Method, device and system for optimizing message pushing strategy based on decision-making platform

Abstract

The invention discloses a method, a device and a system for optimizing a message pushing strategy based on a decision platform. The method comprises the following steps: receiving configuration operation of a user on a message index and the message pushing strategy; respectively generating a target message index and a target message pushing strategy according to the configuration operation; obtaining the first data and the second data, and integrating the first data and the second data to generate multi-dimensional index data; extracting target data from the multi-dimensional index data according to the target message index; inputting the target data into a decision component in sequence to execute a target message pushing strategy; and optimizing the target message pushing strategy based on the execution result. A message operator can automatically generate and execute a target message pushing decision only by configuring message indexes and message pushing decisions of different life cycles on an interface, and meanwhile, the automatically generated target message pushing strategy is optimized according to an execution result; therefore, messages of different life cycles in message pushing can be managed and optimized according to different pushing decisions, and the message pushing efficiency and the message effect are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of message pushing, and in particular, to a method, device, and system for optimizing a message pushing strategy based on a decision-making platform. Background Art

[0002] With the booming development of the mobile Internet, information flow has gradually become popular in message pushing. Among them, channel-pushed information flow has become the main way to attract new customers. At present, message operators push all messages using the same pushing strategy. However, as the cost of buying traffic continues to rise, the user growth of this pushing method has encountered a bottleneck. This requires message operators to carry out refined operations and adopt different pushing strategies for messages in different life cycles to improve the message pushing efficiency and effect. Summary of the Invention

[0003] In view of this, the main purpose of the present invention is to propose a method, device, and system for optimizing a message pushing strategy based on a decision-making platform, so as to at least partially solve at least one of the above technical problems.

[0004] To solve the above technical problems, a first aspect of the present invention proposes a method for optimizing a message pushing strategy based on a decision-making platform, the method comprising:

[0005] Receiving a configuration operation of a user on message metrics and a message pushing strategy for different life cycles;

[0006] Generating a target message metric and a target message pushing strategy respectively according to the configuration operation;

[0007] Respectively obtaining first data and second data from a decision-making platform, and integrating the first data and the second data to generate multi-dimensional metric data;

[0008] Extracting target data from the multi-dimensional metric data according to the target message metric;

[0009] Sequentially inputting the target data into a target message pushing decision component to execute the target message pushing strategy;

[0010] Outputting an execution result, and optimizing the target message pushing strategy based on the execution result.

[0011] According to a preferred embodiment of the present invention, the target message pushing decision component at least includes:

[0012] An object filter for determining the data range and data level to which a target message pushing strategy is applied;

[0013] A link line for determining whether to enter the next target message pushing decision component according to the metric value output by the previous target message pushing decision component;

[0014] A rule set for making logical judgments on multi-dimensional index data and outputting new index data;

[0015] An execution component for performing specific operations according to the output index data of the previous target message push decision component.

[0016] According to a preferred embodiment of the present invention, the execution component includes:

[0017] A decision logic execution component for performing corresponding decisions according to the new index data output by the rule set;

[0018] And / or;

[0019] A decision reach component for modifying the target message push elements according to the new index data output by the rule set, and / or sending an execution notice to the user, where the execution notice includes: execution result and / or modification of the target message push elements.

[0020] According to a preferred embodiment of the present invention, the target message push decision component further includes:

[0021] A random splitter for randomly splitting multi-dimensional index data to implement the AB test of the target message push strategy.

[0022] According to a preferred embodiment of the present invention, the obtaining the first data and the second data from the decision platform respectively and integrating the first data and the second data to generate multi-dimensional index data includes:

[0023] Periodically pulling message data and writing it into the database, and aggregating to generate the first data;

[0024] Receiving real-time queue messages of each service and aggregating relevant data to generate the second data;

[0025] Merging the first data and the second data to generate multi-dimensional index data.

[0026] According to a preferred embodiment of the present invention, the aggregating to generate the first data includes:

[0027] Combining the data table name and the field name to generate a plurality of unique atomic indicators;

[0028] Generating derivative indicators according to a plurality of atomic indicators.

