration page corresponding to the configuration instruction for the user, so that the user can upload operation service related data through the configuration page, wherein the operation service related data comprises feature data, rule data, strategy data, flow experiment configuration data, model data and task data;

verifying rule data, creating a flow experiment and verifying the flow experiment according to the flow experiment configuration data, creating an algorithm strategy and verifying the algorithm strategy according to the strategy data, loading an algorithm model corresponding to the model data and verifying the algorithm model, and creating an operation task and verifying the operation task according to the task data;

and receiving an issuing instruction sent by a user, determining a target operation task and a target access party according to the issuing instruction, and issuing the target operation task to all service instances of the target access party.

According to a third aspect, the present application provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of an embodiment of any of the methods described above when the computer program is executed.

According to a fourth aspect, the present application provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of an embodiment of any of the methods described above.

In the above embodiments of the present application, an intelligent decision management system is provided, which is modularized and can support free configuration of users, so that the system can efficiently and stably support development of intelligent operation services. Specifically, the system provides a technical document to a user, namely an algorithm engineer, and the algorithm engineer uploads operation service related data (including feature data, rule data, policy data, flow experiment configuration data, model data and task data) to the system according to a data format specified in the document, after the system receives the operation service related data, the system verifies the validity of the operation service related data, and after the system passes the validity verification, a series of operations such as verifying rules, creating flow experiments, verifying flow experiments, creating algorithm policies, verifying algorithm policies, loading algorithm models, verifying algorithm models, creating operation tasks and verifying operation tasks are automatically executed. In the whole business process, if the system detects a problem, the system informs related personnel, and after the operation task is successfully verified, algorithm engineers and operators are informed, if no extra operation of the operators is used, the operation task is finished online at a designated time, namely, the system is really applied to online business. Through the system, at least the long-chain cooperation mode which originally requires the common participation and repeated communication of multiple parties can be simplified, the communication cost and the development cost are obviously saved, and further, the development, new service access and landing period are effectively shortened.

Drawings

FIG. 1 is a block diagram of an intelligent decision management system in one embodiment;

FIG. 2 is a flow chart of an operation task issuing method according to an embodiment;

FIG. 3 is a schematic diagram of a configuration flow of an algorithm model in one embodiment;

FIG. 4 is a schematic diagram of a verification process of service correctness in one embodiment;

FIG. 5 is a schematic diagram of a verification process of technical correctness in one embodiment;

fig. 6 is an internal structural diagram of a computer device in one embodiment.

Detailed Description

In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.

In the prior art, an algorithm engineer generates a model file according to service requirements and communicates with a technical engineer to determine requirements, then the technical engineer finishes the development of related service codes, the technical engineer needs to synchronize the developed service codes to a tester, and the algorithm model file and the service codes are stored in a code warehouse together. And testing the service codes by a tester, if a problem is found in the testing process, repeatedly communicating with an algorithm engineer, a technical engineer and an operator to determine whether the service codes need to be modified, and if the service codes need to be modified, the technical engineer is required to upload the model files provided by the algorithm engineer to a passcode warehouse again to restart the service flows and the like. In the prior art, a complete business process involves multiple roles, so that the communication cost is high, the efficiency is low, the stability is poor, the time for landing the business is long, and the time is usually required to be several weeks. In addition, the existing scheme has at least the following problems that the model file and the operation strategy file are placed in the same code warehouse to cause serious coupling with the technical service code file (namely service code), the model file and the algorithm strategy have certain timeliness, a plurality of useless codes and files can be generated after long-time iteration, obviously, the too many useless codes and files occupy system resources, and hidden danger is brought to the service code.

In order to solve at least one of the above problems, the present application provides an intelligent decision management system (which may be simply referred to as a system). The system is modularized and can support free configuration of users, so that the system can efficiently and stably support development of intelligent operation business. Specifically, the system provides a technical document to a user, namely an algorithm engineer, and the algorithm engineer uploads operation service related data (including feature data, rule data, policy data, flow experiment configuration data, model data and task data) to the system according to a data format specified in the document, after the system receives the operation service related data, the system verifies the validity of the operation service related data, and after the system passes the validity verification, a series of operations such as verifying rules, creating flow experiments, verifying flow experiments, creating algorithm policies, verifying algorithm policies, loading algorithm models, verifying algorithm models, creating operation tasks and verifying operation tasks are automatically executed. In the whole business process, if the system detects a problem, the system informs related personnel, and after the operation task is successfully verified, algorithm engineers and operators are informed, if no extra operation of the operators is used, the operation task is finished online at a designated time, namely, the system is really applied to online business.

