CN113919868A

CN113919868A - Equity distribution method and device based on marketing promotion model and electronic equipment

Info

Publication number: CN113919868A
Application number: CN202111147576.7A
Authority: CN
Inventors: 祁云峰; 宋碧莲; 李盛刚
Original assignee: Shanghai Hualong Information Technology Co ltd
Current assignee: Shanghai Hualong Information Technology Co ltd
Priority date: 2021-09-28
Filing date: 2021-09-28
Publication date: 2022-01-11

Abstract

The invention relates to the technical field of computers, in particular to a method, a device and electronic equipment for distributing equity based on a marketing promotion model, which comprises the following steps: reading in resource data through a data template; performing data exploration and data cleaning on the resource data to obtain summary characteristics; establishing a response promotion model and a sales increment model according to the summary characteristics; executing the response promotion model and the sales increment model to obtain a response promotion score and a sales increment score; performing performance verification and value analysis on the response promotion model and the sales increment model; and matching the optimal equity dispatch to the customer according to the value analysis result and the preset equity threshold. The invention adopts a response promotion model and a sales increment model in the intelligent marketing process and combines a promotion interest tool to establish an optimized sales operation system, selects the optimal interest types and distributes the optimal interest types to the most suitable client groups through market activities so as to realize the optimal return on investment.

Description

Equity distribution method and device based on marketing promotion model and electronic equipment

Technical Field

The invention relates to the technical field of computers, in particular to a method and a device for equity distribution based on a marketing promotion model and electronic equipment.

Background

The traditional lifting modeling method appears in Radcliffe and Surry work for the first time, and provides a differential response analysis method for the first time, namely a precursor of the lifting modeling method, the model method is provided for the first time to be applied to a customer response model in a digital marketing process, the lifting modeling method is also called a lifting response model (also called incremental modeling, real lifting modeling or net modeling) is a modeling prediction technology widely used in processes of promotion, renewal, cross recommendation and the like in marketing, and compared with a common response model, the method can directly model the incremental influence of a certain behavior such as a direct marketing behavior. It is therefore possible to answer more accurately whether marketing activities such as tickets, advertisements actually improve sales indicators such as customer liveness, customer purchase amount, customer conversion, customer responsiveness, etc. In addition, the promotion modeling also has functions for promoting sales, cross-sales and retention rate modeling in customer relationship management. In terms of application areas, this method has also been applied to political election and personalized medicine.

The einstein system is a system for enhancing marketing efforts by means of customer sales data and artificial intelligence sales models, which provides customers of the einstein system with predictive analysis, natural language processing functions and machine learning using data collected on every user operation. In terms of artificial intelligence techniques, the cloud platform provides a means to predict whether a customer will purchase a product more or less through prediction and potential customer and opportunity scores to help improve the odds. This system also contains various customer response models, but does not have the functionality of lift-based modeling and can be used to implement sales business operations.

Infer is an artificial intelligence platform integrated with multiple solutions that can be used to predict potential customers and customer scores and manage hyper-subdivided profiles. The user can use the referrer system to predict potential customer scores and use the predicted results of the platform for analysis to assist in marketing, for example, to help the user determine who the desired buyer is, to learn their buying behavior, and to avoid wasting time in making customer value conversions. Infer may be used to analyze customer relationship system data, marketing automation systems, to find potential customers that are most likely to purchase and best suited for your company. It can calculate the exact score to tell your sales person exactly which will be contributing to the incoming call, which will bring more revenue, which will be degraded, etc. The system also comprises various customer grouping models, marketing channel selection models and the like, but does not have the functions of building models based on promotion and being used for realizing the operation of sales business.

Although the corresponding promotion algorithm has been applied to the sales field for more than ten years, no established software system based on promotion modeling and available for realizing sales business operation has been found; the data used in conventional response promotion models is data that is collected after the market promotion campaign is implemented, i.e., the promotion models cannot be developed prior to the market promotion campaign. This wastes data collected prior to the market promotion campaign and can result in waiting periods, resulting in wasted time.

Technologies aiming at predicting a type 0,1 or a multi-point type in response lift models do not relate to new response lift models for predicting continuous increment targets and do not relate to prediction problems such as lift of sales volume; meanwhile, the existing model is a modeling person or a professional with a strong computer or statistical background when facing users, but business personnel or data operators cannot directly use the software to perform business operation.