[0029] To solve the above technical problems, a second aspect of the present invention provides a platform for optimizing the message push strategy based on a decision platform, and the platform includes:

[0030] A receiving module for receiving the configuration operations of the user on the message indicators and message push strategies in different life cycles;

[0031] A generation module, configured to generate a target message metric and a target message push decision respectively according to the configuration operations;

[0032] An integration module, configured to obtain first data and second data from a decision-making platform respectively, and integrate the first data and the second data to generate multi-dimensional metric data;

[0033] An extraction module, configured to extract target data from the multi-dimensional metric data according to the target message metric;

[0034] An execution module, configured to input the target data into a target message push decision component in sequence to execute the target message push policy;

[0035] An output module, configured to output an execution result, and optimize the target message push policy based on the execution result.

[0036] According to a preferred embodiment of the present invention, the target message push decision component at least includes:

[0037] An object filter, configured to determine the data range and data level to which the target message push policy applies;

[0038] A link line, configured to determine whether to enter the next target message push decision component according to the metric value output by the previous target message push decision component;

[0039] A rule set, configured to perform a logical judgment on the multi-dimensional metric data and output new metric data;

[0040] An execution component, configured to perform specific operations according to the output metric data of the previous target message push decision component.

[0041] According to a preferred embodiment of the present invention, the execution component includes:

[0042] A decision logic execution component, configured to perform corresponding decisions according to the new metric data output by the rule set;

[0043] And / or;

[0044] A decision reach component, configured to modify the target message push element according to the new metric data output by the rule set, and / or send an execution notice to a user, where the execution notice includes: an execution result and / or a modification to the target message push element.

[0045] According to a preferred embodiment of the present invention, the target message push decision component further includes:

[0046] A random splitter, configured to randomly split the multi-dimensional metric data to implement an AB test of the target message push policy.

[0047] According to a preferred embodiment of the present invention, the integration module includes:

[0048] A first aggregation module, configured to periodically pull message data and write it into a database, and aggregate and generate first data;

[0049] A second aggregation module, configured to receive real-time queue messages of each service, and aggregate relevant data to generate second data;

[0050] A merging module, configured to merge the first data and the second data to generate multi-dimensional metric data.

[0051] According to a preferred embodiment of the present invention, the first aggregation module includes:

[0052] A first generation module, configured to combine a data table name and a field name to generate a plurality of unique atomic metrics;

[0053] A second generation module, configured to generate derivative metrics according to a plurality of atomic metrics.

[0054] To solve the above technical problems, a third aspect of the present invention provides a system for optimizing a message push policy based on a decision-making platform, including: the platform for optimizing the message push policy based on the decision-making platform according to any one of the above, a supply chain platform, a decision-making platform, and / or a notification platform.

[0055] In summary, the present invention receives, through a visual interface, a user's configuration operations on message metrics and message push policies for different life cycles; generates target message metrics and target message push policies respectively according to the configuration operations; obtains first data and second data from a decision-making platform respectively, and integrates the first data and the second data to generate multi-dimensional metric data; extracts target data from the multi-dimensional metric data according to the target message metrics; sequentially inputs the target data into a target message push decision-making component to execute the target message push policy; outputs an execution result, and optimizes the target message push policy based on the execution result. In this way, message operation personnel only need to operate and configure message metrics and message push decisions for different life cycles on the interface, and can automatically generate and execute target message push decisions, and at the same time optimize the target message push policy according to the execution result. In addition, the execution result is output in a timely manner, which is convenient for users to adjust and optimize the target message push policy according to the message push effect, so as to realize the management and optimization of different push decisions for messages in different life cycles during message push, and improve the message push efficiency and message effect. Description of the Drawings

[0056] To make the technical problems solved by the present invention, the technical means adopted, and the technical effects achieved more clear, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, it should be noted that the drawings described below are only the drawings of the exemplary embodiments of the present invention, and those skilled in the art can obtain the drawings of other embodiments based on these drawings without creative efforts.

