Strategy processing method and device for customer group index, electronic equipment and storage medium

Abstract

The application provides a customer group index strategy processing method and device, electronic equipment and a storage medium. Based on an obtained customer group management instruction, a target customer group is obtained, and a customer group index is obtained based on at least one target index in the customer group management instruction and a pre-constructed index metadata pool. Based on the customer group index, a real-time behavior label of the target customer group, an external variable weight, and a life cycle of the target customer group, an initial processing strategy for the target customer group is determined. In response to the customer group index reaching a preset warning threshold, the initial processing strategy corresponding to the customer group index is executed, a strategy execution result is determined. Based on the strategy execution result, weighted calculation is performed in multiple evaluation dimensions to determine a strategy execution score, and based on a strategy execution evaluation result of a selected target recommended strategy, the target recommended strategy is adjusted and / or stored. In this way, the strategy processing efficiency and accuracy can be improved.

Description

Technical Field

[0001] This application relates to the field of strategy management technology, and in particular to strategy processing methods, apparatus, electronic devices and storage media involving customer group indicators. Background Technology

[0002] Against the backdrop of rapid development in the digital economy, industries such as finance, retail, and e-commerce are increasingly demanding refined customer operations—achieving enhanced customer value and reduced operational risks through precise customer management, real-time indicator monitoring, timely risk warnings, and appropriate strategy execution.

[0003] In current technologies, indicators are typically supplemented by customizing models and then provided externally as a carrier. However, this approach of using custom rule logic cannot adapt to rapidly changing scenarios, resulting in excessively long time cycles for strategy adjustments and customization. Furthermore, single, fixed processing rules cannot accurately match real-time changes in business scenarios, which may lead to delayed or false policy alerts, affecting the accuracy of policy adjustments. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a method, device, electronic device and storage medium for processing customer group indicators. After determining the customer group management instructions, the application quickly obtains the customer group indicators based on the indicator metadata pool. Based on the processing of basic customer group indicators, the application introduces real-time behavior tags and external variable weights to formulate an initial processing strategy to dynamically adapt to the business scenarios in which adjustments are implemented. At the same time, the application processes the strategy warning based on dynamic warning thresholds to ensure that the warning process is more accurate. Furthermore, the application adjusts and evaluates the processing strategy based on multi-dimensional evaluation results, which helps to improve the efficiency and accuracy of strategy processing.

[0005] In a first aspect, embodiments of this application provide a strategy processing method for customer group indicators, the strategy processing method comprising: Based on the obtained customer group management instructions, the target customer group is obtained, and based on at least one target indicator in the customer group management instructions and a pre-built indicator metadata pool, the customer group indicators are obtained. Based on the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group, an initial processing strategy for the target customer group is determined. In response to the customer group indicator reaching a preset warning threshold, the initial processing strategy corresponding to the customer group indicator is executed, and the strategy execution result is determined; wherein, the preset warning threshold is dynamically determined based on the life cycle stage corresponding to the target customer group and the weight of the external variable; Based on the strategy execution results, a weighted calculation is performed on multiple evaluation dimensions to determine the strategy execution score. Based on the strategy execution evaluation results of the selected target recommendation strategies, the target recommendation strategies are adjusted and / or stored. The target recommendation strategies are obtained based on the strategy execution scores.

[0006] In one possible implementation, obtaining customer group metrics based on at least one target metric in the customer group management instruction and a pre-built metric metadata pool includes: Based on at least one target indicator in the customer group management instruction and the calculation logic template and configuration variables configured in the indicator metadata pool, a query statement is automatically generated. Execute the query statement to calculate the customer group data and obtain the customer group indicators.

[0007] In one possible implementation, determining the initial processing strategy for the target customer group based on the customer group metrics, the real-time behavioral tags of the target customer group, external variable weights, and the lifecycle of the target customer group includes: Based on the decision tree algorithm, the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the life cycle of the target customer group are matched with a pre-set strategy template to determine at least one candidate processing strategy for the target customer group. For each candidate processing strategy, calculate the matching degree between the candidate processing strategy and the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group, and determine the strategy matching value corresponding to the candidate processing strategy. The strategy whose matching value is greater than a preset matching value threshold among the at least one candidate processing strategy is determined as the initial processing strategy.

[0008] In one possible implementation, the lifecycle of the target customer group is determined through the following steps: The target customer group is divided into lifecycles based on the primary and secondary phases method, and lifecycle rules are generated for each lifecycle. Configure lifecycle rules for each lifecycle stage and generate a JSON data body; The query statement is generated by calculating the customer group data corresponding to the newly added or adjusted rules in the JSON data body and reusing the unchanged rules. The query statement is executed to obtain the migration data of the target customer group at each life cycle. Based on the migration data, the association between the target customer group and each life cycle is constructed to generate the life cycle of the target customer group.

[0009] In one possible implementation, the step of matching the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group with a pre-set strategy template based on a decision tree algorithm to determine at least one candidate processing strategy for the target customer group includes: For the customer lifecycle to which the target customer group belongs, based on the external variable weights, the real-time behavioral tags of the target customer group, and the customer group indicators, determine the target external weight variable information, target behavioral information, and target customer group indicator information that match the target customer group; Based on the customer lifecycle of the target customer group, and the combination of the target external weight variable information, the target behavior information, and the target customer group indicator information, the information is matched with a pre-set strategy template to determine the target strategy template, and at least one candidate processing strategy is generated based on the target strategy template.

[0010] In one possible implementation, the strategy processing method further includes: The processing strategies whose execution scores are greater than the preset score threshold are identified as candidate recommendation strategies; If the candidate recommendation strategy conflicts with other recommendation strategies, the target recommendation strategy is determined based on the preset strategy priority. If the candidate recommendation strategy does not conflict with other recommendation strategies, the candidate recommendation strategy is determined as the target recommendation strategy.

[0011] In one possible implementation, adjusting and / or storing the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy includes: If the strategy execution evaluation result of the target recommendation strategy meets the preset evaluation criteria, the target recommendation strategy and the customer characteristics and external variable information of the target customer group corresponding to the target recommendation strategy are stored so that the target recommendation strategy can be reused when processing the same customer characteristics and external variable information management instructions. If the strategy execution evaluation result of the target recommendation strategy does not meet the preset evaluation criteria, the abnormal information in the target recommendation strategy is adjusted to obtain an updated target recommendation strategy, and the strategy execution evaluation result of the updated target recommendation strategy is evaluated again until the strategy execution evaluation result of the target recommendation strategy meets the preset evaluation criteria, and the updated target recommendation strategy is stored.