Disclosure of Invention

The invention provides a right and interest distribution method and device based on a marketing promotion model and electronic equipment, which are used for selecting the optimal right and interest types to be distributed to the most suitable client groups through market activities so as to realize the optimal return on investment.

The embodiment of the specification provides a rights and interests dispatching method based on a marketing promotion model, which comprises the following steps:

reading in resource data through a data template, wherein the resource data comprises: transaction data, customer information characteristic data, third party information characteristic data;

performing data exploration and data cleaning on the resource data to obtain summary characteristics;

establishing a response promotion model and a sales increment model according to the summary characteristics;

executing the response promotion model and the sales increment model to obtain a response promotion score and a sales increment score;

performing performance verification and value analysis on the response promotion model and the sales increment model;

and matching the optimal equity dispatch to the customer according to the value analysis result and the preset equity threshold.

Preferably, the reading resource data is divided into two stages, including:

the first stage reads in the customer transaction data before the market promotion activity is implemented;

and the second stage reads in the customer transaction data after the market promotion activity is implemented and is connected with the customer transaction data read in the first stage before the market promotion activity is implemented in a main key mode.

Preferably, the data exploration and data cleansing of the resource data includes:

displaying features related to reading in resource data;

inputting and filling missing data;

processing outliers in the resource data;

converting the resource data classification variables into digital variables;

regularizing the resource data.

Preferably, the summary features include: transaction frequency, transaction amount, transaction time interval, purchase new progress.

Preferably, the establishing of the response promotion model and the sales increment model according to the summary features includes:

screening the summary features by adopting a feature clustering method of variable clustering;

when the summary features are abnormal to establish a response promotion model and a sales increment model, the summary features are subjected to data interception interactively in a data interception mode;

and establishing a response promotion model and a sales increment model according to the intercepted summary features.

Preferably, the response lift model includes: a first response model after the market equity promotion activity has not been conducted for the first time, a second response model after the market equity promotion activity has been conducted, and a third response model after the market equity promotion activity has been conducted; the sales increment model includes: a first value predictive model for which no market interest promotional program has been implemented for the first time, a second value predictive model for which a market interest promotional program has been implemented, and a third value predictive model for which a market interest promotional program has been implemented.

Preferably, the executing the response promotion model and the sales increment model to obtain the response promotion score and the sales increment score includes:

predicting each customer through the first response model, the second response model and the third response model to respectively obtain response scores;

calculating a response promotion score for each customer based on the response scores;

predicting each customer through the first price prediction model, the second price prediction model and the third price prediction model to respectively obtain a value score;

a value delta score is calculated for each customer based on the value scores.

Preferably, the performing performance verification and value analysis on the response promotion model and the sales increment model includes:

performing performance verification on the response promotion model and the sales increment model according to the response scores and the value increment scores and by combining historical resource data;

and analyzing the value of the response promotion model and the sales increment model according to the consumption amount, consumption times and cost.

Preferably, the response promotion model and the sales increment model perform logistic regression by expanding a redefined ROC curve, and are used for promoting performance verification accuracy and value analysis accuracy of the response promotion model and the sales increment model.

Preferably, the sending the best equity matched with a preset equity threshold to the client according to the value analysis result comprises:

when selecting from a plurality of equities, selecting the equity corresponding to the highest response boost score and the value increment score in the value analysis results to serve to the customer;

and when the response promotion score and the value increment score corresponding to the equity are lower than a preset threshold value, no matched equity is distributed to the client.

Preferably, the preset threshold comprises a first threshold and a second threshold, wherein the first threshold is determined by the ratio of the average response promotion score of the whole customer to the average response promotion score of the individual customer; the second threshold is determined by a ratio of the average value delta score for the entire customer and the average value delta score for the individual customer.

An embodiment of the present specification further provides a device for equity distribution based on a marketing promotion model, including:

a data read-in module; reading in resource data through a data template, wherein the resource data comprises: transaction data, customer information characteristic data, third party information characteristic data;

a feature acquisition module; performing data exploration and data cleaning on the resource data to obtain summary characteristics;

the model creating module is used for creating a response promotion model and a sales increment model according to the summary characteristics;

the model execution module is used for executing the response promotion model and the sales increment model and acquiring a response promotion score and a sales increment score;

the verification analysis module is used for performing performance verification and value analysis on the response promotion model and the sales increment model;

and the equity dispatching module is used for dispatching the optimal equity matched with the preset equity threshold to the client according to the value analysis result.