The intelligent decision management system provided by the embodiment of the application is based on the modularized design idea, a user can freely match an algorithm model, an algorithm strategy, a flow experiment and the like which accord with related service scenes, the customization requirement is realized, the engineering development cost required by the landing of an operation service is reduced, and the manpower of a technical engineer is released; the standardization and autonomy of the operation service landing mode also changes the original human factor variable, so that the stability of the service is greatly improved, meanwhile, after the system is configured with the operation task, the system can automatically verify the correctness and reliability of the algorithm task through the simulation environment provided by the relevant access party, the input cost of a tester can be directly saved, the relevant test and verification process can be automatically completed by means of an algorithm engineer and the operator, the single-node service closed loop is realized, the long-chain cooperation mode which originally needs the common participation and repeated communication of multiple parties is simplified, the communication cost and the development cost are remarkably saved, and the development, new service access and landing period is effectively shortened.

The above system will be described in detail below.

In some embodiments, the intelligent decision management system includes modules as shown in FIG. 1, each of which is described below.

The user interaction module 110 is configured to provide a corresponding configuration page for the user according to a configuration instruction sent by the user, so that the user can upload operation service related data through the configuration page, where the operation service related data includes feature data, rule data, policy data, flow experiment configuration data, model data and task data.

The feature data is a feature that a user needs to use by configuring an access related service party (may be simply referred to as an access party) configured by a page, and is a generic term of variables and functions in a rule and/or policy executing process, and may specifically be contents such as functions, service variables, and the like. Further, the feature data may be specifically divided into basic features and service features, where the service features are collected and calculated by the service system of the access party, and then provided to the system.

After the user configures the feature data, the configured feature data can be used as various processing logics to form rule data meeting the service requirement. The rule data is formed by the combination of the logic operation of the characteristics and the arrangement of various functions, and can return the true or false expression, which can be applied to the flow experiment.

The flow experiment configuration data refers to data related to flow experiments configured by a user, wherein the data can comprise an experiment unique identifier of the flow experiment, each flow group included in the flow experiment, and the like. Regarding traffic experiments, it is logic that determines which traffic to use to process a received request according to pre-configured rules in a traffic scenario. The traffic experiment in this embodiment may include two modes, i.e. feature configuration and rule configuration, and the first mode is to select a specific service feature, such as a city ID, and determine which traffic packet in the traffic experiment is hit (multiple traffic packets are set in the traffic experiment) according to a specific city ID value in service request data received by the service. The second way is that each flow packet has its own rule definition, making rule decision according to all data in the service request, and hitting the flow packet if it is decided as true.

The model data includes user-configured model base data and model files. The model base data may include data for model names, model file types, and the like. The system can load the algorithm model in the model file through the model file. The algorithm model is a key embodiment of algorithm and intelligence, and is a computational logic structure of the algorithm for processing characteristics and input.

Policy data user configured data related to algorithm policies. A user can configure a single algorithm policy through a configuration page, and specify the names, the affiliated services, the policy priorities and other contents of the algorithm policy. The algorithm strategy is pseudo codes used for processing different scene types in the operation business and is matched with the algorithm model to form the core of the intelligent operation business.

After the user has configured the feature data, rule data, policy data, flow experiment configuration data and model data, the task data can be configured. The task data is data related to an operation task, and contains previously configured traffic experiments, a model list (including one or more algorithm models) and a strategy list (including one or more algorithm strategies) under each traffic group in the traffic experiments, and the like. The system creates an operation task according to the task data, wherein the operation task is a complete data structure output by the system to the access party. The task data comprises an access party unique identifier configured by a user, and an access party corresponding to an operation task created according to the task data can be determined through the access party unique identifier.

In some embodiments, the above system may further comprise the following modules, namely: the feature module is used for managing all feature data in the system; the rule module is used for managing all rule data in the system; the flow experiment module is used for managing all flow experiment configuration data in the system; the model module is used for managing all model data in the system; the policy module is used for managing all policy data in the system; and the task module is used for managing all task data in the system.

In this embodiment, the data related to the features, rules, flow experiments, algorithm policies, and operation tasks are stored in a dedicated database of the system, and the authorized user can perform operations such as creation, modification, query, and deletion by calling the corresponding module (such as the feature module) through the system at any time. In this embodiment, various data are stored in a dedicated database respectively, rather than being stored in a code warehouse uniformly, so that serious coupling is not caused, and in this embodiment, by opening data operation rights of various modules to related users, the users can delete useless codes and files, thereby releasing system resources occupied by the useless codes and files, and avoiding hidden danger of the codes and files on service codes.