[0012] Secondly, embodiments of this application also provide a strategy processing device for customer group indicators, the strategy processing device comprising: The customer group indicator acquisition module is used to acquire target customer groups based on the acquired customer group management instructions, and to acquire customer group indicators based on at least one target indicator in the customer group management instructions and a pre-built indicator metadata pool. The initial strategy construction module is used to determine the initial processing strategy for the target customer group based on the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group. The strategy execution module is used to execute the initial processing strategy corresponding to the customer group indicator in response to the customer group indicator reaching a preset warning threshold, and determine the strategy execution result; wherein, the preset warning threshold is dynamically determined based on the life cycle stage corresponding to the target customer group and the weight of the external variable; The recommendation strategy feedback module is used to perform weighted calculations on multiple evaluation dimensions based on the strategy execution results to determine the strategy execution score, and to adjust and / or store the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy; wherein, the target recommendation strategy is obtained based on the strategy execution score.

[0013] Thirdly, embodiments of this application also provide an electronic device, including: a processor, a storage medium, and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the customer group indicator strategy processing method as described in any of the first aspects.

[0014] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the customer group indicator strategy processing method as described in any of the first aspects.

[0015] The customer group indicator strategy processing method, apparatus, electronic device, and storage medium provided in this application embodiment acquire a target customer group based on the acquired customer group management instructions, and acquire customer group indicators based on at least one target indicator in the customer group management instructions and a pre-built indicator metadata pool; determine an initial processing strategy for the target customer group based on the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group; in response to the customer group indicators reaching a preset warning threshold, execute the initial processing strategy corresponding to the customer group indicators and determine the strategy execution result; based on the strategy execution result, perform weighted calculations on multiple evaluation dimensions to determine the strategy execution score, and adjust and / or store the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy. In this way, after determining the customer group management instructions, the customer group indicators are quickly obtained based on the indicator metadata pool. On the basis of processing the basic customer group indicators, real-time behavior tags and external variable weights are introduced to formulate the initial processing strategy, so as to dynamically adapt to the business scenarios of implementation and adjustment. At the same time, during the strategy warning process, the dynamic warning threshold is used to ensure that the warning process is more accurate. Furthermore, the processing strategy is adjusted and evaluated based on the multi-dimensional evaluation results, which helps to improve the efficiency and accuracy of strategy processing.

[0016] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating a strategy processing method for customer group indicators provided in an embodiment of this application; Figure 2 A schematic diagram of the structure of a strategy processing device for customer group indicators provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. Based on the embodiments of this application, every other embodiment obtained by those skilled in the art without inventive effort falls within the scope of protection of this application.

[0020] First, the applicable scenarios for this application will be introduced. This application can be applied to the field of strategy management technology.

[0021] Against the backdrop of rapid development in the digital economy, industries such as finance, retail, and e-commerce are increasingly demanding refined customer operations—achieving enhanced customer value and reduced operational risks through precise customer management, real-time indicator monitoring, timely risk warnings, and appropriate strategy execution.

[0022] In current technologies, indicators are typically supplemented by customizing models and then provided externally as a carrier. However, this approach of using custom rule logic cannot adapt to rapidly changing scenarios, resulting in excessively long time cycles for strategy adjustments and customization. Furthermore, single, fixed processing rules cannot accurately match real-time changes in business scenarios, which may lead to delayed or false policy alerts, affecting the accuracy of policy adjustments.

[0023] Based on this, embodiments of this application provide a strategy processing method for customer group indicators to improve the efficiency and accuracy of strategy processing.

[0024] Please see Figure 1 , Figure 1 This is a flowchart illustrating a strategy processing method for customer group metrics provided in an embodiment of this application. Figure 1 As shown in the embodiments of this application, the strategy processing method for customer group indicators includes: S101. Based on the obtained customer group management instructions, obtain the target customer group, and based on at least one target indicator in the customer group management instructions and the pre-built indicator metadata pool, obtain the customer group indicators.

[0025] S102. Based on the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group, determine the initial processing strategy for the target customer group.

[0026] S103. In response to the customer group indicator reaching a preset warning threshold, execute the initial processing strategy corresponding to the customer group indicator and determine the strategy execution result; wherein, the preset warning threshold is dynamically determined based on the life cycle stage corresponding to the target customer group and the weight of the external variable.

[0027] S104. Based on the strategy execution results, perform weighted calculations on multiple evaluation dimensions to determine the strategy execution score, and adjust and / or store the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy; wherein, the target recommendation strategy is obtained based on the strategy execution score.

[0028] The customer group indicator strategy processing method provided in this application embodiment, after determining the customer group management instruction, quickly obtains the customer group indicators based on the indicator metadata pool. On the basis of processing the basic customer group indicators, it introduces real-time behavior tags and external variable weights to formulate an initial processing strategy to dynamically adapt to the business scenario of implementation adjustment. At the same time, during the strategy warning process, it processes based on dynamic warning thresholds to ensure that the warning process is more accurate. Furthermore, it adjusts and evaluates the processing strategy based on multi-dimensional evaluation results, which helps to improve the efficiency and accuracy of strategy processing.

[0029] The exemplary steps of the embodiments of this application are described below: S101. Based on the obtained customer group management instructions, obtain the target customer group, and based on at least one target indicator in the customer group management instructions and the pre-built indicator metadata pool, obtain the customer group indicators.

[0030] In one possible implementation, a graphical user interface can be provided at the front end, where users can input customer group management instructions. Specifically, the customer group management instructions may include key information such as the industry to which the customer group belongs, core management objectives, and the size range of the customer group.

[0031] For example, the customer base may belong to industries such as industrial finance or e-commerce; core management objectives may include risk warning and value enhancement.

[0032] Furthermore, after receiving the customer group management instructions input by the user, the corresponding target customer group can be obtained according to the customer group management instructions.

[0033] Specifically, customer group rules are defined based on the actual business scenarios corresponding to customer group management instructions, and customer group filtering conditions are provided. These rules are configured using a canvas approach. The customer group rules configured in the canvas are converted into JSON data. If the number of bytes in the JSON data exceeds a preset byte threshold, such as 4000 bytes, the JSON data is split and stored in a specified field of the database in an ordered manner. Based on the configured customer group rules, an SQL query statement is generated to obtain customer group data and thus the target customer group.

[0034] Here, the steps for obtaining customer group data are optimized as follows: First, a SQL query is invoked using the preview function to query the cached data in the local database. If the cache is not found, a distributed computing framework, such as Spark, is invoked to execute the SQL query, retrieving detailed customer group data from the big data platform, such as cardholder information. The query results are then synchronized to the local database cache, and the cache expiration tag is updated. The default cache expiration time is 24 hours, which users can adjust according to their needs. If the cached data exceeds the expiration tag, a re-query is automatically triggered to ensure the timeliness of the customer group data.

[0035] The local database can be an Oracle database.