[0057] Figure 1 It is a schematic diagram of a method for optimizing a message push strategy based on a decision-making platform provided by an embodiment of the present invention;

[0058] Figure 2 It is a schematic diagram of generating target message metrics and target message decisions in a screen page according to an embodiment of the present invention;

[0059] Figure 3 It is a schematic diagram of the structural framework of a platform for optimizing a message push strategy based on a decision-making platform provided by an embodiment of the present invention;

[0060] Figure 4 It is a schematic diagram of an architecture for optimizing a message push strategy based on a decision-making platform provided by an embodiment of the present invention;

[0061] Figure 5 It is a schematic diagram of another method for optimizing a message push strategy based on a decision-making platform provided by an embodiment of the present invention;

[0062] Figure 6 It is a schematic diagram of the structural framework of a system for optimizing a message push strategy based on a decision-making platform provided by an embodiment of the present invention. Detailed Embodiments

[0063] On the premise of conforming to the technical concept of the present invention, the structures, performances, effects, or other features described in a specific embodiment can be combined in any suitable manner into one or more other embodiments.

[0064] In the process of introducing specific embodiments, the detailed descriptions of the structures, performances, effects, or other features are for those skilled in the art to fully understand the embodiments. However, it does not exclude that those skilled in the art can implement the present invention with technical solutions that do not contain the above-mentioned structures, performances, effects, or other features in specific situations. The figures in the drawings are only exemplary demonstrations and do not represent that all the contents, operations, and steps in the figures must be included in the solutions of the present invention, nor does it represent that they must be executed in the order shown in the figures.

[0065] Refer to Figure 1 , Figure 1 It is a schematic diagram of a method for optimizing a message push strategy based on a decision-making platform provided by an embodiment of the present invention. As Figure 1, the method for optimizing the message push policy based on the decision-making platform includes:

[0066] S1. Receive the configuration operations of the user on the message metrics and message push policies in different life cycles;

[0067] In the embodiment of the present invention, on the front end, a message push policy automatic construction canvas page is developed using the vue framework and the graph visualization engine g6. Different life cycle message metrics, message metric groups, decision-making components, decision-making sets, and dynamically configurable distributed scheduling periods are preset in the canvas page. The scheduling period refers to the period for collecting message metric data from each message push system. Among them: the message metrics can be bound to the message metric groups according to the user configuration operation, and the decision-making components can be bound to the decision-making sets according to the user configuration operation. In this way, the user can quickly and automatically create a target message push policy for pushing messages through the configuration operations on the message metrics and decision-making components.

[0068] Among them: the configuration operations on the message metrics may include: the selection and / or creation operations of the message metrics, and the binding operations of the message metrics and message metric groups. The configuration operations on the target message push policy may include: the selection and / or creation operations of the decision-making components, and the binding operations of the decision-making components and decision-making sets. The selection operation may be: clicking, dragging, lassoing, etc. The binding operation may be dragging the message metrics into the corresponding message metric groups, or dragging the decision-making components into the corresponding decision-making sets. The scheduling period can be configured according to user customization.

[0069] Exemplarily, the message metrics can reflect the effects of message push. For example, they can be: exposure rate, click-through rate, conversion rate, etc. Among them: The message can be: notification, alarm, advertisement, instruction, etc., which are not specifically limited in the present invention; The decision-making component is the smallest unit for executing the message push strategy, and can include: An object filter for determining the data range and data level to which the message push strategy applies; A link line for determining whether to enter the next decision-making component based on the metric value output by the previous decision-making component. Specifically, it can compare the metric value output by the previous decision-making component with a decision threshold, and determine whether to enter the next decision-making component according to the comparison result. A rule set for grouping and managing the message metric data, and performing logical judgments on the message metric data to output new metric data. Among them: The rule set can include multiple rules related to message metrics, and each target message push strategy can select at least one rule from the rule set according to user configuration. Taking the output of a new metric by the rule set: push value stratification as an example, when the click-through rate in the first 3 days before message push > A and the purchase rate in the first 3 days before message push > B1, the push value stratification is assigned as high-quality push; when the click-through rate in the first 3 days before message push > A and the purchase rate in the first 3 days before message push > B2 and < B1, the push value stratification is assigned as general push; when the click-through rate in the first 3 days before message push > A and the purchase rate in the first 3 days before message push < B3, the push value stratification is assigned as low-quality push. An execution component for performing specific operations according to the output metric data of the previous decision-making component.