The rule verification module 120 is configured to verify rule data uploaded by a user.

The flow experiment verification module 130 is configured to create a flow experiment according to the flow experiment configuration data uploaded by the user, and verify the flow experiment.

The policy verification module 140 is configured to create an algorithm policy according to policy data uploaded by a user, and verify the algorithm policy.

In some embodiments, the model verification module 150 is configured to upload a model file in the model data uploaded by the user to a designated file server to create a corresponding algorithm model, and verify the created algorithm model through a simulation environment service provided by an access party corresponding to the model data.

The model verification module 150 is configured to load an algorithm model corresponding to the model data uploaded by the user, and verify the algorithm model.

The task verification module 160 is configured to create an operation task according to task data uploaded by a user, and verify the operation task.

In some embodiments, the task verification module 160 is configured to create an operation task according to task data uploaded by a user, and verify the created operation task through a simulation environment service provided by an access party corresponding to the operation task.

In some embodiments, the task verification module is specifically configured to, when verifying an operation task: and acquiring rule data, a flow experiment, an algorithm model and an algorithm policy associated with the operation task, copying the currently acquired rule data, flow experiment, algorithm model and algorithm policy as a task snapshot associated with the operation task, and verifying the operation task according to the task snapshot. Accordingly, when the publishing module publishes the target operation task to all service instances of the target access party, the publishing module is specifically configured to: and acquiring a target task snapshot associated with the target operation task, and publishing the target operation task to all service instances of the target access party according to the target task snapshot.

In this embodiment, in order to avoid that key data such as rules, flow experiments, algorithm policies used by an operation task are modified by other users in the process of verifying the operation task and issuing the operation task, so that data seen from a system is different from data actually used on line, the system uses a data snapshot policy, specifically, at the moment of verifying the operation task, the system background automatically reads all the rules, flow experiments, algorithm models and algorithm policies used by the operation task, and then copies a complete definition of the data read at the moment for independent storage, and each data is associated with the operation task. After the policy is used, no matter how long the time span of verification and release of the whole operation task is, other users can freely modify the rules, the policies and other data in the operation task without influence (understandably, the data in the snapshot cannot be modified), the modified rules and policies are configured into the operation task, and the user does not need to worry about tampering with the created operation task, and certainly, the user does not need to worry about tampering with the operation task.

Further, when the task verification module verifies an operation task according to the task snapshot, the task verification module is specifically configured to: checking whether rule data, flow experiments, algorithm models and algorithm strategies contained in the task snapshot are valid or not; when the detection results are valid, acquiring instance data of all valid instances of the access party, which are arranged in the simulation environment, from the service registration and service discovery cluster according to the unique identifier of the access party corresponding to the task snapshot; the related data of the operation task is sent to each effective instance through the operation task pushing interface; and when receiving the successful notification of task loading returned by all the effective examples through the callback interface, determining that the operation task passes the verification.

And the task verification module is further used for returning prompt information to the user when the detection result is not valid or when the task loading failure notification returned by any valid instance is received.

And the issuing module 170 is configured to determine a target operation task and a target access party according to an issuing instruction sent by the user, and issue the target operation task to all service instances of the target access party.

Based on the intelligent decision management system provided by any one of the foregoing embodiments or implementations, the present application further provides an operation task publishing method, as shown in fig. 2, including the following steps:

s110: receiving a configuration instruction sent by a user, and providing a configuration page corresponding to the configuration instruction for the user, so that the user can upload operation service related data through the configuration page, wherein the operation service related data comprises feature data, rule data, strategy data, flow experiment configuration data, model data and task data;

s120: verifying rule data, creating a flow experiment and verifying the flow experiment according to the flow experiment configuration data, creating an algorithm strategy and verifying the algorithm strategy according to the strategy data, loading an algorithm model corresponding to the model data and verifying the algorithm model, and creating an operation task and verifying the operation task according to the task data;

s130: and receiving an issuing instruction sent by a user, determining a target operation task and a target access party according to the issuing instruction, and issuing the target operation task to all service instances of the target access party.

The following describes a process of issuing an operation task using an intelligent decision management system.

1. The user firstly configures the features needed to be used in the service of the access party, and the features configured by the user can be simple features such as city ID and the like, a whole input request structure body and/or a built-in service processing function of the access party; the user may do various processing logic based on the configured features to compose rule data that meets business needs.