An electronic device, wherein the electronic device comprises:

a processor and a memory storing computer executable instructions that, when executed, cause the processor to perform the method of any of the above.

A computer readable storage medium, wherein the computer readable storage medium stores one or more programs which, when executed by a processor, implement the method of any of the above.

The beneficial effects are that:

the invention adopts a response promotion model and a sales increment model in the intelligent marketing process and combines a promotion interest tool to establish an optimized sales operation system, selects the optimal interest types and distributes the optimal interest types to the most suitable client groups through market activities so as to realize the optimal return on investment.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:

fig. 1 is a schematic diagram illustrating an equity distribution method based on a marketing promotion model according to an embodiment of the present disclosure;

fig. 2 is a schematic structural diagram of an equity distribution device based on a marketing promotion model according to an embodiment of the present disclosure;

fig. 3 is a schematic structural diagram of an electronic device provided in an embodiment of the present disclosure;

fig. 4 is a schematic diagram of a computer-readable medium provided in an embodiment of the present specification.

Detailed Description

Exemplary embodiments of the present invention will now be described more fully with reference to the accompanying drawings. The exemplary embodiments, however, may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these exemplary embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the invention to those skilled in the art. The same reference numerals denote the same or similar elements, components, or parts in the drawings, and thus their repetitive description will be omitted.

Features, structures, characteristics or other details described in a particular embodiment do not preclude the fact that the features, structures, characteristics or other details may be combined in a suitable manner in one or more other embodiments in accordance with the technical idea of the invention.

In describing particular embodiments, the present invention has been described with reference to features, structures, characteristics or other details that are within the purview of one skilled in the art to provide a thorough understanding of the embodiments. One skilled in the relevant art will recognize, however, that the invention may be practiced without one or more of the specific features, structures, characteristics, or other details.

The diagrams depicted in the figures are exemplary only, and do not necessarily include all of the contents and operations/steps, nor do they necessarily have to be performed in the order depicted. For example, some operations/steps may be decomposed, and some operations/steps may be combined or partially combined, so that the actual execution sequence may be changed according to the actual situation.

The block diagrams shown in the figures are functional entities only and do not necessarily correspond to physically separate entities. I.e. these functional entities may be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor means and/or microcontroller means.

The term "and/or" and/or "includes all combinations of any one or more of the associated listed items.

Referring to fig. 1, a schematic diagram of an equity distribution method based on a marketing promotion model provided in an embodiment of the present disclosure includes:

s101: reading in resource data through a data template, wherein the resource data comprises: transaction data, customer information characteristic data, third party information characteristic data;

in the preferred embodiment of the present invention, the transactional data alone may be used to build a model if the information provided by the transactional data is deemed sufficient, in general. However, if the user has additional data deemed valuable, the user may optionally add customer base information and/or third party information. The customer basic information table contains some static data of the customer, and the third-party data is data which can be connected to transaction data through other information of the customer, such as telephone and the like, such as payment water and electricity fee summary of the customer, credit rating of the customer and the like; since the data structures of the above three types of data cannot be determined in advance, as a general invention system, the three types of data are read in data in the form of a general data template, that is, the data structures such as table names, field names, data types and classification information in the table are initialized in advance, and are mapped in the form of a dictionary by using standard names, and then are read in the system.

The data template initial information would be stored in a permanent external file and the system of the present invention could update these settings based on the data and user requirements. The source data is read into the system according to the settings, and the data import process is to read the source data into the memory according to the displayed initialization table and generate a data frame with data structure rules.

S102: performing data exploration and data cleaning on the resource data to obtain summary characteristics;

in the preferred embodiment of the present invention, a tree list control is used to place the original data column features on tree nodes, which saves space and facilitates operations such as expansion, contraction, etc. Computer interactive programming is needed for realizing the functions, so that a user can select different characteristics in the tree form table to respectively observe, and customized condition query can be carried out to return a subdata set or a subdata column and establish summary characteristics. The task of data exploration is to display characteristics related to the read-in data, such as the number of rows and columns, basic statistics, the number and proportion of missing characteristics, and the type of data, interpretation, etc.

S103: establishing a response promotion model and a sales increment model according to the summary characteristics;

in the preferred embodiment of the present invention, the user can create a single or batch model object through the system of the present invention, and can also generate a customized mixed model object by using feature engineering, wherein the model object refers to the dependent variable of the model used for predicting the model. In particular, the user needs to specify or define a corresponding feature column in the training data, such as whether to purchase within the next three months, the total amount of purchases within the next year, etc.; a response promotion model and a sales increase model are then created based on the aggregated features according to the created model objectives.

