Recommendation method, device, server and storage medium for financial marketing model

Abstract

The present application provides a method, device, server and storage medium for recommending a financial marketing model. The method includes obtaining characteristic information of a target financial marketing model, obtaining delivery list information and converted list information of multiple historical financial marketing models from a preset database based on the characteristic information; inputting the delivery list information and converted list information of multiple historical financial marketing models into a complete preprocessing model to obtain multiple mutually exclusive delivery list information; obtaining the conversion rates of multiple historical financial marketing models based on the multiple mutually exclusive delivery list information and converted list information; determining the estimated conversion rate of the target financial marketing model based on the Bayesian formula and the conversion rates of multiple historical financial marketing models, and recommending the target financial marketing model that meets the recommendation conditions to the terminal device based on the estimated conversion rate. By normalizing the basic data, the deviation of the estimated conversion rate is reduced, and the digital financial model is accurately delivered to the target customer group.

Description

Technical Field

[0001] The present application relates to the field of digital finance, and in particular to a method, device, server, and storage medium for recommending a financial marketing model. Background Art

[0002] In a highly interconnected digital world, various internet technologies are maturing, and many companies are embarking on digital transformation strategies. The foundation of digitalization lies in a vast amount of complex, highly business-relevant basic data. This data is processed, analyzed, and presented to generate optimization plans. These plans require predictions of their implementation effectiveness to evaluate them. In financial marketing projects, multiple marketing recommendation models can be optimized and constructed by combining a range of customer-authorized characteristics, such as historical customer behavior and asset status. This requires predicting the proportion of potential customers who will actually purchase from a marketing recommendation model, i.e., the model's conversion rate. This serves as an indicator for evaluating the conversion effectiveness of the marketing recommendation model, thereby enabling precision marketing.

[0003] While a large amount of complex data provides comprehensive feedback on the actual situation, it also causes data redundancy, resulting in a larger deviation between the predicted implementation effect of the plan and the actual implementation effect. It is difficult to accurately quantify and evaluate the implementation effect of the plan, and thus it is impossible to accurately push appropriate financial marketing models to customers. Summary of the Invention

[0004] The main purpose of this application is to provide a recommendation method, device, server and storage medium for a financial marketing model, aiming to reduce the deviation of the model's estimated conversion rate by normalizing the basic data, and accurately deliver the model to the target customer group, thereby improving the accuracy of the evaluation results and the recommendation efficiency of the digital financial marketing model.

[0005] In a first aspect, the present application provides a method for recommending a financial marketing model, the method comprising the following steps:

[0006] Acquiring characteristic information of a target financial marketing model, and acquiring a plurality of related historical financial marketing models from the preset database based on the characteristic information;

[0007] Obtaining delivery list information and converted list information of multiple historical financial marketing models;

[0008] Inputting the delivery list information of the plurality of historical financial marketing models and the converted list information into a complete pre-processing model to obtain a plurality of mutually exclusive delivery list information, wherein the complete pre-processing model is used to assign duplicate list information between the plurality of historical financial marketing models to the delivery list information of the corresponding historical financial marketing model;

[0009] Determining the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information, thereby obtaining the conversion rates of multiple historical financial marketing models;

[0010] Determining an estimated conversion rate of a target financial marketing model based on the Bayesian formula and the conversion rates of the multiple historical financial marketing models, and determining whether the target financial marketing model meets the recommendation conditions based on the estimated conversion rate;

[0011] If the target financial marketing model meets the recommendation condition, the target financial marketing model is recommended to the terminal device.

[0012] In a second aspect, the present application further provides a financial marketing model recommendation device, the financial marketing model recommendation device comprising:

[0013] a feature information acquisition module, configured to acquire feature information of a target financial marketing model, and acquire a plurality of related historical financial marketing models from the preset database based on the feature information;

[0014] A list information acquisition module, used to obtain the delivery list information and converted list information of multiple historical financial marketing models;

[0015] An information mutual exclusion processing module, configured to input the delivery list information of the plurality of historical financial marketing models and the converted list information into a complete pre-processing model to obtain a plurality of mutually exclusive delivery list information, wherein the complete pre-processing model is configured to allocate duplicate list information between the plurality of historical financial marketing models to the delivery list information of the corresponding historical financial marketing model;

[0016] A historical conversion rate module, configured to determine the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information, thereby obtaining the conversion rates of multiple historical financial marketing models;

[0017] a conversion rate estimation module, configured to determine an estimated conversion rate of a target financial marketing model based on a Bayesian formula and the conversion rates of the plurality of historical financial marketing models, and determine whether the target financial marketing model meets a recommendation condition based on the estimated conversion rate;

[0018] The target model recommendation module is configured to recommend the target financial marketing model to the terminal device if the target financial marketing model meets the recommendation conditions.

[0019] In a third aspect, the present application further provides a server, the server comprising a memory, a processor, and a preset database;

[0020] The memory is used to store computer programs;

[0021] The processor is used to execute the computer program and implement the recommendation method of the financial marketing model provided in any one of the embodiments of the present application when executing the computer program.

[0022] In a fourth aspect, the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the processor implements the recommendation method of the financial marketing model described in any one of the embodiments of the present application.