[0036] Furthermore, to ensure the accuracy and availability of the acquired customer data, after obtaining the target customer group through a query, the acquired customer group can be approved. Specifically, an approval request is initiated for the target customer group, and the approval request is sent to the personnel at the designated approval node according to the preset approval process. The approval results of each approval node are received. If the approval results of all approval nodes are approved, the status of the target customer group is updated to the effective and online status, which is used for subsequent indicator analysis and processing strategy binding.

[0037] In one possible implementation, after obtaining the target customer group, the corresponding customer group indicators can be obtained based on at least one target indicator in the customer group management instructions and a pre-built indicator metadata pool.

[0038] Specifically, the step "obtaining customer group metrics based on at least one target metric in the customer group management instruction and a pre-built metric metadata pool" includes: a1: Based on at least one target indicator in the customer group management instruction and the calculation logic template and configuration variables configured in the indicator metadata pool, a query statement is automatically generated.

[0039] a2: Execute the query statement to calculate the customer group data and obtain the customer group indicators.

[0040] In one possible implementation, a corresponding indicator metadata pool can be pre-built, which includes three processing modules: indicator basic attributes, calculation logic templates, and data source mapping.

[0041] For example, the basic attributes of an indicator may include information such as indicator type and statistical dimensions; the calculation logic template may include logic processing templates such as summation, average, and turnover days; and the data source mapping represents the source information of customer group data, such as transaction tables and customer tables.

[0042] Here, when storing customer group data, large-scale customer group data can be split into multiple database tables according to dimensions such as region and industry. The Spark distributed computing framework can be used to simultaneously calculate indicators for multiple tables associated with customer group management commands, thereby obtaining the corresponding indicator data and further improving computing efficiency.

[0043] In one possible implementation, the user inputs at least one target indicator through a front-end graphical user interface. Specifically, the user can select from multiple indicator types displayed in the graphical user interface, such as financial or behavioral target indicators. Then, the user inputs and configures corresponding key variables in the front-end graphical user interface, such as statistical period and dimension fields. Based on the user's input of at least one target indicator and the calculation logic template and configuration variables configured in the indicator metadata pool, an optimized query statement is automatically generated without modifying the underlying basic processing code, further improving the efficiency of obtaining customer group indicators.

[0044] Here, the query statement can be an SQL statement that includes database sharding and table partitioning logic.

[0045] Furthermore, after generating the query statement, the distributed computing framework (Spark) is invoked to execute the SQL query statement, perform parallel computing on large-scale customer data, and obtain statistical information corresponding to the target indicators, such as the distribution of transaction amount, activity frequency, and accounts receivable turnover days of the customer group within a specified period, thereby obtaining the corresponding customer group indicators.

[0046] Here, in order for users to observe the acquired customer segment data, the statistical information obtained from the query can be displayed in the form of bar charts, pie charts, or line charts to obtain customer segment metrics.

[0047] In one possible implementation, the history of changes in indicator configuration can also be recorded, and the trend of indicator changes can be determined by recording different versions of historical indicators. At the same time, it is possible to roll back to any historical version of the indicator so that different indicator versions can be accurately located during subsequent statistical processing and processing strategy adjustment, thereby determining specific abnormal situations and the direction of processing strategy adjustment, and improving the accuracy of subsequent processing strategy adjustment.

[0048] For example, in an industrial finance scenario, if business personnel need to add a supply chain accounts receivable turnover days indicator, they only need to select the financial indicator type in the front-end graphical user interface, configure the variable statistical period as 30 days and the dimension field as supplier ID, and the system will automatically call the turnover days calculation template from the indicator metadata pool to generate an SQL query statement containing database sharding and table sharding query logic. Through Spark parallel computing, a large amount (e.g., millions) of supplier indicator data can be quickly obtained without modifying the basic code, thereby improving the efficiency of customer group indicator acquisition.

[0049] In one possible implementation, to improve the efficiency of acquiring customer metrics, for frequently queried metric data, i.e., metric data whose query frequency is greater than a preset query frequency threshold within a preset historical time period, such as daily active user frequency, the frequently queried metric data can be cached. The cache duration is dynamically adjusted according to the metric update frequency. When it is determined that the frequently queried metric data needs to be acquired, it can be retrieved directly from the cache, thus extending the metric query time and improving the efficiency of acquiring customer metrics.

[0050] For example, the caching period could be 12 hours for daily updated metrics and 7 days for weekly updated metrics.

[0051] In another possible implementation, computing resources can be automatically allocated based on the size of the customer base and the computational complexity of the metrics, in order to avoid resource waste or shortage and further improve the efficiency of acquiring customer base metrics.

[0052] For example, the automatic allocation of computing resources could be as follows: allocating 8-core CPUs and 16GB of memory to a customer base of one million.

[0053] Furthermore, after obtaining customer group metrics, an initial processing strategy can be constructed based on these metrics.

[0054] S102. Based on the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group, determine the initial processing strategy for the target customer group.

[0055] In existing technologies, taking a financial scenario as an example, a customer group may be in a "stable cooperation phase" but may experience multiple abnormal transfers in real time. If the binding strategy cannot be adjusted based on this real-time feature and the conventional maintenance strategy is still executed, risks may be missed or business opportunities may be lost. Therefore, in order to improve the accuracy of the initial processing strategy, it is necessary to obtain the real-time behavioral tags and external variable weights of the target customer group based on customer group indicators, as reference factors for determining the initial processing strategy, thereby improving the accuracy of the initial processing strategy determination.

[0056] Specifically, the step "determine the initial processing strategy for the target customer group based on the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group" includes: b1: Based on the decision tree algorithm, the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the life cycle of the target customer group are matched with the pre-set strategy template to determine at least one candidate processing strategy for the target customer group.

[0057] b2: For each candidate processing strategy, calculate the matching degree between the candidate processing strategy and the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group, and determine the strategy matching value corresponding to the candidate processing strategy.

[0058] b3: The strategy whose strategy matching value is greater than a preset matching value threshold among the at least one candidate processing strategy is determined as the initial processing strategy.

[0059] In one possible implementation, data such as customer transaction behavior, browsing history, and operation records can be collected in real time using a streaming computing framework (such as Flink), and real-time behavior tags can be generated based on the collected data.

[0060] For example, real-time behavior tags may include abnormal transfers, frequent browsing of a certain product, etc.

[0061] In one possible implementation, external environment data may include at least one of the following: market interest rates, industry policies, and competitor dynamics. The above data can be obtained by accessing an external environment data interface, and external variable weights can be established based on the obtained data.

[0062] Here, the weights of external variables can be determined based on the acquired external environment data and the degree of influence of the external environment data on the current customer group indicators.

[0063] For example, the weight of the external variable corresponding to the market interest rate can be set to 0.3, and the weight of the external variable corresponding to the adjustment of industry policies can be set to 0.4.