The invention has a new response lifting model for predicting the continuous increment target besides the traditional response lifting model for predicting the 0,1 type or multi-classification type target. We introduce in the following scenarios:

modeling data for market promotion campaigns has not been implemented for the first time: the data here is built for the sales incremental model, so the model target is a continuous numerical feature, such as sales or money or usage for the next three months;

modeling data after implementation of a market promotion campaign: the data here is built for both sales incremental models and response promotion models, so the model targets can be continuous numeric features or binary features of type 0,1, such as whether there is a purchase, activation, etc. in the next three months.

S104: executing the response promotion model and the sales increment model to obtain a response promotion score and a sales increment score;

in the preferred embodiment of the invention, a response promotion model and the sales increment model are executed, and a response promotion score and a sales increment score are output after the response promotion model and the sales increment model are executed.

S105: performing performance verification and value analysis on the response promotion model and the sales increment model;

in the preferred embodiment of the invention, the performance verification is carried out on the response promotion model and the sales increment model according to the response promotion score and the sales increment score and by combining historical resource data; and analyzing the value of the response promotion model and the sales increment model according to the consumption amount, consumption times and cost, wherein the final analysis result can be used for optimizing equity distribution.

S106: and matching the optimal equity dispatch to the customer according to the value analysis result and the preset equity threshold.

In the preferred embodiment of the present invention, equity optimization is performed according to the generated value analysis results, and the existing equity categories are distributed to the most suitable client population to optimize the overall benefit index. If a selection from multiple equity is desired, then the response promotion model and the sales delta model should be multi-equity response promotion models or multi-equity sales delta models, then eventually there should be multiple promotion prediction scores or multiple sales delta prediction scores for each customer. For example, if three equity categories (A, B, C) are available for selection, and each customer receives three forecast scores after forecasting the value of the response promotion model and the sales incremental model for each customer, then for a customer, the system assigns the equity with the highest selected score to that customer.

In terms of cost, not all customers will receive market promotion equity, since even the highest scoring equity will not cause significant response and sales increments, and so here in addition to the above analysis of the highest equity lift score, the absolute score of the highest equity lift score needs to be considered, i.e., an equity threshold needs to be found below which equity lift scores will not be selected for distribution, but above which they can be distributed.

Further, the reading resource data is divided into two stages, including:

In the preferred embodiment of the present invention, the resource data is read in stages, the first stage will read in the customer transaction data before the market promotion is implemented, the customer will create some kind of response model using the model engine of the system; the second stage is to read in the customer transaction data after the market promotion event has been conducted, noting that this is only incremental data, so the system needs to be linked to the data read in the first stage by way of primary keys such as customer ID, product number, etc.

Further, the data exploration and data cleansing of the resource data includes:

displaying features related to reading in resource data;

inputting and filling missing data;

processing outliers in the resource data;

converting the resource data classification variables into digital variables;

regularizing the resource data.

In a preferred embodiment of the present invention, after displaying the features associated with the read-in resource data, further performing data cleansing on the resource data, the data cleansing including: inputting and filling missing data, processing abnormal values in the resource data, converting the resource data classification variables into digital variables, normalizing the resource data, and performing one or more corresponding processing modes on the resource data according to different conditions of the data.

Inputting and filling missing data: while some machine learning models automatically process rows containing missing values during the training phase, but most algorithms do not accept datasets with missing values, the simplest solution to process missing values here is to delete a row or entire column, without an optimal deletion threshold, but the present invention user can use a threshold chosen to be, say, 70% as an example value and try to delete rows and columns with missing values above this threshold; the system also provides for padding to address data loss, where the user can evaluate important content based on the missing values, and also consider default values for possible missing values in the column. For example, if there is a column containing only 1 and NA, then the NA row may correspond to 0, except in the case where a default value is used for the missing value, the data fill is by using the median of the feature column.

Outliers in the process data: two methods of dealing with outliers are listed, which will be detected using standard deviation and percentile, one is with outliers having a standard deviation, which can be considered outliers if a value is more distant from the mean than the standard deviation; another is that the system provides percentile outlier detection, which may assume a certain percentage of the values from the top or bottom as outliers.