[0023] The present application provides a method, device, server, and storage medium for recommending a financial marketing model. The present application obtains characteristic information of a target financial marketing model, obtains multiple related historical financial marketing models from a preset database based on the characteristic information, obtains delivery list information and converted list information of multiple historical financial marketing models, inputs the delivery list information and converted list information of multiple historical financial marketing models into a complete preprocessing model to obtain multiple mutually exclusive delivery list information, and the complete preprocessing model is used to assign duplicate list information between the multiple historical financial marketing models to the delivery list information of the corresponding historical financial marketing models; determines the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information to obtain the conversion rates of multiple historical financial marketing models; determines an estimated conversion rate of the target financial marketing model based on a Bayesian formula and the conversion rates of the multiple historical financial marketing models, and determines whether the target financial marketing model meets the recommendation conditions based on the estimated conversion rate; if the target financial marketing model meets the recommendation conditions, recommends the target financial marketing model to a terminal device. By normalizing the basic data, we can reduce the deviation of the model's estimated conversion rate and accurately deliver the model to the target customer group, thereby improving the accuracy of the evaluation results and the recommendation efficiency of the digital financial marketing model. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0025] Figure 1 This is a flowchart of a method for recommending a financial marketing model provided in an embodiment of the present application;

[0026] Figure 2 A diagram showing a usage scenario of a recommendation method for a financial marketing model provided in an embodiment of the present application;

[0027] Figure 3This is a flowchart of another method for recommending a financial marketing model provided in an embodiment of the present application;

[0028] Figure 4 A schematic block diagram of a financial marketing model recommendation device provided in an embodiment of the present application;

[0029] Figure 5 This is a schematic block diagram of the structure of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0031] The flowcharts shown in the accompanying drawings are for illustrative purposes only and do not necessarily include all contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined, or partially merged, so the actual execution order may vary depending on the actual situation.

[0032] The term "and / or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0033] The following describes some embodiments of the present application in detail with reference to the accompanying drawings. In the absence of conflict, the following embodiments and features therein may be combined with each other.

[0034] See also Figure 1 , Figure 1 This is a flowchart of a method for recommending a financial marketing model, as provided in an embodiment of the present application. This method aims to reduce the deviation in the model's estimated conversion rate by normalizing the underlying data and accurately targeting the model to the target customer base, thereby improving the accuracy of the evaluation results and the efficiency of the digital financial marketing model recommendation.

[0035] See also Figure 2 , Figure 2 This is a usage scenario diagram of a method for recommending a financial marketing model provided in an embodiment of the present application. The method for recommending a financial marketing model provided in an embodiment of the present application is applied to a server, which corresponds to a preset database and can establish a communication connection with a user terminal.

[0036] like Figure 1 As shown, the recommendation method of the financial marketing model includes steps S101 to S106.

[0037] S101 : Acquire characteristic information of a target financial marketing model, and acquire a plurality of related historical financial marketing models from the preset database according to the characteristic information.

[0038] The characteristic information is determined based on the target financial marketing model's application scenario and target customer groups. For example, in a digital financial marketing model for a financial marketing project, in the credit card account opening model application scenario, if the target group A is labeled as a fashionable car owner, the characteristic information of the target financial marketing model could be the car category and the credit card category. The corresponding marketing plan for the target financial marketing model could be to associate the benefits of interest to group A with gas cards and car wash coupons, and to associate the product with a car owner-themed credit card.

[0039] Specifically, if Figure 2 As shown, the target financial marketing model's characteristic information is comprehensively analyzed based on its application scenarios, customer groups, and other information. This characteristic information is then compared with historical financial marketing models in a pre-set database on the server to match historical financial marketing models with business relevance. To improve conversion rate accuracy, historical financial marketing models with a high degree of business relevance to the target financial marketing model are selected. In specific implementations, the closeness of the business relevance can also be set based on needs. It should be noted that a certain number of historical financial marketing models is required to accurately predict conversion rates, and the specific number can be flexibly set based on needs.

[0040] S102: Obtaining delivery list information and converted list information of a plurality of historical financial marketing models.

[0041] The placement list information is the information of all users who have been placed during the implementation of the historical financial marketing model, and the converted list information is the information of users who have engaged in consumption behavior based on the historical financial marketing model. It should be noted that in the preset database, users in each list can be linked by user ID.

[0042] Specifically, the user information and converted user information of multiple historical financial marketing models that are highly business-related to the target financial marketing model are obtained as basic data for predicting the conversion rate of the target financial marketing model.

[0043] S103. Input the delivery list information of multiple historical financial marketing models and the converted list information into a complete preprocessing model to obtain multiple mutually exclusive delivery list information. The complete preprocessing model is used to assign the repeated list information between the multiple historical financial marketing models to the delivery list information of the corresponding historical financial marketing model.

[0044] Among them, the complete preprocessing model conducts a comprehensive analysis based on the model's feature information, the delivery list information of each historical financial marketing model, the converted list information, and the historical delivery behavior of users contained in the delivery list information, and decides to assign duplicate list information to the delivery list information of the corresponding historical financial marketing model, and deletes the duplicate parts in other models accordingly.