[0064] Furthermore, after obtaining customer group metrics, real-time behavioral tags of the target customer group, and weights of external variables, it is necessary to construct the lifecycle of the target customer group, and then, in combination with the lifecycle of the target customer group, construct a preliminary processing strategy.

[0065] Specifically, the lifecycle of the target customer group is determined through the following steps: c1: Divide the life cycle for the target customer group based on the primary and secondary phase method, and generate life cycle rules for each life cycle.

[0066] c2: Configure the lifecycle rules for each lifecycle and generate a JSON data body.

[0067] c3: Calculate the customer group data corresponding to the newly added or adjusted rules in the JSON data body, reuse the unchanged rules, and generate a query statement.

[0068] c4: Execute the query statement to obtain the migration data of the target customer group in each life cycle, and based on the migration data, construct the association between the target customer group and each life cycle to generate the life cycle of the target customer group.

[0069] In one possible implementation, the main stage rules and auxiliary stage rules in the main-auxiliary stage method are generated by combining preset separation rule fields through logical operations on fields controlled by metadata. Each main stage rule or auxiliary stage rule is displayed through fields controlled by metadata.

[0070] In one possible implementation, a graphical user interface can be provided on the terminal to receive user selections and determine the main stage rules and auxiliary stage rules. Specifically, this can be done via a canvas, with the left side of the canvas providing components for the main stage, main stage rules, auxiliary stage, and auxiliary stage rules. The main stage rules and auxiliary stage rules are generated by logically combining fields controlled by metadata.

[0071] Furthermore, the rules for the main stage and the auxiliary stage can be configured according to the business requirements contained in the customer group management instructions, and a JSON data body can be generated after the configuration is completed.

[0072] In one possible implementation, to improve data processing efficiency, the generated JSON data body can be optimized through trial calculations. During the optimization process, only the customer group data corresponding to the newly added or adjusted rules in the JSON data body is calculated. For customer group data that has already been calculated and has not been changed, the cached results of the previous calculations are directly reused, and the two are concatenated to generate an SQL query statement.

[0073] Here, taking a customer base of millions as an example, after applying the above calculation process, the efficiency of lifecycle calculation is improved by 50%, and for a customer base of millions, the calculation time is shortened from 1 hour to less than 30 minutes.

[0074] Furthermore, the generated SQL query statements are submitted to the big data platform for calculation to obtain customer migration data between the main and auxiliary phases; the calculated customer lists corresponding to the main and auxiliary phases are used to construct the association between the target customer group and each life cycle, and generate the life cycle of the target customer group.

[0075] Furthermore, after determining the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group, the multi-dimensional features of the target customer group are matched with the strategy template using the decision tree algorithm to generate at least one suitable candidate processing strategy.

[0076] In one possible implementation, matching can be performed based on preset matching information according to the target customer group's life cycle, external variable weights, real-time behavioral tags of the target customer group, and customer group indicators. After determining the target customer group's life cycle, the target external weight variable information, target behavioral information, and target customer group indicator information that match the target customer group are determined sequentially under the target customer group's life cycle, thereby obtaining at least one candidate processing strategy.

[0077] Specifically, the step "based on the decision tree algorithm, matching the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group with a pre-set strategy template to determine at least one candidate processing strategy for the target customer group" includes: d1: For the customer lifecycle to which the target customer group belongs, based on the external variable weights, the real-time behavioral tags of the target customer group, and the customer group indicators, determine the target external weight variable information, target behavioral information, and target customer group indicator information that match the target customer group.

[0078] d2: Based on the customer lifecycle of the target customer group, and the combination of the target external weight variable information, the target behavior information, and the target customer group indicator information, match them with the pre-set strategy template to determine the target strategy template, and generate at least one candidate processing strategy based on the target strategy template.

[0079] In one possible implementation, when determining at least one candidate processing strategy using a decision tree algorithm, since there are different decision branches in the decision tree, they can be matched sequentially according to different dimensions.

[0080] Here, the lifecycle includes the expansion phase, growth phase, stable phase, decline phase, and churn phase; external variable weights include: high external variable weight, medium external variable weight, and low external variable weight; real-time behavior tags include: abnormal behavior, positive behavior, and silent or no obvious behavior; customer group indicators include: high-quality indicators, normal indicators, and high-risk indicators.

[0081] In one possible implementation, regarding the weights of external variables, a high weight indicates a strong influence from the external environment; a medium weight indicates a certain influence; and a low weight indicates a weak influence. Higher external variable weights tend to favor risk-based strategies, while lower weights tend to favor conventional operational strategies. Regarding real-time behavioral tags, abnormal behavior can include at least one of the following: abnormal transfers, high-frequency logins, or violations; positive behavior can include at least one of the following: active trading, high-frequency browsing, or normal activity; and inactive behavior can include at least one of the following: long-term inactivity or low activity. Abnormal behavior directly leads to risk warnings; positive behavior leads to value enhancement; and inactive behavior is assessed for recovery or maintenance based on its lifecycle. Regarding customer group indicators, they can be categorized based on preset business thresholds. Specifically, high-performing indicators refer to healthy customer turnover, active transactions, and good asset quality; normal indicators refer to customer group indicators within a reasonable range; and high-risk indicators refer to excessively long customer turnover, high overdue risk, and drastic fluctuations.

[0082] In one possible implementation, firstly, the life cycle of the current target customer group is determined based on the life cycle of the target customer group; the life cycle of the customer group can be determined in three dimensions: growth period or stable period, decline period or churn period, and expansion period for new customers.

[0083] Furthermore, after determining the customer lifecycle of the target customer group, the group is then segmented based on the weights of external variables to determine whether the external variable weights corresponding to the current target customer group are high, medium, or low.

[0084] Furthermore, after determining the weights of external variables, the target customer group is then categorized based on its real-time behavioral tags to determine whether the real-time behavior of the target customer group is abnormal, positive, silent, or without obvious behavior.

[0085] Furthermore, after determining the real-time behavior of the target customer group, the target customer group is further divided according to its customer indicators and preset business thresholds to determine whether the customer indicators of the target customer group are high-quality indicators, normal indicators, or high-risk indicators.

[0086] Furthermore, after determining the customer lifecycle of the target customer group, the system determines the combination of target external weight variable information, target behavior information, and target customer group indicator information, and automatically matches the target strategy template from the pre-built strategy template library to generate at least one candidate processing strategy.

[0087] In one possible implementation, the processing strategies that can be matched with different combinations of target external weight variables, target behavior information, and target customer group indicator information may vary depending on the different customer group lifecycles.

[0088] For example, if the target customer group is in the growth or stable stage: high external variable weight + abnormal behavior match risk warning strategy; high external variable weight + normal behavior + high-quality indicators match value enhancement strategy; medium external variable weight + normal behavior + normal indicators match value enhancement strategy; low external variable weight + arbitrary behavior + normal indicators match routine maintenance strategy; high-risk indicators + arbitrary behavior match risk warning strategy.