[0070] Optionally, the execution component includes:

[0071] A decision logic execution component for performing corresponding decisions according to the new metric data output by the rule set; for example: operations such as enabling, disabling, adjusting the bid, and adjusting the message playback period can be performed on the message. Thus, "empty consumption" messages can be closed in a timely manner, and the message push budget can be controlled to be consumed smoothly. Among them: Adjusting the bid can include: adjusting the account budget, adjusting the message push budget, adjusting the message bid, etc.

[0072] And / or;

[0073] A decision reach component for modifying the message push elements according to the new metric data output by the rule set, and / or sending an execution notice to the user. The execution notice includes: execution result and / or modification of the message push elements. Among them: The message push elements can include: message plan, message group, message creative, etc. The execution notice can be sent by email, text message, instant messaging. Thus, the target message push decision can be automatically executed, and the message push elements such as message plan and message account can be automatically created / adjusted after reaching the decision threshold, or the message can be notified to the business party, enabling the business party to perceive in a timely manner and make timely adjustments to the message push elements such as message account and plan, improving the message push efficiency.

[0074] Furthermore, the decision-making component further includes:

[0075] A random splitter, which is used to randomly split multi-dimensional index data to implement the AB test of the strategy. Thus, after the target message push strategy is executed, an AB test comparison is performed on different strategy branches in the target message push strategy, data is collected, and an effect report is generated, which facilitates the business side to adjust the strategy plan according to the effect report, forming a closed loop supported by the target message push strategy and improving the message push effect.

[0076] S2. Generate a target message index and a target message push strategy respectively according to the configuration operation;

[0077] For the target message index, the user can directly select an existing message index group according to needs, or create a new message index group. Through the index group, indexes with similar attributes can be grouped together. For example, the message indexes of basic attributes can be grouped together, or the indexes of effect attributes can be grouped together. In this way, it can be more intuitively displayed and convenient for the operation personnel to view. Subsequently, a target message index is generated according to the user's selection operation of the message index, and an index binding relationship is generated according to the user's binding operation of the target message index and the message index group.

[0078] For the target message push strategy, the user can directly select an existing decision set according to needs, or create a new decision set. Among them: The decision set includes relevant decision-making components established around a certain theme, which can more intuitively display the purpose of the decision set. Subsequently, a target message push decision-making component is generated according to the user's selection operation of the decision-making component, and then a target message push decision is generated. The target message push decision constitutes the target message push decision tree for this message push. By customizing the dynamic configuration scheduling period by the user, a decision binding relationship is generated according to the user's binding operation of the decision-making component and the decision set.

[0079] Furthermore, the target message index and the target message push decision can be published.

[0080] For example Figure 2 , on the left side of the front-end canvas page, there are displayed a decision form, decision-making components, the decision set, and decision scheduling. Among them: The decision form includes: message basic settings (i.e., message index settings) and index grouping. The decision-making components include: an object selector, a rule set, a random splitter, an AB splitter, a SMS notification, an adjustment of account budget, an adjustment of message budget, an adjustment of message bid, and an adjustment of message time period. The user performs a "drag and drop" configuration operation on the specific options in the decision form, decision-making components, the decision set, and decision scheduling on the front-end canvas page to generate the target message push decision displayed on the right side of the canvas page.

[0081] S3. Obtain the first data and the second data from the decision-making platform respectively, and integrate the first data and the second data to generate multi-dimensional index data;

[0082] In this embodiment, the first data may be front-end data, and the second data may be back-end conversion data, where: the front end can support data from multiple data sources. The back end uses the springboot framework + mybatis framework + doris database + tidb database to develop an index configuration page, and uses self-service configuration to access the target message index data source without business code development. Users can dynamically configure the distributed scheduling period on the front-end canvas page to achieve dynamic configurability of the index data running period.