[0045] For example, there is some duplication of delivery list information between historical financial marketing model A and historical financial marketing model B. After comprehensive analysis, the complete preprocessing model can assign all the duplicate delivery list information to the delivery list information of historical financial marketing model A, or assign part of the delivery list information to the delivery list information of historical financial marketing model A and the other part to the delivery list information of historical financial marketing model B.

[0046] It should be noted that the complete pre-processing model can comprehensively analyze various dimensions, such as the number of users in the delivery list information of each historical financial marketing model, the user's historical delivery behavior, and the repetition rate of the delivery list information of each historical financial marketing model. This allows for mutually exclusive processing of each model while improving the accuracy of subsequent conversion rate calculations. For example, if historical financial marketing model A and historical financial marketing model B have duplicate delivery list information, and a large amount of user information in historical financial marketing model B duplicates the delivery list information of other historical financial marketing models, that is, the repetition rate of the delivery list information of historical financial marketing model B is higher, then the duplicated portion can be preferentially assigned to model B.

[0047] Specifically, the delivery list information from multiple historical financial marketing models is compared, and user information within these lists is pre-processed for completeness, ensuring that any user appears only on the delivery list of a single historical financial marketing model. It should be noted that completeness pre-processing is intended to improve grouping completeness and cannot guarantee a strictly complete event group. However, the larger the sample size and the more refined the grouping, the higher the probability of a complete event, and the higher the accuracy of the data obtained through Bayes' theorem.

[0048] For example, in a financial marketing project, multiple marketing recommendation models will be pushed for different information of the same customer to meet the different needs of customers in different scenarios. Therefore, the same customer may be on the delivery list information of multiple historical financial marketing models. In order to improve the accuracy of the predicted conversion rate, the delivery list information of multiple historical financial marketing models needs to be pre-processed for completeness.

[0049] In some embodiments, the multiple historical financial marketing models include at least a first historical financial marketing model and a second historical financial marketing model; the inputting of the delivery list information and the converted list information of the multiple historical financial marketing models into a complete preprocessing model includes: taking the intersection of the delivery list information of the first historical financial marketing model and the delivery list information of the second historical financial marketing model to obtain first repeated delivery list information; analyzing the delivery list information of the first historical financial marketing model, the delivery list information of the second historical financial marketing model, and the converted list information to obtain a first allocation strategy; and deleting the first repeated delivery list information from the corresponding delivery list information based on the first allocation strategy.

[0050] The first allocation strategy includes deleting some or all duplicate placement list information from the placement list information of the first historical financial marketing model or the second historical financial marketing model. It should be understood that the first allocation strategy may include deleting duplicate placement list information from the placement list information in the first historical financial marketing model, or deleting some duplicate placement list information from the placement list information in the first historical financial marketing model and then deleting another portion of duplicate placement list information from the placement list information in the second historical financial marketing model.

[0051] Specifically, in order to ensure that any user only appears in the delivery list information of one historical financial marketing model, the delivery list information of two historical financial marketing models is compared, and the duplicate user information is deleted from the corresponding delivery list information based on the first allocation strategy output by the completeness preprocessing model.

[0052] In some embodiments, the multiple historical financial marketing models also include a third historical financial marketing model; the inputting of the delivery list information and the converted list information of the multiple historical financial marketing models into the complete preprocessing model also includes: taking the union of the delivery list information of the first historical financial marketing model and the delivery list information of the second historical financial marketing model to obtain the delivered list information; taking the intersection of the delivery list information of the third historical financial marketing model and the delivered list information to obtain the second repeated delivery list information; analyzing the delivery list information of the first historical financial marketing model, the delivery list information of the second historical financial marketing model, the delivery list information of the third historical financial marketing model, and the converted list information to obtain a second allocation strategy; and deleting the second repeated delivery list information from the corresponding delivery list information based on the second allocation strategy.

[0053] The second allocation strategy includes which model to delete which portion of duplicate delivery list information from. It should be understood that the second allocation strategy can be to select and delete portions of duplicate delivery list information from the first, second, and third historical financial marketing models, respectively, thereby achieving a state where duplicate delivery list information is eliminated from the first, second, and third historical financial marketing models. The second allocation strategy can also be to select any one or two of the first, second, and third historical financial marketing models and delete duplicate delivery list information from their delivery list information.

[0054] Specifically, the sum of the delivery list information of the first historical financial marketing model and the second historical financial marketing model is obtained and compared with the delivery list information of the third historical financial marketing model. The duplicate user information is deleted from the corresponding delivery list information based on the first allocation strategy output by the completeness preprocessing model.

[0055] It should be noted that in order to improve the accuracy of the estimated conversion rate, the number of historical financial marketing models needs to reach a certain level to obtain as many prior times as possible and improve the accuracy of the prediction. Therefore, the completeness preprocessing provided in the embodiments of this application is not limited to three historical financial marketing models. In other words, the first historical financial marketing model, the second historical financial marketing model, and the third financial marketing model can be any three models in the historical marketing model. On this basis, continuous iteration is carried out until the delivery list information of each model has undergone completeness preprocessing, thereby reducing the impact of data redundancy on the conversion rate calculation.