[0089] If the target customer group is in a period of decline or churn: normal behavior should be matched with churn recovery strategies; abnormal behavior or high-risk indicators should be matched with risk warning strategies.

[0090] If the target customer group is in the process of acquiring new customers: use a routine maintenance strategy with low or medium external variable weights and normal behavior; use a risk warning strategy with abnormal behavior or high external variable weights.

[0091] Specifically, depending on which stage of the customer lifecycle the target customer group is in, if a strong external shock, high-risk indicators, and abnormal behavior occur, an emergency intervention strategy will be directly matched.

[0092] In one possible implementation, after matching through a decision tree algorithm, at least one candidate processing strategy can be obtained. Specifically, each candidate processing strategy may include strategy type, strategy execution action, strategy applicable conditions, and corresponding customer group lifecycle, target external weight variable information, target behavior information, and target customer group indicator information.

[0093] In this way, the process of obtaining candidate processing strategies through matching using the decision tree algorithm is executed automatically, and the judgment threshold and branching rules can be dynamically adjusted according to the business scenario, which can improve the efficiency and accuracy of processing strategy generation.

[0094] For example, a certain industrial finance customer group is in the "stable cooperation stage". The customer group indicator shows "stable transaction amount", but the real-time behavior label is "three consecutive abnormal cross-regional transfers". The external variable shows "the industry policy is tightening". The system uses the decision tree algorithm to match risk control strategy templates and generate a processing strategy of "suspending large credit lines + manual verification".

[0095] Specifically, the system first assesses the customer lifecycle, identifying it as a stable, cooperative customer in the main stage (stable period). This customer group is considered high-value and has high business relevance, thus the decision tree enters the dedicated branch for stable period customers. The strategy under this branch prioritizes risk control while avoiding excessive intervention in normal business operations. Regarding external variable weights, tightening industry policies are considered a strong external influencing factor, resulting in a high weight for the external variable. The decision tree further converges towards risk warning and risk control strategy branches. The real-time behavioral tag is continuous abnormal cross-regional transfers, a clearly high-risk behavior, leading the decision tree to directly lock onto a risk control strategy template. Since the customer group's transaction amount is stable with no significant deterioration, the highest-level emergency shutdown strategy is not triggered; only moderate risk control is implemented. Based on the combined results of the four feature matching dimensions, the system matches risk control strategy templates from the strategy template library, ultimately generating a processing strategy of suspending large credit lines and conducting manual verification.

[0096] In one possible implementation, in order to improve the accuracy of the determined initial processing strategy, at least one candidate processing strategy may be determined first based on customer group indicators, real-time behavioral tags of the target customer group, external variable weights, and the life cycle of the target customer group. For each good candidate processing strategy, a strategy matching value corresponding to the candidate processing strategy is determined, and the candidate processing strategies are filtered according to the corresponding strategy matching value to obtain the initial processing strategy.

[0097] In one possible implementation, after identifying at least one candidate processing strategy, the cosine similarity algorithm can be used to calculate the matching degree between each candidate processing strategy and the customer group indicators, the real-time behavioral labels of the target customer group, the weights of external variables, and the lifecycle of the target customer group, thereby determining the strategy matching value corresponding to each candidate processing strategy.

[0098] Furthermore, the strategies whose matching values ​​are greater than a preset matching value threshold are determined as the initial processing strategies.

[0099] Here, the preset matching value threshold can be set according to the strategy filtering requirements. For example, the preset matching value threshold can be set to 0.8.

[0100] Furthermore, in order to ensure that the correspondence between the target customer group's data in each dimension and the matching initial processing strategy can be intuitively observed, the initial processing strategy can be bound to the corresponding customer group indicators, real-time behavior tags, and external variables to generate a correspondence list containing trigger conditions, execution strategies, and priorities. In subsequent processing, the processing strategy to be executed can be selected based on the corresponding trigger conditions, etc.

[0101] Here, the priority is determined based on the business value of the corresponding business in the customer group management instructions. For example, risk prevention and control has a higher priority than marketing and promotion.

[0102] Furthermore, after determining the initial processing strategy, once the customer group indicators reach the preset warning threshold, the corresponding initial processing strategy can be executed, and the corresponding strategy execution result can be determined, so as to complete the subsequent strategy analysis based on the strategy execution result.

[0103] S103. In response to the customer group indicator reaching a preset warning threshold, execute the initial processing strategy corresponding to the customer group indicator and determine the strategy execution result; wherein, the preset warning threshold is dynamically determined based on the life cycle stage corresponding to the target customer group and the weight of the external variable.

[0104] In existing technologies, predictions are typically triggered using fixed warning thresholds. However, this method fails to adaptively adjust for dynamic factors such as customer lifecycle migration speed and market fluctuations. For instance, during e-commerce promotional seasons, customer activity frequency metrics are significantly higher than normal, potentially leading to numerous false warnings triggered by a fixed threshold. Conversely, during industry downturns, the rate of decline in customer transaction volume accelerates, and the fixed threshold may fail to adjust in a timely manner, resulting in delayed warnings and impacting the accuracy of the strategy's alerts.

[0105] In this embodiment, dynamic warning thresholds can be set for each lifecycle to match different real-world scenarios, thereby improving the accuracy of policy warnings.

[0106] Specifically, deep learning models can be trained based on historical customer group indicator data, life cycle migration speed, market environment fluctuation data, etc., to obtain an adaptive early warning value calculation model.

[0107] Furthermore, based on manually configured or imported customer groups, and combined with the historical fluctuation range of customer group indicators, a first adaptive early warning threshold can be generated. For customer groups corresponding to each main and auxiliary stage in the life cycle, a second adaptive early warning threshold can be generated based on the model and the stage migration speed and the weight of external environmental variables. The triggering conditions for setting the early warning threshold can be dynamically adjusted. For example, if the change in customer group migration speed exceeds 20% or the adjustment of the weight of external environmental variables exceeds 0.1, the early warning threshold will be automatically recalculated and updated when the triggering conditions are met.

[0108] Furthermore, if it is determined that the customer group indicator reaches the preset warning threshold (such as the first adaptive warning threshold or the second adaptive warning threshold mentioned above), the initial processing strategy corresponding to the customer group indicator is executed, and the strategy execution result is determined.

[0109] For example, during promotional seasons, when the external environmental variable is a surge in transaction volume, the second adaptive warning threshold for the e-commerce customer activity index is automatically increased by 30% to avoid false triggers; while during industry downturns, when the external environmental variable is a decline in transaction volume, the second adaptive warning threshold is automatically decreased by 25% to ensure timely warnings.