Converting the categorical variables into numerical variables: the main purpose of the herein merging groups, which can be applied to the merging of classification data and digital data into groups, is to make the model more robust and prevent overfitting, but this reduces performance. Each time a group is merged, some useful information is lost, but the data is made more normalized, and the tradeoff between performance and overfitting is critical to the merging-into-group process. It is noted here that merging into groups may be redundant for boosting algorithms, but for classification columns, merging makes the model more stable.

Data regularization: after the data features undergo a canonical scaling process, the range of the continuous features becomes the same. For many algorithms, this process is not mandatory, but may be applied to improve model accuracy.

Further, the summary features include: transaction frequency, transaction amount, transaction time interval, purchase new progress.

Further, the establishing a response promotion model and a sales increase model according to the summary features comprises:

In the preferred embodiment of the present invention, the system generates summary features from the resource data, and the following types of features are generally necessary in the customer response model, including:

number of purchases in past time periods: i.e., transaction frequency, such as the average number of months over the past three months to purchase a product;

number of purchases in past time period: such as a monthly transaction amount for purchasing a product in the past three months;

the client age: the time from the customer's first purchase to the present, such as days;

length of time between first and last purchase: and purchasing the recency.

Further, if the basic information of the client is provided, information such as age group, sex, income, and the like is important. The method of the system is that some important characteristics are preset, and then addition or deletion modification and the like are carried out according to the requirements of users; since too long data may have a detrimental effect on modeling, the system of the present invention provides the user with a data interception function, and the user may interactively select time recency to limit the modeled data. For example, a user may intercept the transaction data of the last two years as the modeled data sample.

Further, the response promotion model includes: a first response model after the market equity promotion activity has not been conducted for the first time, a second response model after the market equity promotion activity has been conducted, and a third response model after the market equity promotion activity has been conducted; the sales increment model includes: a first value predictive model for which no market interest promotional program has been implemented for the first time, a second value predictive model for which a market interest promotional program has been implemented, and a third value predictive model for which a market interest promotional program has been implemented.

In the preferred embodiment of the present invention, the several types of models involved include the following:

the first response model MR for a market interest promotion campaign has not been conducted for the first time, which is to build a 0,1 goal model for all customer samples prior to conducting the market promotion campaign. The goal of 1 means that the customer has purchased during a certain period of time from the current point of time, and the goal of 0 means none. The results of this predictive model will be combined with the results of the underlying model to form the final result of the response boosting model.

The first value prediction model MV, in which the market interest promotion activity has not been conducted for the first time, represents the value of the purchase of the customer during a certain period of time from the current time point, such as the amount of the repurchase, the amount of money, etc. Or may be specified by the user according to a specific service. The results of this predictive model will be combined with the results of the underlying model to form the final result of the sales incremental model.

Modeling after implementation of market equity promotion campaigns: the data here is built for both the sales incremental model and the response promotion model. The following categories are covered:

establishing two customer response models after the market interest promotion activity is implemented, wherein one is to establish a second response model MRT aiming at the customer samples implementing the market interest promotion activity; one is to build a third customer response model MRC for those customer samples for which no marketing campaign is being conducted, the model predicting the output as a probability value falling within the interval [ 0,1 ].

Establishing two customer value models after the market equity promotion activity is implemented, wherein one customer value model is a second value prediction model MVT aiming at customer samples implementing the market equity promotion activity; one is to build a second value prediction model MVC for those customer samples that do not implement market interest promotion campaigns.

Further, the executing the response promotion model and the sales increment model to obtain the response promotion score and the sales increment score includes:

a value delta score is calculated for each customer based on the value scores.

In a preferred embodiment of the present invention, response promotion model and sales delta model predictions are performed: the main function is to carry out summary calculation according to the several staged models generated in the last step to obtain a response promotion model and a sales increment model prediction score. The specific implementation process is as follows:

responding to the lifting model: respectively predicting all client groups by using the models MR, MRT and MRC generated in the previous step to respectively obtain client response scores: MR (X), MRT (X) and MRC (X), and then calculating a response boost score for each customer according to equation (1):