[0056] It should be understood that completeness preprocessing is performed on two models, namely the first and second historical financial marketing models, to determine the mutually exclusive delivery list information for each of the first and second historical financial marketing models. A third model is then taken and the third model is subjected to completeness preprocessing along with the list information from the first two models to determine the delivery list information for the mutually exclusive third historical financial marketing model. Based on this, completeness preprocessing is performed on other historical financial marketing models, and their delivery list information is compared with the preprocessed delivery list information to identify any duplicates. These duplicates are then deleted from their delivery list information, and this iteration is repeated to obtain mutually exclusive delivery list information for conversion rate calculation.

[0057] S104: Determine the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information, and obtain the conversion rates of multiple historical financial marketing models.

[0058] Specifically, after completeness preprocessing, the delivery list information of each historical financial marketing model may change. Based on this, the corresponding converted list information is updated according to the user ID in the delivery list information after completeness preprocessing, and then the conversion rate of each historical financial marketing model is obtained based on the ratio of the number of users in the converted list information to the number of users in the delivery list information.

[0059] In some embodiments, in order to increase the number of priors and improve the accuracy of predicted conversion rates, before determining the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information, it also includes: obtaining the total number of users in the delivery list information of each historical financial marketing model; when the total number of users is greater than a preset value, randomly sampling and dividing the delivery list information to obtain multiple delivery list information subsets that meet the preset value; based on the delivery list information subsets, generating multiple historical financial marketing model subsets.

[0060] The preset value can be set according to demand or determined based on the characteristic information of the target financial marketing model. The random sampling and segmentation operation can be implemented by computer technology and is not limited here.

[0061] Specifically, to improve the accuracy of predicted conversion rates, the information on delivery lists with a total number of users exceeding a certain value is randomly sampled and divided into multiple subsets. It should be understood that based on the law of large numbers, in statistical activities, the frequency of an event converges to a stable value as the number of experiments increases. Therefore, the delivery list information with a large sample size is divided into subsets to obtain a greater number of experiments. For example, in financial marketing projects, a marketing recommendation model is often pushed to many potential customers separately. It is necessary to divide these many potential customers into multiple groups for statistical analysis to increase the number of priors and, in turn, improve the accuracy of predicted conversion rates.

[0062] In some embodiments, the random sampling and segmentation of the delivery list information further includes: setting a preset threshold and an initial segmentation number based on accuracy requirements; segmenting the delivery list information according to the initial segmentation number to obtain a segmented delivery list information subset, and calculating the conversion rate of multiple delivery list information subsets; comparing the relationship between the variance of multiple conversion rates and the preset threshold, if the variance of multiple conversion rates is less than the preset threshold, gradually increasing the initial segmentation number, each increase corresponds to performing the segmentation operation and the conversion rate calculation operation, and comparing the relationship between the variance of multiple conversion rates and the preset threshold, until the variance of multiple conversion rates is greater than the preset threshold, stopping increasing the initial segmentation number, and using the current number as the segmentation result. Through multiple segmentations, the appropriate subset sample size is determined, the a priori number is increased as much as possible, and the accuracy of the predicted conversion rate is improved.

[0063] In some embodiments, the determining of the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information includes: determining the conversion rates of multiple historical financial marketing model subsets based on the delivery list information subsets and the converted list information; and using the conversion rates of the multiple historical financial marketing model subsets as the conversion rates of the historical financial marketing models before random sampling and segmentation.

[0064] Specifically, the converted list information corresponding to each delivery list information subset is obtained through the user ID in each delivery list information subset, and then the conversion rate of each delivery list information subset is obtained based on the ratio of the number of users in the converted list information to the number of users in the delivery list information. The conversion rates of multiple historical financial marketing model subsets are used to replace the conversion rates of the historical financial marketing models before random sampling and segmentation.

[0065] S105 . Determine an estimated conversion rate of the target financial marketing model based on the Bayesian formula and the conversion rates of the multiple historical financial marketing models, and determine whether the target financial marketing model meets the recommendation condition based on the estimated conversion rate.

[0066] Among them, Bayes' formula can be used to describe the relationship between two conditional probabilities.

[0067] Among them, the recommended conditions are the standards for the target financial marketing model to be put into use, which can be restrictions on the estimated conversion rate value or restrictions on other values ​​derived from the estimated conversion rate.

[0068] Specifically, the Bayesian formula is used to calculate the conditional probability of the target financial marketing model being deployed under each historical financial marketing model, resulting in multiple conditional probabilities. The estimated conversion rate of the target financial marketing model is then predicted based on these multiple conditional probabilities. Based on these estimated conversion rates, it is determined whether the target financial marketing model meets the recommendation criteria.

[0069] In some embodiments, see Figure 3 , Figure 3 This is a flow chart of another method for recommending a financial marketing model provided in an embodiment of the present application, which determines the estimated conversion rate of the target financial marketing model based on the Bayesian formula and the conversion rates of the multiple historical financial marketing models, including steps S1051 to S1053.

[0070] S1051: Obtain the target financial marketing model's delivery list information. Specifically, obtain the target financial marketing model's target potential customer group's list information.

[0071] S1052: Based on the delivery list information of the target financial marketing model and the delivery list information of each historical financial marketing model, determine the conditional probability parameters of each historical financial marketing model according to the Bayesian formula.