[0110] Furthermore, after determining the strategy execution results after executing the initial processing strategy, the strategy execution score can be calculated and determined on different evaluation dimensions based on the strategy execution results, and the target recommended strategy can be adjusted and / or stored based on the strategy execution evaluation results of the selected target recommended strategy.

[0111] S104. Based on the strategy execution results, perform weighted calculations on multiple evaluation dimensions to determine the strategy execution score, and adjust and / or store the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy; wherein, the target recommendation strategy is obtained based on the strategy execution score.

[0112] Current technologies lack a complete closed-loop iterative mechanism for strategy execution effectiveness. Effectiveness evaluation only focuses on whether metrics return to normal ranges, without considering business benefits. For example, core value dimensions such as average revenue per customer (ARPC) growth, customer experience, and complaint rate reduction are not considered. For instance, while a marketing strategy might bring customer activity metrics back to normal, it could lead to an increase in the average complaint rate per customer. Existing solutions fail to identify this problem and still classify it as an effective strategy, impacting the accuracy of strategy selection.

[0113] In one possible implementation, the rating dimensions may include at least one of the following: indicator improvement dimension, business value dimension, and customer experience dimension. Specifically, the indicator improvement dimension may include parameters such as the speed at which customer indicators return to normal range and the magnitude of improvement in customer indicators; the business value dimension may include parameters such as the growth in average revenue per customer, the percentage reduction in costs, and the reduction in the number of risk events; the customer experience dimension may include parameters such as customer complaint rate, satisfaction score, and changes in business processing time. This establishes a lineage between indicators and the initial marketing strategy, monitors changes in customer indicators, real-time changes in behavioral tags, and the impact of external environmental variables before, during, and after the implementation of the initial marketing strategy, and determines the strategy execution score for the initial handling strategy based on the above evaluation dimensions.

[0114] Furthermore, the weights of each evaluation dimension can be determined based on the analytic hierarchy process (AHP). For example, the weight of the indicator improvement dimension is 0.3, the weight of the business value dimension is 0.4, and the weight of the customer experience dimension is 0.3. Then, based on the dimension parameters of the initial processing strategy under each evaluation dimension, the initial processing strategy is weighted and scored to obtain the strategy execution score.

[0115] In one possible implementation, after determining the policy execution score of the initial processing strategy, the initial processing strategy can be filtered based on the policy execution score to obtain the target recommended strategy.

[0116] Specifically, the strategy processing method further includes: e1: Select the processing strategies whose strategy execution scores are greater than the preset score threshold as candidate recommendation strategies.

[0117] e2: If the candidate recommendation strategy conflicts with other recommendation strategies, the target recommendation strategy is determined based on the preset strategy priority.

[0118] e3: If the candidate recommendation strategy does not conflict with other recommendation strategies, the candidate recommendation strategy shall be determined as the target recommendation strategy.

[0119] In one possible implementation, processing strategies with execution scores greater than a preset score threshold can be selected as candidate recommendation strategies based on a pre-set preset score threshold. For each candidate recommendation strategy, it can be determined whether the current candidate recommendation strategy conflicts with other recommendation strategies. If the candidate recommendation strategy does not conflict with other recommendation strategies, the candidate recommendation strategy is determined as the target recommendation strategy.

[0120] In another possible implementation, if a candidate recommendation strategy conflicts with other recommendation strategies, the strategy with higher priority can be retained as the target recommendation strategy according to the previously determined priority rules; alternatively, manual intervention can be used to coordinate conflicts among target recommendation strategies and determine the target recommendation strategy.

[0121] Here, a conflict between candidate recommendation strategies and other recommendation strategies can occur when both credit increase and credit suspension strategies are executed simultaneously. Priority rules can be set according to business value; for example, risk control has a higher priority than customer maintenance, and customer maintenance has a higher priority than marketing promotion.

[0122] In one possible implementation, after the target recommendation strategy is determined, the target recommendation strategy and the initial processing strategy can be displayed through the graphical user interface of the terminal. The initial processing strategy can be displayed in descending order of its execution score, while the target recommendation strategy is highlighted.

[0123] Furthermore, after determining the target recommendation strategy, the target recommendation strategy is executed, and the target recommendation strategy is adjusted and / or stored based on the strategy execution evaluation results.

[0124] Specifically, the step "adjusting and / or storing the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy" includes: f1: If the strategy execution evaluation result of the target recommendation strategy meets the preset evaluation criteria, store the target recommendation strategy and the customer characteristics and external variable information of the target customer group corresponding to the target recommendation strategy, so as to reuse the target recommendation strategy when processing the same customer characteristics and external variable information management instructions.

[0125] f2: If the strategy execution evaluation result of the target recommendation strategy does not meet the preset evaluation criteria, adjust the abnormal information in the target recommendation strategy to obtain an updated target recommendation strategy, and evaluate the strategy execution evaluation result of the updated target recommendation strategy again until the strategy execution evaluation result of the target recommendation strategy meets the preset evaluation criteria, and store the updated target recommendation strategy.

[0126] Here, the target recommendation strategy is piloted, recording the initial distribution of customer indicators, real-time behavioral tags, and external environmental variables before execution, the dynamic changes during execution, and the final status data after execution. A multi-dimensional quantitative evaluation report is generated by combining the above data. The report includes core content such as indicator improvement effect, business value enhancement, and changes in customer experience. At the same time, manual evaluation intervention analysis is combined with the quantitative evaluation report and manual scoring to generate strategy execution evaluation results.

[0127] Here, the preset evaluation criteria can be set based on historical strategy execution data, strategy execution objectives, business requirements, etc.

[0128] In one possible implementation, if the strategy execution evaluation result is determined to meet the preset evaluation criteria, the target recommendation strategy and the customer characteristics and external variable information of the target customer group corresponding to the target recommendation strategy are stored in the strategy library. If the same customer characteristics and external variable information management instructions are processed later, the stored target recommendation strategy does not need to be calculated again, and the data processing time is directly reused, thereby reducing data processing time and improving the efficiency of strategy generation and processing.

[0129] In another possible implementation, if it is determined that the strategy execution evaluation result does not meet the preset evaluation criteria, then based on the abnormal information in the evaluation report, such as a decline in customer experience, the strategy binding logic, warning threshold, or evaluation dimension weight is adjusted to obtain an updated target recommendation strategy. The target recommendation strategy is then executed again, and the strategy execution evaluation result of the updated target recommendation strategy is evaluated again, until the strategy execution evaluation result of the target recommendation strategy meets the preset evaluation criteria, and the updated target recommendation strategy is stored.