SCORE＝MRT(X)–0.5*MR(X)–0.5*MRC(X)(1)

the traditional calculation method is SCORE (MRT) (x) -MRC (x), i.e. the promotion SCORE is the difference between the Control group Treatment group (Treatment group), note that MRT (x), which is the prediction of customer response after the market interest promotion activity is performed, whereas the present invention considers not only this difference but also the prediction of customer response before the market interest promotion activity, i.e. mr (x), which is the advantage that MRT (x) and MRC (x) are modeled separately in the traditional promotion model, the MRT model only represents those customers who participate in the market interest activity, and the MRC represents those customers who do not participate in the market interest activity. Although the initial equity distribution experiment design is theoretically established on a random sample, namely, any equity is randomly distributed to any client, the actual operation is not random when the equity is really received, for example, the client which can receive the equity has a larger response probability, so that the client behavior cannot be really changed after the equity is obtained. Therefore, the deviation caused by different customer samples can be generated by calculating the promotion score by adopting the difference value of the two in the traditional response promotion model, and the MR (X) is added in the method to adjust the deviation of the samples, because the MR reflects the response probability of all the people before the market interest activity.

Sales increment model: and respectively predicting all client groups by using the models MV, MVT and MVC generated in the previous step to respectively obtain client response scores: MV (X), MVT (X) and MVC (X), and then calculate a value increment score for each customer according to equation (2):

SCORE＝MVT(X)–0.5*MV(X)–0.5*MVC(X)(2)

the principle of the above improved method is the same as that of the response promotion model, and is not further described here.

Further, the performing performance verification and value analysis on the response promotion model and the sales increment model comprises:

In the preferred embodiment of the invention, the performance verification is carried out on the response promotion model and the sales increment model according to the response score and the value increment score and by combining historical resource data; and analyzing the value of the response promotion model and the sales increment model according to the consumption amount, consumption times and cost, and performing logistic regression on the response promotion model and the sales increment model by expanding a redefined ROC curve when performing performance verification and value analysis on the model to improve the performance verification accuracy and the value analysis accuracy of the response promotion model and the sales increment model.

Further, the response promotion model and the sales increment model adopt expansion redefined ROC curves to carry out logistic regression, and the logistic regression is used for promoting performance verification accuracy and value analysis accuracy of the response promotion model and the sales increment model.

In a preferred embodiment of the invention, the validation of the response promotion model and the validation of the general response model are different because there are two types of people in the validation data, i.e., those who have performed equity promotions and those who have not, such as customers who have received neither equity nor equity in the validation data when considering the promotion response model, and therefore the predicted promotion amount of the validation data cannot be directly compared with the actual level truth of the validation data. The present invention therefore employs a redefined ROC curve to perform logistic regression on the response promotion model and the sales increase model for promotion model evaluation. Because the real information of the promotion value can not be obtained only by implementing the equity promotion data, the system can check the overall promotion value by classifying the verification data, and the value analysis only needs to expand the longitudinal index in the verification into the sales amount and then consider the equity cost.

Further, the selecting the equity best matched with the response promotion model and the sales increment model according to the value analysis result to be dispatched to the customer comprises the following steps:

In the preferred embodiment of the present invention, if a selection from multiple equity is desired, then the response promotion model and the sales incremental model should be either a multi-equity response promotion model or a multi-equity sales incremental model, and finally there should be multiple promotion prediction scores or multiple sales incremental prediction scores for each customer. For example, if three equity categories (A, B, C) are available for selection, and each customer receives three forecast scores after forecasting the value of the response promotion model and the sales incremental model for each customer, then for a customer, the system assigns the equity with the highest selected score to that customer.

In terms of cost, not all customers will receive market promotion equity, since even the highest scoring equity will not cause significant response and sales increments, and so here in addition to the above analysis of the highest equity lift score, the absolute score of the highest equity lift score needs to be considered, i.e., a threshold value needs to be found below which equity lift scores will not be selected to be assigned.

Further, the preset threshold comprises a first threshold and a second threshold, wherein the first threshold is determined by the ratio of the average response promotion score of the whole customer to the average response promotion score of the individual customer; the second threshold is determined by a ratio of the average value delta score for the entire customer and the average value delta score for the individual customer.

In the preferred embodiment of the present invention, this threshold is determined by the ratio of the average equity accrual score for the entire customer to the equity accrual score for the individual customer. For example, if a certain customer equity promotion score is 150% of the overall average equity promotion score, then the system may observe that the number of customers reaching 150% of the ratio is only 20% of the total customers, then 150% may be selected as the threshold, and in addition, the user may determine his/her preferred threshold according to similar methods and results.