[0072] Specifically, the target financial marketing model and the historical financial marketing model are both considered as statistical events, and the conditional probability parameters of each historical financial marketing model are determined according to the Bayesian formula, which are recorded as P(A i |X), the specific formula is as follows:

[0073]

[0074] Among them, A i represents the i-th historical financial marketing model event, P(A i ) represents the probability of the occurrence of the i-th historical financial marketing model event, X represents the target financial marketing model event, P(A i |X) represents the conditional probability of the i-th historical financial marketing model event occurring under the condition that the target financial marketing model event occurs, P(X|A i ) represents the conditional probability of the target financial marketing model event occurring under the condition that the i-th historical financial marketing model event occurs.

[0075] For example, the repeated part of the current historical financial marketing model and the target financial marketing model list can be taken, and the ratio of the repeated part to the total number of users of the current historical financial marketing model can be used as P(X|A i ); The ratio of the total number of users deployed in the current historical financial marketing model to the total number of users deployed in all historical financial marketing models is taken as P(A i ).

[0076] Among them, A j represents the jth historical financial marketing model event, P(A j ) represents the probability of the jth historical financial marketing model event, P(X|A j ) represents the conditional probability of the target financial marketing model event occurring under the condition that the jth historical financial marketing model event occurs, ∑ j P(X|A j )P(A j ) represents P(X|A j )P(A j )The sum of all terms.

[0077] For example, j P(X|A j )P(A j) corresponds to the total probability formula. If there are n historical financial marketing model events, it can be expanded into P(X|A1)+P(X|A2)……+P(X|A n )·P(A n ), which is used to calculate the probability of the target financial marketing model event occurring.

[0078] S1053: Determine an estimated conversion rate of the target financial marketing model based on the conditional probability parameters of each historical financial marketing model and the conversion rates of the multiple historical financial marketing models.

[0079] Specifically, the expected conversion rate of the target financial marketing model is estimated based on the conversion rates of multiple historical financial marketing models and their corresponding conditional probability parameters. The expected conversion rate is used to approximate the estimated conversion rate of the target financial marketing model. The specific formula is as follows:

[0080] E(x)=P(A1|X)·Rate1+P(A2|X)·Rate2...+P(A j |X)·Rate j

[0081] Where E(x) represents the expected conversion rate of the target financial marketing model, P(A j |X) represents the conditional probability parameter of the j-th historical financial marketing model, Rate j represents the conversion rate of the j-th historical financial marketing model.

[0082] For example, the product of the conversion rate of each historical financial marketing model and its corresponding conditional probability parameter is calculated, and the multiple products are added together to obtain the expected conversion rate. It should be understood that when the sample size meets the standard, the expected conversion rate approximates the estimated conversion rate of the target financial marketing model.

[0083] S106: If the target financial marketing model meets the recommendation condition, recommend the target financial marketing model to the terminal device.

[0084] Specifically, if Figure 2 As shown, if the target financial marketing model meets the recommendation conditions, the server recommends the target financial marketing model to multiple user terminal devices corresponding to the target financial marketing model list information, and uses the server to realize the function of intelligent content distribution to accurately deliver the content to the terminal devices of the target customer group.

[0085] In some embodiments, in order to improve the accuracy of the evaluation results, determining whether the target financial marketing model meets the recommendation conditions based on the estimated conversion rate also includes: determining the confidence of the conversion rate based on the estimated conversion rate; judging whether the confidence of the conversion rate meets the preset value; if the confidence of the conversion rate meets the preset value, the target financial marketing model meets the recommendation conditions.

[0086] The preset value may be set according to the accuracy requirement of the evaluation result, and may be a fixed value or a threshold range.

[0087] The confidence level shows the degree to which the true value of the parameter has a certain probability of falling within the range of the measurement result. The confidence level gives the range of credibility of the measured value of the measured parameter.

[0088] Specifically, a probability density map is created based on the estimated conversion rate, and combined with the probability density function, the confidence level of achieving the estimated conversion rate is calculated. When the confidence level of the conversion rate meets the preset value, the target financial marketing model meets the recommendation conditions; when the confidence level of the conversion rate does not meet the preset value, the target financial marketing model does not meet the recommendation conditions.

[0089] See also Figure 4 , Figure 4 This is a schematic block diagram of a financial marketing model recommendation device provided in an embodiment of the present application. The financial marketing model recommendation device can be configured in a server to execute the aforementioned financial marketing model recommendation method.

[0090] like Figure 4 As shown, the recommendation device of the financial marketing model includes: a feature information acquisition module 201, a list information acquisition module 202, an information mutual exclusion processing module 203, a historical conversion rate module 204, an estimated conversion rate module 205, and a target model recommendation module 206.

[0091] A feature information acquisition module 201 is configured to acquire feature information of a target financial marketing model and obtain a plurality of related historical financial marketing models from the preset database based on the feature information;

[0092] A list information acquisition module 202 is used to acquire the delivery list information and the converted list information of the plurality of historical financial marketing models;

[0093] An information mutual exclusion processing module 203 is configured to input the delivery list information of the multiple historical financial marketing models and the converted list information into a complete pre-processing model to obtain multiple mutually exclusive delivery list information. The complete pre-processing model is configured to allocate duplicate list information between the multiple historical financial marketing models to the delivery list information of the corresponding historical financial marketing model;

[0094] A historical conversion rate module 204 is configured to determine the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information, thereby obtaining the conversion rates of multiple historical financial marketing models;

[0095] The estimated conversion rate module 205 is configured to determine an estimated conversion rate of a target financial marketing model based on the Bayesian formula and the conversion rates of the multiple historical financial marketing models, and determine whether the target financial marketing model meets the recommendation conditions based on the estimated conversion rate;

[0096] The target model recommendation module 206 is configured to recommend the target financial marketing model to the terminal device if the target financial marketing model meets the recommendation conditions.