[0130] The customer group indicator strategy processing method provided in this application embodiment obtains the target customer group based on the acquired customer group management instructions, and obtains the customer group indicators based on at least one target indicator in the customer group management instructions and a pre-built indicator metadata pool; determines the initial processing strategy for the target customer group based on the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group; in response to the customer group indicators reaching a preset warning threshold, executes the initial processing strategy corresponding to the customer group indicators, and determines the strategy execution result; based on the strategy execution result, performs weighted calculations on multiple evaluation dimensions to determine the strategy execution score, and adjusts and / or stores the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy. In this way, after determining the customer group management instructions, the customer group indicators are quickly obtained based on the indicator metadata pool. On the basis of processing the basic customer group indicators, real-time behavior tags and external variable weights are introduced to formulate the initial processing strategy, so as to dynamically adapt to the business scenarios of implementation and adjustment. At the same time, during the strategy warning process, the dynamic warning threshold is used to ensure that the warning process is more accurate. Furthermore, the processing strategy is adjusted and evaluated based on the multi-dimensional evaluation results, which helps to improve the efficiency and accuracy of strategy processing.

[0131] Based on the same inventive concept, this application also provides a customer group indicator strategy processing device corresponding to the customer group indicator strategy processing method. Since the principle of the device in this application is similar to the above-mentioned customer group indicator strategy processing method in this application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0132] Please see Figure 2 , Figure 2 This is a schematic diagram of the structure of a customer group indicator strategy processing device provided in an embodiment of this application. Figure 2 As shown, the strategy processing device 200 includes: The customer group indicator acquisition module 210 is used to acquire target customer groups based on the acquired customer group management instructions, and acquire customer group indicators based on at least one target indicator in the customer group management instructions and a pre-built indicator metadata pool. The initial strategy construction module 220 is used to determine an initial processing strategy for the target customer group based on the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group. The strategy execution module 230 is used to execute the initial processing strategy corresponding to the customer group indicator in response to the customer group indicator reaching a preset warning threshold, and determine the strategy execution result; wherein, the preset warning threshold is dynamically determined based on the life cycle stage corresponding to the target customer group and the weight of the external variable; The recommendation strategy feedback module 240 is used to perform weighted calculations on multiple evaluation dimensions based on the strategy execution results to determine the strategy execution score, and to adjust and / or store the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy; wherein, the target recommendation strategy is obtained based on the strategy execution score.

[0133] In one possible implementation, when the customer group indicator acquisition module 210 acquires customer group indicators based on at least one target indicator in the customer group management instruction and a pre-built indicator metadata pool, the customer group indicator acquisition module 210 is used to: Based on at least one target indicator in the customer group management instruction and the calculation logic template and configuration variables configured in the indicator metadata pool, a query statement is automatically generated. Execute the query statement to calculate the customer group data and obtain the customer group indicators.

[0134] In one possible implementation, when the initial strategy construction module 220 determines an initial processing strategy for the target customer group based on the customer group metrics, the real-time behavioral tags of the target customer group, external variable weights, and the lifecycle of the target customer group, the initial strategy construction module 220 is used to: Based on the decision tree algorithm, the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the life cycle of the target customer group are matched with a pre-set strategy template to determine at least one candidate processing strategy for the target customer group. For each candidate processing strategy, calculate the matching degree between the candidate processing strategy and the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group, and determine the strategy matching value corresponding to the candidate processing strategy. The strategy whose matching value is greater than a preset matching value threshold among the at least one candidate processing strategy is determined as the initial processing strategy.

[0135] In one possible implementation, the strategy processing device 200 further includes a lifecycle determination module (not shown in the figure), the lifecycle determination module being used for: The target customer group is divided into lifecycles based on the primary and secondary phases method, and lifecycle rules are generated for each lifecycle. Configure lifecycle rules for each lifecycle stage and generate a JSON data body; The query statement is generated by calculating the customer group data corresponding to the newly added or adjusted rules in the JSON data body and reusing the unchanged rules. The query statement is executed to obtain the migration data of the target customer group at each life cycle. Based on the migration data, the association between the target customer group and each life cycle is constructed to generate the life cycle of the target customer group.

[0136] In one possible implementation, when the initial strategy construction module 220 is used to match the customer group indicators, the real-time behavioral labels of the target customer group, the weights of external variables, and the lifecycle of the target customer group with a pre-set strategy template based on a decision tree algorithm to determine at least one candidate processing strategy for the target customer group, the initial strategy construction module 220 is used to: For the customer lifecycle to which the target customer group belongs, based on the external variable weights, the real-time behavioral tags of the target customer group, and the customer group indicators, determine the target external weight variable information, target behavioral information, and target customer group indicator information that match the target customer group; Based on the customer lifecycle of the target customer group, and the combination of the target external weight variable information, the target behavior information, and the target customer group indicator information, the information is matched with a pre-set strategy template to determine the target strategy template, and at least one candidate processing strategy is generated based on the target strategy template.

[0137] In one possible implementation, the strategy processing device 200 further includes a recommendation strategy determination module (not shown in the figure), the recommendation strategy determination module being used for: The processing strategies whose execution scores are greater than the preset score threshold are identified as candidate recommendation strategies; If the candidate recommendation strategy conflicts with other recommendation strategies, the target recommendation strategy is determined based on the preset strategy priority. If the candidate recommendation strategy does not conflict with other recommendation strategies, the candidate recommendation strategy is determined as the target recommendation strategy.

[0138] In one possible implementation, when the recommendation strategy feedback module 240 adjusts and / or stores the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy, the recommendation strategy feedback module 240 is used to: If the strategy execution evaluation result of the target recommendation strategy meets the preset evaluation criteria, the target recommendation strategy and the customer characteristics and external variable information of the target customer group corresponding to the target recommendation strategy are stored so that the target recommendation strategy can be reused when processing the same customer characteristics and external variable information management instructions. If the strategy execution evaluation result of the target recommendation strategy does not meet the preset evaluation criteria, the abnormal information in the target recommendation strategy is adjusted to obtain an updated target recommendation strategy, and the strategy execution evaluation result of the updated target recommendation strategy is evaluated again until the strategy execution evaluation result of the target recommendation strategy meets the preset evaluation criteria, and the updated target recommendation strategy is stored.

[0139] The customer group indicator strategy processing device provided in this application embodiment obtains a target customer group based on the acquired customer group management instructions, and obtains customer group indicators based on at least one target indicator in the customer group management instructions and a pre-built indicator metadata pool; determines an initial processing strategy for the target customer group based on the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group; in response to the customer group indicators reaching a preset warning threshold, executes the initial processing strategy corresponding to the customer group indicators and determines the strategy execution result; based on the strategy execution result, performs weighted calculations on multiple evaluation dimensions to determine the strategy execution score, and adjusts and / or stores the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy. In this way, after determining the customer group management instructions, the customer group indicators are quickly obtained based on the indicator metadata pool. On the basis of processing the basic customer group indicators, real-time behavior tags and external variable weights are introduced to formulate the initial processing strategy, so as to dynamically adapt to the business scenarios of implementation and adjustment. At the same time, during the strategy warning process, the dynamic warning threshold is used to ensure that the warning process is more accurate. Furthermore, the processing strategy is adjusted and evaluated based on the multi-dimensional evaluation results, which helps to improve the efficiency and accuracy of strategy processing. Please see Figure 3 , Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 3 As shown, the electronic device 300 includes a processor 310, a memory 320, and a bus 330.