[0083] Exemplarily, this step may include:

[0084] S31. Regularly pull message data and write it into the database, and aggregate it to generate the first data;

[0085] For example: The message procurement trading platform (connected to the media) can regularly (for example, every half hour) pull the message data of the media MKT-API and write it into the local databases TIDB and DORIS, and aggregate it to generate the message push front-end data. Exemplarily, during the aggregation process, the table name field names in the source data can be combined to uniquely identify an atomic index; in addition, derivative indexes can be generated from multiple atomic indexes. For example: Derivative indexes are generated by performing arithmetic operations on multiple atomic indexes. Among them: Atomic indexes and derivative indexes are used as a basic index condition in the judgment logic executed by the target message push decision component in subsequent operations. For example: Take the number of days for message plan push as an atomic index. In a target message push decision component, when the number of days for message plan push is between 1 and 4, continue to execute the logic of a target message push decision component.

[0086] S32. Receive the real-time queue messages of each business, and aggregate the relevant data to generate the back-end conversion data, the second data;

[0087] For example: Receive the real-time queue messages sent by the key business nodes of each business system through the advertising message attribution system, and aggregate the relevant data to generate the back-end conversion data. Among them: The relevant data refers to the back-end conversion effect data of the message. For example, if a purchase user is generated by a certain message at a certain time, the relevant data within the same time range of the same message can be added during aggregation.

[0088] S33. Merge the first data and the second data to generate multi-dimensional index data;

[0089] The front - end data and the back - end conversion data can be merged according to different dimensions to obtain the metric data corresponding to the respective dimensions, thereby obtaining multi - dimensional metric data. Among them: the different dimensions can be multiple single dimensions, or composite dimensions generated from multiple dimensions. For example: a composite dimension of the message push time and the four elements in the message push landing page link (main channel label, sub - channel, message content, message keyword).

[0090] S4. Extract target data from the multi - dimensional metric data according to the target message metrics;

[0091] Specifically, the data corresponding to the target message metrics in the multi - dimensional metric data can be extracted as the target data.

[0092] S5. Input the target data into the target message push decision component in sequence to execute the target message push strategy;

[0093] Exemplarily, such as Figure 2 , it is possible to scan the target message push decision component that needs to be executed currently at regular intervals (such as every 1 hour), send the target message metric data of the message object through a message middleware (such as: rocket MQ), and push the extracted target data to this target message push decision component. The decision - making consumer end caches the decisions that need to be executed currently and starts to receive the target data of the message object. When the decision - making consumer node receives the target data, it selects the execution data within the data range and data hierarchy determined by the object selector, executes the logical rules of the rule set, and calls the media API interface to adjust the target message push strategy or send a message notification.

[0094] Among them: the size of the target data sent by the message middleware can be adjusted as needed. At the same time, when receiving the target data, the number of consumer threads and the number of consumer nodes can also be adjusted to achieve horizontal expansion.

[0095] S6. Output the execution result and optimize the target message push strategy based on the execution result.

[0096] Among them: the execution result can include: execution records, execution details, and execution effects. When the AB splitter is included in the target message push decision component, after the decision is executed, AB testing and comparison can be carried out for different policy branches in the target message push strategy, data can be collected, and an execution effect report can be generated. On the one hand, the target message push strategy can be automatically optimized according to the execution result. For example: different optimization methods corresponding to different execution results can be preset, and then the target message push strategy can be optimized according to the optimization method corresponding to the current execution result. Among them: the optimization of the target message push strategy can be the optimization of the message push time and frequency, message push budget, enabling and disabling of message push elements, and automatic creation, etc. On the other hand, generating the execution effect report can also facilitate the business side to adjust the target message push strategy according to the effect report.

[0097] Further, the target message push policy can be graded according to the execution result of the output, and different push methods (such as push channels, push times, push frequencies, etc.) can be adopted for different grades to push messages.