[0097] Exemplarily, the information mutual exclusion processing module 203 further includes a first intersection determination submodule, a first strategy analysis submodule, and a first list deletion submodule.

[0098] A first intersection determination submodule is configured to obtain first repeated delivery list information by taking the intersection of the delivery list information of the first historical financial marketing model and the delivery list information of the second historical financial marketing model;

[0099] A first strategy analysis submodule is configured to analyze the delivery list information of the first historical financial marketing model, the delivery list information of the second historical financial marketing model, and the converted list information to obtain a first allocation strategy;

[0100] The first list deletion submodule is used to delete the first duplicate delivery list information from the corresponding delivery list information based on the first allocation strategy.

[0101] Exemplarily, the information mutual exclusion processing module 203 further includes a list union determination submodule, a second intersection determination submodule, a second strategy analysis submodule, and a second list deletion submodule.

[0102] a list union determination submodule, configured to obtain the union of the delivery list information of the first historical financial marketing model and the delivery list information of the second historical financial marketing model to obtain delivered list information;

[0103] A second intersection determination submodule is configured to obtain the intersection of the delivery list information of the third historical financial marketing model and the delivered list information to obtain second repeated delivery list information;

[0104] A second strategy analysis submodule is configured to analyze the delivery list information of the first historical financial marketing model, the delivery list information of the second historical financial marketing model, the delivery list information of the third historical financial marketing model, and the converted list information to obtain a second allocation strategy;

[0105] The second list deletion submodule is used to delete the second repeated delivery list information from the corresponding delivery list information based on the second allocation strategy.

[0106] Exemplarily, the recommendation device for the financial marketing model further includes a quantity acquisition module, a sampling and segmentation module, and a model subset module.

[0107] The quantity acquisition module is used to obtain the total number of users in the delivery list information of each historical financial marketing model;

[0108] A sampling and segmentation module, configured to perform random sampling and segmentation on the delivery list information when the total number of users is greater than a preset value, to obtain a plurality of delivery list information subsets that meet the preset value;

[0109] The model subset module is used to generate multiple historical financial marketing model subsets based on the delivery list information subset.

[0110] Exemplarily, the historical conversion rate module 204 further includes a model conversion rate determination submodule and a model conversion rate determination submodule.

[0111] A subset conversion rate determination submodule, configured to determine the conversion rates of a plurality of historical financial marketing model subsets based on the delivery list information subset and the converted list information;

[0112] The model conversion rate determination submodule is configured to use the conversion rates of the plurality of historical financial marketing model subsets as the conversion rates of the historical financial marketing models before random sampling and segmentation.

[0113] Exemplarily, the estimated conversion rate module 205 further includes a target model list submodule, a conditional probability parameter submodule, and an estimated conversion rate submodule.

[0114] The target model list submodule is used to obtain the delivery list information of the target financial marketing model;

[0115] a conditional probability parameter submodule, configured to determine, based on the delivery list information of the target financial marketing model and the delivery list information of each historical financial marketing model, the conditional probability parameter of each historical financial marketing model under the target financial marketing model according to the Bayesian formula;

[0116] The estimated conversion rate submodule is configured to determine the estimated conversion rate of the target financial marketing model based on the conditional probability parameters of each historical financial marketing model under the target financial marketing model and the conversion rates of the multiple historical financial marketing models.

[0117] Exemplarily, the target model recommendation module 206 further includes a confidence determination submodule, a condition judgment submodule, and a condition satisfaction submodule.

[0118] a confidence determination submodule, configured to determine the confidence of the conversion rate based on the estimated conversion rate;

[0119] A condition judgment submodule, used to judge whether the confidence level of the conversion rate meets a preset value;

[0120] The condition satisfaction submodule is used to determine that if the confidence level of the conversion rate meets a preset value, the target financial marketing model meets the recommendation condition.

[0121] For example, the above method and apparatus may be implemented in the form of a computer program. The computer program may be implemented in the form of a computer program. Figure 5 Run on the server shown.

[0122] See also Figure 5 , Figure 5 This is a schematic block diagram of the structure of a server provided in an embodiment of the present application.

[0123] like Figure 5 As shown, the server includes a processor, a memory, a network interface, and a preset database connected through a system bus, wherein the memory may include a storage medium and an internal memory.

[0124] The storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute a recommendation method for any financial marketing model.

[0125] The processor is used to provide computing and control capabilities to support the operation of the entire server.

[0126] The internal memory provides an environment for the operation of the computer program in the storage medium. When the computer program is executed by the processor, the processor can execute the recommendation method of any financial marketing model.