2. The user selects the previously configured features or rules to configure the flow experiment, which includes a plurality of flow packets.

3. The user configures an algorithm model and an algorithm policy for each traffic packet.

Regarding the algorithm model, a user only needs to fill in basic information (such as model name, model file type and the like) of the model through a web page, then uploads a corresponding model file to a designated file server through a system, and after the model file passes through the data verification of the file server, an algorithm model is successfully created; after the algorithm model is successfully established, model verification (verification through simulation environment service of an access party) is needed, and the algorithm model can be used in subsequent operation tasks after the model verification;

regarding algorithm policies, a user may configure the algorithm policies in a web page input box using a pseudo code form, for example, configure a single algorithm policy, specify the names of the policies, the services to which the policies belong, the policy priorities, and other contents, and may upload the algorithm policies to an intelligent decision management platform in the form of an upload file from which the algorithm policies are automatically extracted by the system. The algorithm strategy is also required to be verified by using the simulation environment service of the access party, and the algorithm strategy can be used in the operation task after the verification is passed.

4. After the steps 1-3 are performed, the final operation task (also referred to as an algorithm task) can be configured after the user configures the features, the rule data, the flow experiment, the algorithm model and the algorithm policy. The complete operation tasks comprise flow experiments, a model list, a strategy list and the like under each flow group in the flow experiments.

5. After the configuration of the operation task is completed, the correctness (including technical correctness and business correctness) of the operation task needs to be checked. The correctness of the operation task can be checked by issuing simulation operation on the system, wherein the technical correctness of the operation task can be verified through a service callback result of the simulation environment provided by the access party, and the service correctness of the operation task can be verified through detailed output of logs of the simulation environment.

6. After the operation task passes the verification of the simulation environment, the user can instruct the intelligent decision management system to issue the operation task to each online instance of the access party. The system firstly searches all effective examples of related services of the access party from the service registration center, and then pushes related data of operation tasks to all the effective examples through the agreed interface. When each effective instance is loaded to complete the operation task, the complete intelligent operation service is completed once and falls to the ground (namely, the operation task is successfully issued).

More specifically, the configuration process of the algorithm model may be that basic information of the algorithm model, such as an access party and related services served by the model, a model name, a model type, and the like, is first configured, and then the following steps are performed, that is:

(1) The user selects a local model file to upload through a front-end web page, the system receives the model file through a rear-end interface and then caches the model file to the local of the system, then judges whether the model file is a newly built model according to model basic information and MD5 values of the model file, if the model file is not a new model, a model folder path which is established before in a designated file server (such as an OSS file server, OSS refers to Object Storage Service, namely object storage service) is found, and if the model file is a new model folder is established in the designated file server according to naming rules;

(2) The background of the system automatically uploads the model file to a designated path (namely a model folder) in the file server determined in the step (1);

(3) If the uploading is successful, the file server returns the URL (Uniform Resource Locator ) of the file, the system stores the received URL of the file and other related information in a database (such as a MYSQL database), specifically, a new database record is created in the database as a data storage, and the user can perform data operations such as inquiring, modifying, deleting and the like later according to the requirement;

(4) If the database is successfully operated, the interface returns a success prompt to the front-end page of the system, and if the operation is failed, the interface returns an error prompt and displays the error prompt on the web page, so that a user can process the error according to the related prompt;

(5) If the uploading of the model file to the designated file server fails, the back-end interface of the system returns a corresponding error prompt to the front-end page to inform the user.

Illustratively, steps (1) - (5) above may be described with reference to FIG. 3.

Further, the following describes the process of the system automatically verifying the business correctness and technical correctness of the operational tasks.

The verification process of the service correctness is as follows:

1-1, firstly checking whether each element contained in an operation task, such as characteristics, rules, algorithm models, algorithm strategies, flow experiments and the like, is effective, namely whether the current state of each element is a verified state, if the current state of one element is illegal, prompting a user for error reasons, and helping the user to locate the element with the problem;

1-2, if the states of all the constituent elements in the operation task are valid in the previous step, acquiring all the valid examples (including specific IP addresses and network ports of all the examples) of the related business of the access party from the unified service registration and service discovery cluster according to the unique identifier of the access party corresponding to the operation task, and simultaneously setting an initial state of all the examples to write the initial state into a task release detail database table;

1-3, calling an operation task pushing interface for all the examples to send the related data of the operation task to the simulation environment of the access party;

1-4, if the interface call in the last step fails, retrying for a certain number of times according to the algorithm for avoiding network congestion;

1-5, if all the above examples push data successfully (in general, the simulation environment has only one example), updating a task release detail database table (task table) representing that the task release detail database table is currently in a waiting access side callback state;

1-6, simultaneously, if the data pushing is successful, the back end interface of the system returns successful information to the front end page so as to prompt the user to wait for verification at present; if the push data is unsuccessful, a rollback instruction is issued for each instance of the just pushed data.