[0098] Based on Figure 1 the method for optimizing the message push policy based on the decision-making platform, an embodiment of the present invention further provides a platform for optimizing the message push policy based on the decision-making platform, as Figure 3 shown, the platform includes:

[0099] A receiving module 31, configured to receive a user's configuration operation on message metrics and message push policies for different lifecycles;

[0100] A generating module 32, configured to generate target message metrics and target message push decisions respectively according to the configuration operation;

[0101] An integrating module 33, configured to obtain first data and second data from the decision-making platform respectively, and integrate the first data and the second data to generate multi-dimensional metric data;

[0102] An extracting module 34, configured to extract target data from the multi-dimensional metric data according to the target message metrics;

[0103] An executing module 35, configured to sequentially input the target data into a target message push decision component to execute the target message push policy;

[0104] An output module 36, configured to output an execution result, and optimize the target message push policy based on the execution result.

[0105] Wherein, the target message push decision component at least includes:

[0106] An object filter, configured to determine the data range and data level to which the target message push policy applies;

[0107] A link line, configured to determine whether to enter the next target message push decision component according to the metric value output by the previous target message push decision component;

[0108] A rule set, configured to perform a logical judgment on the multi-dimensional metric data and output new metric data;

[0109] An execution component, configured to perform a specific operation according to the output metric data of the previous target message push decision component.

[0110] Optionally, the execution component includes:

[0111] A decision logic execution component for executing corresponding decisions according to the new metric data output by the rule set;

[0112] And / or;

[0113] A decision reach component for modifying target message push elements according to the new metric data output by the rule set, and / or sending an execution notice to the user, where the execution notice includes: an execution result and / or a modification to the target message push element.

[0114] Furthermore, the target message push decision component further includes:

[0115] A random splitter for randomly splitting multi-dimensional metric data to implement A / B testing of the target message push strategy.

[0116] In a specific implementation manner, the integration module 33 includes:

[0117] A first aggregation module for periodically pulling message data and writing it into a database, and aggregating to generate first data;

[0118] A second aggregation module for receiving real-time queue messages of each service and aggregating relevant data to generate second data;

[0119] A merging module for merging the first data and the second data to generate multi-dimensional metric data.

[0120] Wherein: the first aggregation module includes:

[0121] A first generation module for combining data table names and field names to generate multiple unique atomic metrics;

[0122] A second generation module for generating derived metrics according to multiple atomic metrics.

[0123] The above platform for optimizing the message push strategy based on the decision platform can be implemented through Figure 4 The architecture for optimizing the message push strategy based on the decision platform shown, such as Figure 4As shown in the figure, the architecture for optimizing message push policies based on a decision-making platform includes: infrastructure services, a database, a metric calculation layer, a decision execution layer, and a display layer. Among them: Infrastructure services include: A scheduling service that can dynamically configure distributed scheduling to enable dynamic configurability of the message metric running cycle. A message queue service that can send business node information in real time. An observability service that can intuitively display message metrics, target message push policies, message push effects, etc. The database is used to provide a data source for the target message push policy and can include: TIDB and DORIS databases. The metric calculation layer can include: A grouping management module that can group message metrics according to the user's binding operation. A metric management module that manages existing message metrics and supports data from both TIDB and DORIS data sources. An atomic metric calculation module that can identify the atomic metrics in the data source. For example, if the data source is stored in the form of a data table, the combination of the table name and field name can be used to uniquely identify an atomic metric. A derived metric calculation module that can generate derived metrics by performing arithmetic operations on multiple atomic metrics. A metric data assembly module that can assemble atomic metrics and / or derived metrics into message metrics, and a push data synchronization module. The decision execution layer can include: A decision management module that can manage message push policies. Specifically, it can include: A list displaying existing message push policies, a decision canvas, a decision set management unit, and a decision version. A decision release management module that can manage the release of decisions. Specifically, it can include: A decision online / offline management unit, a decision rollback unit, and a release record unit. A decision component that can generate target message push decisions. Specifically, it can include: An object filter, a rule set, a link, and a random shunt. A decision reach module that can reach the corresponding decision component according to the decision execution result to execute the corresponding logic. Specifically, it can include: Modifying message push elements, modifying message accounts, sending SMS notifications, etc. A decision monitoring module that can monitor the decision execution process. Specifically, it can include: Running records, running details, and effect reports. The display layer can include: A decision display module that can display the currently automatically generated target message push decisions. A report visualization module that can display the execution results of the current target message push decisions. Further, a log recording module can be configured in the decision execution layer and the metric calculation layer to record the execution process of specific decisions. At the same time, a permission control module can be configured in the display layer, the decision execution layer, and the metric calculation layer to set different permissions for different users.