In the preferred embodiment of the present invention, the system of the present invention solves the general problem of out-of-time sample prediction for a product by using standardized data templates, in terms of data type; from the industrial application, the method can be applied to general demand prediction problems in a supply chain prediction system, and can also be applied to out-of-time sample prediction problems of any index of any product in the transaction process; the system is very simple to use and operate, and professional knowledge of algorithms and models is not needed to be known by professionals; the characteristic clustering method used by the system screens the predicted characteristics of a plurality of products, the execution efficiency of characteristic dimension reduction is very high, and therefore a user can quickly find out the most relevant predicted characteristics; because the invention solves the problem of universal time-out sample prediction, the system and the use interface can help companies to carry out pre-sale training, demonstration and scheme guidance of different industries and application scenes on clients and pre-sale personnel.

Fig. 2 is a schematic structural diagram of an equity distribution device based on a marketing promotion model according to an embodiment of the present disclosure, including:

a data reading module 201; reading in resource data through a data template, wherein the resource data comprises: transaction data, customer information characteristic data, third party information characteristic data;

a feature acquisition module 202; performing data exploration and data cleaning on the resource data to obtain summary characteristics;

the model creating module 203 is used for creating a response promotion model and a sales increment model according to the summary characteristics;

the model execution module 204 is used for executing the response promotion model and the sales increment model and acquiring a response promotion score and a sales increment score;

a verification analysis module 205, which performs performance verification and value analysis on the response promotion model and the sales increment model;

and the equity dispatching module 206 is used for dispatching the best equity matched with the preset equity threshold to the client according to the value analysis result.

Based on the same inventive concept, the embodiment of the specification further provides the electronic equipment.

In the following, embodiments of the electronic device of the present invention are described, which may be regarded as specific physical implementations for the above-described embodiments of the method and apparatus of the present invention. Details described in the embodiments of the electronic device of the invention should be considered supplementary to the embodiments of the method or apparatus described above; for details which are not disclosed in embodiments of the electronic device of the invention, reference may be made to the above-described embodiments of the method or the apparatus.

Fig. 3 is a schematic structural diagram of an electronic device provided in an embodiment of the present specification. An electronic device 300 according to this embodiment of the invention is described below with reference to fig. 3. The electronic device 300 shown in fig. 3 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present invention.

As shown in fig. 3, electronic device 300 is embodied in the form of a general purpose computing device. The components of electronic device 300 may include, but are not limited to: at least one processing unit 310, at least one memory unit 320, a bus 330 connecting different device components (including the memory unit 320 and the processing unit 310), a display unit 340, and the like.

Wherein the storage unit stores program code executable by the processing unit 310 to cause the processing unit 310 to perform the steps according to various exemplary embodiments of the present invention described in the above-mentioned processing method section of the present specification. For example, the processing unit 310 may perform the steps as shown in fig. 1.

The storage unit 320 may include readable media in the form of volatile storage units, such as a random access memory unit (RAM)3201 and/or a cache storage unit 3202, and may further include a read only memory unit (ROM) 3203.

The storage unit 320 may also include a program/utility 3204 having a set (at least one) of program modules 3205, such program modules 3205 including, but not limited to: an operating device, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.

Bus 330 may be one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

The electronic device 300 may also communicate with one or more external devices 400 (e.g., keyboard, pointing device, bluetooth device, etc.), with one or more devices that enable a user to interact with the electronic device 300, and/or with any devices (e.g., router, modem, etc.) that enable the electronic device 300 to communicate with one or more other computing devices. Such communication may occur via an input/output (I/O) interface 350. Also, the electronic device 300 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the internet) via the network adapter 360. Network adapter 360 may communicate with other modules of electronic device 300 via bus 330. It should be appreciated that although not shown in FIG. 3, other hardware and/or software modules may be used in conjunction with electronic device 300, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID devices, tape drives, and data backup storage devices, to name a few.

Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments of the present invention described herein may be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiment of the present invention can be embodied in the form of a software product, which can be stored in a computer-readable storage medium (which can be a CD-ROM, a usb disk, a removable hard disk, etc.) or on a network, and includes several instructions to make a computing device (which can be a personal computer, a server, or a network device, etc.) execute the above-mentioned method according to the present invention. The computer program, when executed by a data processing apparatus, enables the computer readable medium to implement the above-described method of the invention, namely: such as the method shown in fig. 1.

Fig. 4 is a schematic diagram of a computer-readable medium provided in an embodiment of the present disclosure.