[0127] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the server to which the solution of the present application is applied. The specific server may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0128] It should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0129] In one embodiment, the processor is configured to execute a computer program stored in the memory to implement the following steps:

[0130] Acquire characteristic information of a target financial marketing model, and acquire multiple related historical financial marketing models from the preset database based on the characteristic information; acquire delivery list information and converted list information of multiple historical financial marketing models; input the delivery list information and converted list information of multiple historical financial marketing models into a complete preprocessing model to obtain multiple mutually exclusive delivery list information, wherein the complete preprocessing model is used to assign duplicate list information between the multiple historical financial marketing models to the delivery list information of the corresponding historical financial marketing models; determine the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information to obtain the conversion rates of multiple historical financial marketing models; determine the estimated conversion rate of the target financial marketing model based on the Bayesian formula and the conversion rates of the multiple historical financial marketing models, and determine whether the target financial marketing model meets the recommendation conditions based on the estimated conversion rate; if the target financial marketing model meets the recommendation conditions, recommend the target financial marketing model to the terminal device.

[0131] In one embodiment, the multiple historical financial marketing models include at least a first historical financial marketing model and a second historical financial marketing model; the processor implements the inputting of the delivery list information and the converted list information of the multiple historical financial marketing models into a complete preprocessing model, including: taking the intersection of the delivery list information of the first historical financial marketing model and the delivery list information of the second historical financial marketing model to obtain first repeated delivery list information; analyzing the delivery list information of the first historical financial marketing model, the delivery list information of the second historical financial marketing model, and the converted list information to obtain a first allocation strategy; and deleting the first repeated delivery list information from the corresponding delivery list information based on the first allocation strategy.

[0132] In one embodiment, the multiple historical financial marketing models also include a third historical financial marketing model; the processor, when implementing the inputting of the delivery list information and the converted list information of the multiple historical financial marketing models into the complete preprocessing model, also includes: taking the union of the delivery list information of the first historical financial marketing model and the delivery list information of the second historical financial marketing model to obtain the delivered list information; taking the intersection of the delivery list information of the third historical financial marketing model and the delivered list information to obtain the second repeated delivery list information; analyzing the delivery list information of the first historical financial marketing model, the delivery list information of the second historical financial marketing model, the delivery list information of the third historical financial marketing model, and the converted list information to obtain a second allocation strategy; and deleting the second repeated delivery list information from the corresponding delivery list information based on the second allocation strategy.

[0133] In one embodiment, when implementing a method for recommending a financial marketing model, the processor is used to: obtain the total number of users in the delivery list information of each historical financial marketing model; when the total number of users is greater than a preset value, randomly sample and split the delivery list information to obtain multiple delivery list information subsets that meet the preset value; and generate multiple historical financial marketing model subsets based on the delivery list information subsets.

[0134] In one embodiment, when the processor determines the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information, it is used to achieve: determining the conversion rates of multiple historical financial marketing model subsets based on the delivery list information subset and the converted list information; and using the conversion rates of the multiple historical financial marketing model subsets as the conversion rates of the historical financial marketing models before random sampling and segmentation.

[0135] In one embodiment, when determining the estimated conversion rate of the target financial marketing model based on the Bayesian formula and the conversion rates of the multiple historical financial marketing models, the processor is used to: obtain the delivery list information of the target financial marketing model; determine the conditional probability parameters of each historical financial marketing model under the target financial marketing model based on the Bayesian formula based on the delivery list information of the target financial marketing model and the delivery list information of each historical financial marketing model; and determine the estimated conversion rate of the target financial marketing model based on the conditional probability parameters of each historical financial marketing model under the target financial marketing model and the conversion rates of the multiple historical financial marketing models.

[0136] In one embodiment, when implementing the determination of whether the target financial marketing model meets the recommendation conditions based on the estimated conversion rate, the processor is used to implement: determining the confidence of the conversion rate based on the estimated conversion rate; judging whether the confidence of the conversion rate meets a preset value; if the confidence of the conversion rate meets the preset value, the target financial marketing model meets the recommendation conditions.

[0137] It should be noted that those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the recommendation of the financial marketing model described above can refer to the corresponding process in the aforementioned financial marketing model recommendation method embodiment, and will not be repeated here.

[0138] An embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. The computer program includes program instructions. The method implemented when the program instructions are executed can refer to the various embodiments of the recommendation method of the financial marketing model of the present application.

[0139] The computer-readable storage medium may be an internal storage unit of the server described in the aforementioned embodiment, such as a hard disk or memory of the server. The computer-readable storage medium may also be an external storage device of the server, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the server.

[0140] It should be understood that the terms used in this specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in this specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0141] It should also be understood that the term "and / or" used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations. It should be noted that, in this article, the terms "include", "comprise" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system that includes a series of elements includes not only those elements, but also other elements that are not explicitly listed, or also includes elements that are inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "including a..." does not exclude the presence of other identical elements in the process, method, article or system that includes the element.