The verification process of the technical correctness is as follows:

2-1, each instance of the related service of the access party starts to load data after receiving the related data of the operation task, and calls a callback interface appointed in advance after completing the loading to inform the loading state of the instance of the system;

2-2, if the notification returned by a certain effective instance through the callback interface is successful in loading, updating the instance state corresponding to the instance in the task release detail table, and verifying the technical correctness of the instance;

2-3, if only a successful callback for a partial instance is received, continue waiting. If a success callback of all the examples is received, the technical correctness is considered to be verified, the state of a task table (namely the task release detail table) in the database is updated, meanwhile, the user is reminded that the technical verification of the operation task is successful, and the user can check data such as related service logs, service reports and the like according to the actual requirements of the user to verify the service correctness of the operation task again;

2-4, if the notification returned by the instance is a loading failure, the system immediately ends logic waiting for the callback state after receiving the callback notification, displays the callback state in the web page according to the error reason returned by the callback, and simultaneously updates the task table state in the database to remind the user of verification failure so as to enable the user to carry out relevant modification according to the specific reason;

for example, the above-mentioned verification process of service correctness and technical correctness can be referred to the flow shown in fig. 4 and 5, respectively.

Further, when the operation task passes verification, the issuing process of the operation task can be performed, and the issuing process is similar to the verification process of the operation task, wherein the two processes are different in that a simulation environment of an access party is used during verification, and the simulation environment usually has only a single instance, and the operation task is issued to a plurality of instances at the same time during issuing; in addition, when the operation task is released, the corresponding data snapshot is not generated any more, because the snapshot is generated again at this time, the data of the possible snapshot is different from the related data of the operation task which is already verified, and further the operation task data released online is inconsistent with the operation task data which is verified, and finally the operation of the verification operation task is lost as a key step.

In a word, the embodiment of the application adopts a modularized design idea to abstract and decompose intelligent operation business into 'building block' modules such as characteristics, rules, flow experiments, algorithm models, algorithm strategies, operation tasks and the like, and algorithm engineers and operators can be matched with data meeting business scenes at will according to specific businesses to produce. Meanwhile, the introduction of the simulation environment of the access party and the use of the automatic verification logic can help the user to realize the verification of the service correctness and the technical correctness of related data (such as operation tasks, algorithm models and algorithm strategies). Therefore, the research and development period can be effectively shortened, the online iteration period is shortened, and the algorithm is enabled to be possible to carry out quick verification; the development process comprising a plurality of role nodes can be shortened to be a self-closed loop of an algorithm engineer, zero participation of a technical engineer and a tester is realized, and the labor cost is greatly reduced by improving the whole release process; moreover, the intelligent operation service does not need technical development and test, communication cost is saved, a chain which is easy to cause problems is standardized, the robustness of the whole service system is improved, errors in the manual implementation process are avoided, and the stability of the technical system is improved.

Fig. 2 is a flow chart of an operation task issuing method in one embodiment. It should be understood that, although the steps in the flowchart of fig. 2 are shown in sequence as indicated by the arrows, the steps are not necessarily performed in sequence as indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in fig. 2 may include multiple sub-steps or stages that are not necessarily performed at the same time, but may be performed at different times, nor do the order in which the sub-steps or stages are performed necessarily performed in sequence, but may be performed alternately or alternately with at least a portion of the sub-steps or stages of other steps or other steps.

In one embodiment, a computer device is provided, the internal structure of which may be as shown in FIG. 6. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is used for storing data such as operation service related data configured by a user, and the specific stored data can also be referred to as limitation in the embodiment of the method. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program when executed by a processor implements an intelligent decision management method.

It will be appreciated by those skilled in the art that the structure shown in fig. 6 is merely a block diagram of some of the structures associated with the present application and is not limiting of the computer device to which the present application may be applied, and that a particular computer device may include more or fewer components than shown, or may combine certain components, or have a different arrangement of components.

The present embodiment also provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method provided in any of the method embodiments described above when the computer program is executed.