A computer program implementing the method shown in fig. 1 may be stored on one or more computer readable media. The computer readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor device, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The computer readable storage medium may include a propagated data signal with readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A readable storage medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution apparatus, device, or apparatus. Program code embodied on a readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).

In summary, the invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that some or all of the functionality of some or all of the components in embodiments in accordance with the invention may be implemented in practice using a general purpose data processing device such as a microprocessor or a Digital Signal Processor (DSP). The present invention may also be embodied as apparatus or device programs (e.g., computer programs and computer program products) for performing a portion or all of the methods described herein. Such programs implementing the present invention may be stored on computer-readable media or may be in the form of one or more signals. Such a signal may be downloaded from an internet website or provided on a carrier signal or in any other form.

While the foregoing embodiments have described the objects, aspects and advantages of the present invention in further detail, it should be understood that the present invention is not inherently related to any particular computer, virtual machine or electronic device, and various general-purpose machines may be used to implement the present invention. The invention is not to be considered as limited to the specific embodiments thereof, but is to be understood as being modified in all respects, all changes and equivalents that come within the spirit and scope of the invention.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.

The above description is only an example of the present application and is not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A equity distribution method based on a marketing promotion model is characterized by comprising the following steps:

2. The marketing promotion model-based equity distribution method of claim 1, wherein the reading resource data is divided into two phases, including:

3. The marketing promotion model-based equity distribution method of claim 1, wherein the data exploration and data cleansing of the resource data comprises:

displaying features related to reading in resource data;

inputting and filling missing data;

processing outliers in the resource data;

converting the resource data classification variables into digital variables;

regularizing the resource data.

4. The marketing promotion model-based equity distribution method of claim 1, wherein said building response promotion models and sales increment models from said aggregated features comprises:

screening the summary features by adopting a feature clustering method of variable clustering, wherein the summary features comprise: transaction frequency, transaction amount, transaction time interval, new purchase progress;

establishing a response promotion model and a sales increment model according to the intercepted summary features;

the response promotion model includes: a first response model after the market equity promotion activity has not been conducted for the first time, a second response model after the market equity promotion activity has been conducted, and a third response model after the market equity promotion activity has been conducted;

the sales increment model includes: a first value predictive model for which no market interest promotional program has been implemented for the first time, a second value predictive model for which a market interest promotional program has been implemented, and a third value predictive model for which a market interest promotional program has been implemented.

5. The marketing promotion model-based equity distribution method of claim 4, wherein said executing the response promotion model and the sales increment model to obtain a response promotion score and a sales increment score comprises:

a value delta score is calculated for each customer based on the value scores.

6. The method of claim 5, wherein the performing performance verification and value analysis on the response promotion model and the sales increment model comprises:

performing performance verification on the response promotion model and the sales increment model according to the response promotion score and the value increment score and by combining historical resource data;

analyzing the value of the response promotion model and the sales increment model according to the consumption amount, consumption times and cost;

and performing logistic regression on the response promotion model and the sales increment model by expanding a redefined ROC curve, and improving the performance verification accuracy and the value analysis accuracy of the response promotion model and the sales increment model.

7. The method of claim 1, wherein the assigning the best equity to the client according to the value analysis result and combined with a predetermined equity threshold comprises:

selecting the highest response promotion score and the corresponding interest of the value increment score in the value analysis result when selecting a plurality of interests;

when the response promotion score and the value increment score corresponding to the equity are higher than a preset equity threshold value, the equity is dispatched to the client;

when the response promotion score and the value increment score corresponding to the equity are lower than a preset equity threshold value, no matched equity is distributed to the client;

the preset equity threshold comprises a first threshold and a second threshold, the first threshold is determined by the ratio of the average response promotion score of the whole customer and the average response promotion score of the individual customer;

the second threshold is determined by a ratio of the average value delta score for the entire customer and the average value delta score for the individual customer.

8. An equity distribution device based on a marketing promotion model, comprising:

the data reading module reads resource data through a data template, wherein the resource data comprises: transaction data, customer information characteristic data, third party information characteristic data;

the characteristic acquisition module is used for carrying out data exploration and data cleaning on the resource data to obtain summary characteristics;

9. An electronic device, wherein the electronic device comprises:

a processor and a memory storing computer-executable instructions that, when executed, cause the processor to perform the method of any of claims 1-7.

10. A computer readable storage medium, wherein the computer readable storage medium stores one or more programs which, when executed by a processor, implement the method of any of claims 1-7.