[0124] Those skilled in the art can understand that the various modules in the above device embodiments can be distributed in the device as described, or can be correspondingly changed and distributed in one or more devices different from the above embodiments. The modules in the above embodiments can be combined into one module, or further split into multiple sub-modules.

[0125] Figure 5It is a schematic flowchart of another method for optimizing a message push policy based on a decision-making platform provided in this embodiment of the present invention. As Figure 5 shown, the method includes:

[0126] S501. The platform for optimizing the message push policy based on the decision-making platform verifies the user login information and completes the user login.

[0127] The message operation staff can log in to the platform for optimizing the message push policy based on the decision-making platform through the account name and password. The platform for optimizing the message push policy based on the decision-making platform receives and verifies the account name and password input by the user for login. After successful verification, the user login is completed, and the canvas page of the user is displayed. After failed verification, the user is prompted to log in again.

[0128] S502. The platform for optimizing the message push policy based on the decision-making platform receives the configuration operations of the user on the message metrics and message push policies for different lifecycles, and respectively generates target message metrics and target message push policies according to the user's configuration operations;

[0129] The user can perform configuration operations such as "drag and drop", click, and circle selection on the canvas page, so as to complete functions such as creating metric groups, creating metrics, creating decision sets, creating decisions, and publishing decisions, and finally generate target message metrics and target message push policies.

[0130] For specific configuration operations, reference can be made to Figure 1 step S1 in, and for the generation of the target message metrics and the target message push policy, reference can be made to Figure 1 step S2 in.

[0131] S503. The platform for optimizing the message push policy based on the decision-making platform extracts target data from the front and back ends of the decision-making platform according to the target message metrics;

[0132] This step S503 is the same as steps S3 and S4 in Figure 1 , and will not be elaborated here.

[0133] S504. The platform for optimizing the message push policy based on the decision-making platform sequentially inputs the target data into the target message push decision component to execute the target message push policy, and decides to reach the transaction procurement platform and / or the notification platform. At the same time, the execution process of the decision logic is monitored;

[0134] This step can refer to Figure 1 step S5 in. The decision logic can be executed through the rule set, and the decision can be reached to the transaction procurement platform and / or the notification platform through the execution component. Further, a monitoring component can be added to monitor the execution process of the decision logic through the monitoring component.

[0135] S505. The platform for optimizing the message push policy based on the decision-making platform outputs the execution result, and optimizes the target message push policy based on the execution result.

[0136] Exemplarily, the execution result can be displayed in the form of a report on the canvas page, and the target message push policy can be automatically optimized according to the execution result. On the basis of this automatic optimization, the message staff can also manually adjust and optimize the target message push decision according to the optimized operation report.

[0137] Based on Figure 5 the method for optimizing the message push policy based on the decision-making platform described above, the embodiment of the present invention further provides a system for optimizing the message push policy based on the decision-making platform, as Figure 6 shown, including: the platform 61 for optimizing the message push policy based on the decision-making platform as Figure 5 shown, the decision-making platform 62, the supply chain platform 63, and / or the notification platform 64.

[0138] In summary, the present invention can be implemented by a method, a platform, an electronic device or a computer-readable medium that can execute a computer program. Some or all functions of the present invention can be implemented by using a general-purpose data processing device such as a microprocessor or a digital signal processor (DSP) in practice.