[0142] The serial numbers of the embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments. The above description is only a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for recommending a financial marketing model, characterized in that: Applied to a server, the server corresponding to a preset database, the method includes: Acquiring characteristic information of a target financial marketing model, and acquiring a plurality of related historical financial marketing models from the preset database based on the characteristic information; Obtaining delivery list information and converted list information of multiple historical financial marketing models; Inputting the delivery list information of the plurality of historical financial marketing models and the converted list information into a complete pre-processing model to obtain a plurality of mutually exclusive delivery list information, wherein the complete pre-processing model is used to assign duplicate list information between the plurality of historical financial marketing models to the delivery list information of the corresponding historical financial marketing model; Determining the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information, thereby obtaining the conversion rates of multiple historical financial marketing models; Determining an estimated conversion rate of a target financial marketing model based on the Bayesian formula and the conversion rates of the multiple historical financial marketing models, and determining whether the target financial marketing model meets the recommendation conditions based on the estimated conversion rate; If the target financial marketing model meets the recommendation condition, the target financial marketing model is recommended to the terminal device.

2. The method according to claim 1, characterized in that The multiple historical financial marketing models include at least a first historical financial marketing model and a second historical financial marketing model; The step of inputting the delivery list information and the converted list information of the plurality of historical financial marketing models into a complete pre-processing model includes: Take the intersection of the delivery list information of the first historical financial marketing model and the delivery list information of the second historical financial marketing model to obtain the first repeated delivery list information; Analyze the delivery list information of the first historical financial marketing model, the delivery list information of the second historical financial marketing model, and the converted list information to obtain a first allocation strategy; The first duplicate delivery list information is deleted from the corresponding delivery list information based on the first allocation strategy.

3. The method according to claim 2, characterized in that The plurality of historical financial marketing models further includes a third historical financial marketing model; The step of inputting the delivery list information and the converted list information of the plurality of historical financial marketing models into a complete pre-processing model further includes: Taking the union of the delivery list information of the first historical financial marketing model and the delivery list information of the second historical financial marketing model to obtain delivered list information; Taking the intersection of the delivery list information of the third historical financial marketing model and the delivered list information to obtain second repeated delivery list information; Analyze the delivery list information of the first historical financial marketing model, the delivery list information of the second historical financial marketing model, the delivery list information of the third historical financial marketing model, and the converted list information to obtain a second allocation strategy; The second duplicate delivery list information is deleted from the corresponding delivery list information based on the second allocation strategy.

4. The method according to claim 1, wherein Before determining the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information, the method further includes: Get the total number of users in the delivery list information of each historical financial marketing model; When the total number of users is greater than a preset value, the delivery list information is randomly sampled and divided to obtain multiple delivery list information subsets that meet the preset value; Based on the delivery list information subset, multiple historical financial marketing model subsets are generated.

5. The method according to claim 4, characterized in that The determining the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information includes: Determining conversion rates of multiple historical financial marketing model subsets based on the delivery list information subset and the converted list information; The conversion rates of the plurality of historical financial marketing model subsets are used as the conversion rates of the historical financial marketing models before random sampling and segmentation.

6. The method according to claim 1, wherein Determining the estimated conversion rate of the target financial marketing model based on the Bayesian formula and the conversion rates of the multiple historical financial marketing models includes: Obtain the delivery list information of the target financial marketing model; Based on the delivery list information of the target financial marketing model and the delivery list information of each historical financial marketing model, determining the conditional probability parameters of each historical financial marketing model under the target financial marketing model according to the Bayesian formula; An estimated conversion rate of the target financial marketing model is determined based on the conditional probability parameters of each historical financial marketing model under the target financial marketing model and the conversion rates of the multiple historical financial marketing models.

7. The method according to claim 6, characterized in that The determining whether the target financial marketing model meets the recommendation condition according to the estimated conversion rate further includes: determining a confidence level of the conversion rate based on the estimated conversion rate; Determining whether the confidence level of the conversion rate meets a preset value; If the confidence level of the conversion rate meets a preset value, the target financial marketing model meets the recommendation condition.

8. A financial marketing model recommendation device, characterized in that: The recommendation device of the financial marketing model includes: A feature information acquisition module is used to acquire feature information of a target financial marketing model and obtain a plurality of related historical financial marketing models from a preset database based on the feature information; A list information acquisition module, used to obtain the delivery list information and converted list information of multiple historical financial marketing models; An information mutual exclusion processing module, configured to input the delivery list information of the plurality of historical financial marketing models and the converted list information into a complete pre-processing model to obtain a plurality of mutually exclusive delivery list information, wherein the complete pre-processing model is configured to allocate duplicate list information between the plurality of historical financial marketing models to the delivery list information of the corresponding historical financial marketing model; A historical conversion rate module, configured to determine the conversion rate of each historical financial marketing model based on the multiple mutually exclusive delivery list information and the converted list information, thereby obtaining the conversion rates of multiple historical financial marketing models; a conversion rate estimation module, configured to determine an estimated conversion rate of a target financial marketing model based on a Bayesian formula and the conversion rates of the plurality of historical financial marketing models, and determine whether the target financial marketing model meets a recommendation condition based on the estimated conversion rate; The target model recommendation module is configured to recommend the target financial marketing model to the terminal device if the target financial marketing model meets the recommendation conditions.

9. A server, characterized in that: The server includes a memory, a processor, and a preset database; The memory is used to store computer programs; The processor is configured to execute the computer program and implement the method for recommending a financial marketing model according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, enables the processor to implement the method for recommending a financial marketing model according to any one of claims 1 to 7.

Images