[0140] The memory 320 stores machine-readable instructions executable by the processor 310. When the electronic device 300 is running, the processor 310 and the memory 320 communicate via the bus 330. When the machine-readable instructions are executed by the processor 310, they can perform the operations described above. Figure 1 The steps of the customer group indicator strategy processing method in the method embodiment shown are described in detail in the method embodiment, and will not be repeated here.

[0141] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, can perform the above-described actions. Figure 1The steps of the customer group indicator strategy processing method in the method embodiment shown are described in detail in the method embodiment, and will not be repeated here.

[0142] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0143] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the shown or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0144] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0145] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0146] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0147] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The scope of protection of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A strategy processing method for customer group indicators, characterized in that, The strategy processing method includes: Based on the obtained customer group management instructions, the target customer group is obtained, and based on at least one target indicator in the customer group management instructions and a pre-built indicator metadata pool, the customer group indicators are obtained. Based on the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group, an initial processing strategy for the target customer group is determined. In response to the customer group indicator reaching a preset warning threshold, the initial processing strategy corresponding to the customer group indicator is executed, and the strategy execution result is determined; wherein, the preset warning threshold is dynamically determined based on the life cycle stage corresponding to the target customer group and the weight of the external variable; Based on the strategy execution results, a weighted calculation is performed on multiple evaluation dimensions to determine the strategy execution score. Based on the strategy execution evaluation results of the selected target recommendation strategies, the target recommendation strategies are adjusted and / or stored. The target recommendation strategies are obtained based on the strategy execution scores.

2. The strategy processing method according to claim 1, characterized in that, The process of obtaining customer group metrics based on at least one target metric in the customer group management instruction and a pre-built metric metadata pool includes: Based on at least one target indicator in the customer group management instruction and the calculation logic template and configuration variables configured in the indicator metadata pool, a query statement is automatically generated. The query statement is executed to calculate the customer group data and obtain the customer group indicators.

3. The strategy processing method according to claim 1, characterized in that, The initial processing strategy for the target customer group is determined based on the customer group metrics, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group, including: Based on the decision tree algorithm, the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the life cycle of the target customer group are matched with a pre-set strategy template to determine at least one candidate processing strategy for the target customer group. For each candidate processing strategy, calculate the matching degree between the candidate processing strategy and the customer group indicators, the real-time behavior tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group, and determine the strategy matching value corresponding to the candidate processing strategy. The strategy whose strategy matching value is greater than a preset matching value threshold among the at least one candidate processing strategies is determined as the initial processing strategy.

4. The strategy processing method according to claim 1, characterized in that, The lifecycle of the target customer group is determined through the following steps: The target customer group is divided into lifecycles based on the primary and secondary phases method, and lifecycle rules are generated for each lifecycle. Configure lifecycle rules for each lifecycle stage and generate a JSON data body; The query statement is generated by calculating the customer group data corresponding to the newly added or adjusted rules in the JSON data body and reusing the unchanged rules. The query statement is executed to obtain the migration data of the target customer group at each life cycle. Based on the migration data, the association between the target customer group and each life cycle is constructed to generate the life cycle of the target customer group.

5. The strategy processing method according to claim 3, characterized in that, The decision tree algorithm-based approach matches the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group with a pre-set strategy template to determine at least one candidate processing strategy for the target customer group, including: For the customer lifecycle to which the target customer group belongs, based on the external variable weights, the real-time behavioral tags of the target customer group, and the customer group indicators, determine the target external weight variable information, target behavioral information, and target customer group indicator information that match the target customer group; Based on the customer lifecycle of the target customer group, and the combination of the target external weight variable information, the target behavior information, and the target customer group indicator information, the information is matched with a pre-set strategy template to determine the target strategy template, and at least one candidate processing strategy is generated based on the target strategy template.

6. The strategy processing method according to claim 1, characterized in that, The strategy processing method further includes: The processing strategies whose execution scores are greater than the preset score threshold are identified as candidate recommendation strategies; If the candidate recommendation strategy conflicts with other recommendation strategies, the target recommendation strategy is determined based on the preset strategy priority. If the candidate recommendation strategy does not conflict with other recommendation strategies, the candidate recommendation strategy is determined as the target recommendation strategy.

7. The strategy processing method according to claim 1, characterized in that, The adjustment and / or storage of the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy includes: If the strategy execution evaluation result of the target recommendation strategy meets the preset evaluation criteria, the target recommendation strategy and the customer characteristics and external variable information of the target customer group corresponding to the target recommendation strategy are stored so that the target recommendation strategy can be reused when processing the same customer characteristics and external variable information management instructions. If the strategy execution evaluation result of the target recommendation strategy does not meet the preset evaluation criteria, the abnormal information in the target recommendation strategy is adjusted to obtain an updated target recommendation strategy, and the strategy execution evaluation result of the updated target recommendation strategy is evaluated again until the strategy execution evaluation result of the target recommendation strategy meets the preset evaluation criteria, and the updated target recommendation strategy is stored.

8. A strategy processing device for customer group indicators, characterized in that, The strategy processing device includes: The customer group indicator acquisition module is used to acquire target customer groups based on the acquired customer group management instructions, and to acquire customer group indicators based on at least one target indicator in the customer group management instructions and a pre-built indicator metadata pool. The initial strategy construction module is used to determine the initial processing strategy for the target customer group based on the customer group indicators, the real-time behavioral tags of the target customer group, the weights of external variables, and the lifecycle of the target customer group. The strategy execution module is used to execute the initial processing strategy corresponding to the customer group indicator in response to the customer group indicator reaching a preset warning threshold, and determine the strategy execution result; wherein, the preset warning threshold is dynamically determined based on the life cycle stage corresponding to the target customer group and the weight of the external variable; The recommendation strategy feedback module is used to perform weighted calculations on multiple evaluation dimensions based on the strategy execution results to determine the strategy execution score, and to adjust and / or store the target recommendation strategy based on the strategy execution evaluation results of the selected target recommendation strategy; wherein, the target recommendation strategy is obtained based on the strategy execution score.

9. An electronic device, characterized in that, include: The device includes a processor, a storage medium, and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the strategy processing method for customer metrics as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the strategy processing method for customer group indicators as described in any one of claims 1 to 7.

Images