CN114266601A - Marketing strategy determination method and device, terminal equipment and storage medium - Google Patents
Marketing strategy determination method and device, terminal equipment and storage medium Download PDFInfo
- Publication number
- CN114266601A CN114266601A CN202111606916.8A CN202111606916A CN114266601A CN 114266601 A CN114266601 A CN 114266601A CN 202111606916 A CN202111606916 A CN 202111606916A CN 114266601 A CN114266601 A CN 114266601A
- Authority
- CN
- China
- Prior art keywords
- user
- marketing
- marketing strategy
- target
- feature information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000013145 classification model Methods 0.000 claims abstract description 40
- 238000003066 decision tree Methods 0.000 claims description 57
- 238000007637 random forest analysis Methods 0.000 claims description 38
- 238000012549 training Methods 0.000 claims description 25
- 230000006872 improvement Effects 0.000 claims description 11
- 238000000605 extraction Methods 0.000 claims description 3
- 230000006399 behavior Effects 0.000 abstract description 11
- 230000008901 benefit Effects 0.000 abstract description 6
- 230000008569 process Effects 0.000 description 7
- 239000013598 vector Substances 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 238000004590 computer program Methods 0.000 description 6
- 238000006243 chemical reaction Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 230000009471 action Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 230000001133 acceleration Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000005484 gravity Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000013138 pruning Methods 0.000 description 1
- 238000010079 rubber tapping Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Abstract
The invention discloses a marketing strategy determination method, a marketing strategy determination device, terminal equipment and a storage medium, wherein the method comprises the following steps: acquiring observability data of a user, and extracting basic feature information of the user from the observability data; inputting the basic feature information into a pre-trained target classification model, classifying the user based on the basic feature information, and obtaining a decision result of the user; and determining a target marketing strategy corresponding to the user according to the decision result. According to the invention, users are classified through the observability data of the users, the structural difference of the customer base is eliminated, the marketing strategy is evaluated, the influence of the customer base on the tendency of the objective condition of the customer base can be eliminated, and the influence of different marketing strategies on the user behaviors can be accurately estimated, so that the selected marketing strategy is more accurate, the matching degree with the users is higher, and the marketing benefit is further improved.
Description
Technical Field
The invention relates to the technical field of intelligent marketing, in particular to a marketing strategy determination method, a marketing strategy determination device, terminal equipment and a storage medium.
Background
With the development of internet technology, internet marketing activities are also related to aspects of people's lives gradually. In internet marketing campaigns, the most common problems are: different marketing channels and marketing modes, for example information flow, APP start-up picture or backstage propelling movement etc. which can promote the profit most. Further, the intelligent marketing aims to find the optimal marketing strategy corresponding to each customer so as to facilitate personalized marketing service. Most of the existing intelligent marketing modes predict the consumption behaviors of users through historical consumption records of the users or historical behavior records of browsing and clicking webpages, so that an individualized marketing scheme is generated. However, this method ignores the structural difference of the guest group, and affects the prediction result. For example, customers entering targeted pages with advertisements in luxury APPs are generally more expensive than customers entering targeted pages with advertisements in the information stream, which does not necessarily indicate that the advertisements pushed in luxury APPs themselves are better marketed, but because the average consumption of customers themselves using luxury APPs is stronger. When structural differences of the customer groups are ignored, the accuracy of the predicted marketing strategy is affected, and when the matching degree of the marketing strategy and the users is not high, the purpose of income improvement is difficult to achieve.
Disclosure of Invention
The invention mainly aims to provide a marketing strategy determination method, a marketing strategy determination device, terminal equipment and a storage medium, and aims to solve the technical problem that the marketing strategy prediction corresponding to a user is inaccurate due to the fact that the structural difference of a customer group is ignored in the existing intelligent marketing mode.
In addition, in order to achieve the above object, the present invention further provides a marketing strategy determining method, including the steps of:
acquiring observability data of a user, and extracting basic feature information of the user from the observability data;
inputting the basic feature information into a pre-trained target classification model, classifying the user based on the basic feature information, and obtaining a decision result of the user;
and determining a target marketing strategy corresponding to the user according to the decision result.
Optionally, the decision result includes a plurality of candidate marketing strategies, and the step of determining the target marketing strategy corresponding to the user according to the decision result includes:
calculating a first profit value corresponding to each candidate marketing strategy in the decision result;
acquiring a second profit value corresponding to the user at present, and calculating a target difference between each first profit value and each second profit value;
and determining a target marketing strategy corresponding to the user from the plurality of marketing strategies according to the target difference.