In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored which, when executed by a processor, carries out the steps of a method as provided in any of the method embodiments described above.

Those skilled in the art will appreciate that implementing all or part of the above-described method embodiments may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the method embodiments described above. Any reference to memory, storage, database, or other medium used in the various embodiments provided herein may include non-volatile and/or volatile memory. The nonvolatile memory can include Read Only Memory (ROM), programmable ROM (PROM), electrically Programmable ROM (EPROM), electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double Data Rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link DRAM (SLDRAM), memory bus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), among others.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The above examples merely represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application is to be determined by the claims appended hereto.

Claims (10)

1. An intelligent decision management system, comprising:

the user interaction module is used for providing a corresponding configuration page for a user according to a configuration instruction sent by the user, so that the user can upload operation service related data through the configuration page, wherein the operation service related data comprises feature data, rule data, strategy data, flow experiment configuration data, model data and task data;

the rule verification module is used for verifying the rule data uploaded by the user;

the flow experiment verification module is used for creating a flow experiment according to the flow experiment configuration data uploaded by the user and verifying the flow experiment;

the strategy verification module is used for creating an algorithm strategy according to strategy data uploaded by a user and verifying the algorithm strategy;

the model verification module is used for loading an algorithm model corresponding to the model data uploaded by the user and verifying the algorithm model;

the task verification module is used for creating an operation task according to task data uploaded by a user and verifying the operation task;

and the issuing module is used for determining a target operation task and a target access party according to an issuing instruction sent by a user and issuing the target operation task to all service instances of the target access party.

2. The system of claim 1, wherein the model data comprises model files; the model verification module is specifically used for: and uploading the model file in the model data uploaded by the user to a designated file server to create a corresponding algorithm model, and verifying the created algorithm model through a simulation environment service provided by an access party corresponding to the model data.

3. The system of claim 1, wherein the task verification module is specifically configured to: and creating an operation task according to task data uploaded by a user, and verifying the created operation task through a simulation environment service provided by an access party corresponding to the operation task.

4. A system according to claim 1 or 3, wherein the task verification module, when verifying an operational task, is specifically configured to: acquiring rule data, a flow experiment, an algorithm model and an algorithm policy associated with the operation task, copying the currently acquired rule data, flow experiment, algorithm model and algorithm policy as a task snapshot associated with the operation task, and verifying the operation task according to the task snapshot;

correspondingly, when the publishing module publishes the target operation task to all service instances of the target access party, the publishing module is specifically configured to: and acquiring a target task snapshot associated with the target operation task, and publishing the target operation task to all service instances of the target access party according to the target task snapshot.

5. The system of claim 4, wherein the task verification module, when verifying the operational task from the task snapshot, is specifically configured to:

checking whether rule data, flow experiments, algorithm models and algorithm strategies contained in the task snapshot are effective or not;

when the detection results are valid, acquiring the instance data of all the valid instances of the access party, which are arranged in the simulation environment, from the service registration and service discovery cluster according to the unique identifier of the access party corresponding to the task snapshot;

the related data of the operation task is sent to each effective instance through an operation task pushing interface;

and when receiving the task loading success notification returned by all the effective examples through the callback interface, determining that the operation task passes the verification.

6. The system of claim 5, wherein the task verification module is further configured to return a prompt message to a user when the detection result is not valid or when a task loading failure notification returned by any valid instance is received.

7. The system of claim 1, wherein the system further comprises:

the feature module is used for managing all feature data;

the rule module is used for managing all rule data;

the flow experiment module is used for managing all flow experiment configuration data;

the model module is used for managing all model data;

the policy module is used for managing all policy data;

and the task module is used for managing all task data.

8. An operation task issuing method, characterized in that the method comprises the following steps:

receiving a configuration instruction sent by a user, and providing a configuration page corresponding to the configuration instruction for the user, so that the user can upload operation service related data through the configuration page, wherein the operation service related data comprises feature data, rule data, strategy data, flow experiment configuration data, model data and task data;

verifying the rule data, creating a flow experiment and verifying the flow experiment according to the flow experiment configuration data, creating an algorithm strategy and verifying the algorithm strategy according to the strategy data, loading an algorithm model corresponding to the model data and verifying the algorithm model, and creating an operation task and verifying the operation task according to the task data;

and receiving an issuing instruction sent by a user, determining a target operation task and a target access party according to the issuing instruction, and issuing the target operation task to all service instances of the target access party.

9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of claim 8 when the computer program is executed by the processor.

10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of claim 8.

Images