[0139] The specific embodiments described above have further elaborated on the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the present invention is not inherently related to any specific computer, virtual device or electronic device, and various general-purpose devices can also implement the present invention. The above are only specific embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for optimizing a message push policy based on a decision-making platform, characterized in that The method includes: Receiving the user's configuration operations on message metrics and message push policies for different lifecycles; Generating target message metrics and target message push policies respectively according to the configuration operations; Obtaining first data and second data from the decision-making platform respectively, and integrating the first data and the second data to generate multi-dimensional metric data; Extracting target data from the multi-dimensional metric data according to the target message metrics; Sequentially inputting the target data into the target message push decision-making component to execute the target message push policy; 2. The method according to claim 1, characterized in that, Outputting the execution result, and optimizing the target message push policy based on the execution result. The target message push decision-making component at least includes: An object filter, used to determine the data range and data level to which the target message push policy applies; A link line, used to judge whether to enter the next target message push decision-making component according to the metric value output by the previous target message push decision-making component; A rule set, used to perform logical judgment on the multi-dimensional metric data and output new metric data; 3. The method according to claim 2, characterized in that, An execution component, used to perform specific operations according to the output metric data of the previous target message push decision-making component. The execution component includes: A decision logic execution component, used to execute corresponding decisions according to the new metric data output by the rule set; And / or; 4. The method according to claim 3, characterized in that, A decision reach component, used to modify the target message push elements according to the new metric data output by the rule set, and / or send an execution notice to the user, where the execution notice includes: the execution result and / or the modification of the target message push elements. The target message push decision-making component further includes:

5. The method according to claim 1, characterized in that A random splitter, used to randomly split the multi-dimensional metric data to implement the A / B test of the target message push policy. The obtaining first data and second data from the decision-making platform respectively, and integrating the first data and the second data to generate multi-dimensional metric data includes: Regularly pulling message data and writing it into the database, and aggregating to generate the first data; Receiving the real-time queue messages of each service, and aggregating relevant data to generate the second data; 6. The method according to claim 5, characterized in that, Merging the first data and the second data to generate multi-dimensional metric data. The aggregating to generate the first data includes: Combining the data table name and the field name to generate multiple unique atomic metrics; 7. A platform for optimizing message push strategies based on a decision-making platform, characterized in that, Generating derived metrics according to multiple atomic metrics. The platform includes: A receiving module, used to receive the user's configuration operations on message metrics and message push policies for different lifecycles; A generating module, used to generate target message metrics and target message push decisions respectively according to the configuration operations; An integrating module, used to obtain first data and second data from the decision-making platform respectively, and integrate the first data and the second data to generate multi-dimensional metric data; An extracting module, used to extract target data from the multi-dimensional metric data according to the target message metrics; An executing module, used to sequentially input the target data into the target message push decision-making component to execute the target message push policy; 8. The platform according to claim 7, characterized in that, An output module, used to output the execution result, and optimize the target message push policy based on the execution result. The target message push decision-making component at least includes: An object filter for determining the data scope and data level to which the target message push policy is applied; A link line for determining whether to enter the next target message push decision component based on the metric value output by the previous target message push decision component; A rule set for performing logical judgments on multi-dimensional metric data and outputting new metric data; An execution component for performing specific operations based on the output metric data of the previous target message push decision component.

9. The platform according to claim 8, characterized in that, The execution component includes: A decision logic execution component for performing corresponding decisions based on the new metric data output by the rule set; And / or; A decision reach component for modifying the target message push element according to the new metric data output by the rule set, and / or sending an execution notice to the user, where the execution notice includes: an execution result and / or a modification to the target message push element.

10. The platform according to claim 9, characterized in that, The target message push decision component further includes: A random splitter for randomly splitting multi-dimensional metric data to implement an AB test of the target message push policy.

11. The platform according to claim 7, characterized in that, The integration module includes: A first aggregation module for periodically pulling message data and writing it into a database, and aggregating to generate first data; A second aggregation module for receiving real-time queue messages of each service and aggregating relevant data to generate second data; A merging module for merging the first data and the second data to generate multi-dimensional metric data.

12. The platform according to claim 11, wherein The first aggregation module includes: A first generation module for combining the data table name and field name to generate multiple unique atomic metrics; A second generation module for generating derived metrics based on multiple atomic metrics.

13. A system for optimizing a message push strategy based on a decision-making platform, characterized in that, Includes: The platform for optimizing the message push policy based on the decision platform according to any one of claims 7-12, the decision platform, the supply chain platform, and / or the notification platform.

Images