Optionally, the step of determining the target marketing strategy corresponding to the user from the plurality of marketing strategies according to the target difference includes:
calculating a marketing cost value for converting the current marketing strategy of the user into each candidate marketing strategy;
calculating a profit-improvement value for the user based on the marketing cost value and the target difference value;
determining the target marketing strategy with the maximum profit increase value from the plurality of candidate marketing strategies.
Optionally, after the step of calculating the revenue improvement value of the user based on the marketing cost value and the target difference value, the method further comprises:
and if the income improvement value is negative, setting the current marketing strategy of the user as a target marketing strategy.
Optionally, the target classification model includes a random forest model, and the step of inputting the basic feature information into a pre-trained target classification model, classifying the user based on the basic feature information, and obtaining a decision result for the user includes:
inputting the basic feature information into a pre-trained random forest model, and generating a plurality of decision trees by taking the basic feature information as a classification condition, wherein the maximum depth of each decision tree is smaller than a preset depth threshold;
classifying the users based on the decision trees, determining target leaf nodes corresponding to the users in the decision trees, and obtaining decision results of the users, wherein the leaf nodes of the decision trees are candidate marketing strategies.
Optionally, before the step of inputting the basic feature information into a pre-trained classification model, classifying the user by using the basic feature information as a classification condition, and obtaining a decision result for the user, the method further includes:
acquiring observability data of each user in a user set to form an observability data set, and constructing a sample data set based on the observability data set;
and pre-training a preset basic classification model by using the sample data set to obtain a target classification model.
Optionally, the target classification model includes a random forest model, the sample data set includes basic feature information of each user in the user set, and the step of pre-training a preset basic classification model by using the sample data set to obtain the target classification model includes:
obtaining model training parameters, wherein the model training parameters comprise a depth threshold and a sample threshold;
and splitting by taking the basic characteristic information of each user in the user set as a classification condition to generate a random forest model, wherein the random forest model comprises a plurality of decision trees, the maximum depth of each decision tree is smaller than the depth threshold, leaf nodes of each decision tree are candidate marketing strategies, and the number of samples corresponding to each candidate marketing strategy is not smaller than the sample threshold.
In addition, to achieve the above object, the present invention provides a marketing strategy determining apparatus, including:
the characteristic extraction module is used for acquiring observability data of a user and extracting basic characteristic information of the user from the observability data;
the classification decision module is used for inputting the basic characteristic information into a pre-trained target classification model, classifying the user by taking the basic characteristic information as a classification condition, and obtaining a decision result of the user;
and the strategy selection module is used for determining a target marketing strategy corresponding to the user according to the decision result.
In addition, to achieve the above object, the present invention also provides a terminal device, including: a memory, a processor, and a marketing strategy determination program stored on the memory and executable on the processor, the marketing strategy determination program when executed by the processor implementing the steps of the marketing strategy determination method as described above.
Further, to achieve the above object, the present invention also provides a computer readable storage medium having a marketing strategy determination program stored thereon, which when executed by a processor, implements the steps of the marketing strategy determination method as described above.
Furthermore, to achieve the above object, the present invention also provides a computer program product comprising a computer program which, when being executed by a processor, realizes the steps of the marketing strategy determination method as described above.
The embodiment of the invention provides a marketing strategy determination method, a marketing strategy determination device, terminal equipment and a storage medium. In the existing marketing strategy determining mode, prediction is mostly carried out based on historical consumption records or historical behavior records of users, and the difference of customer groups of the users is ignored, so that the generated marketing strategy is inaccurate, the matching degree with the users is not high, and marketing benefits cannot be improved. In the embodiment of the invention, the basic characteristic information of a user is extracted from observability data by acquiring the observability data of the user; inputting the basic feature information into a pre-trained target classification model, classifying the user based on the basic feature information, and obtaining a decision result of the user; and determining a target marketing strategy corresponding to the user according to the decision result. The users are classified based on the observability data of the users, structural differences do not exist among user groups of the same category, the influence of the tendency of the user groups based on self objective conditions can be eliminated through the evaluation of marketing strategies, the influence of different marketing strategies on the user behaviors can be accurately estimated, the selected marketing strategies are more accurate, the matching degree with the users is higher, and the marketing benefits are improved.
Drawings
Fig. 1 is a schematic hardware structure diagram of an implementation manner of a terminal device according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a marketing strategy determination method according to a first embodiment of the present invention;
fig. 3 is a functional block diagram of a marketing strategy determining apparatus according to an embodiment of the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
In the following description, suffixes such as "module", "component", or "unit" used to denote elements are used only for facilitating the explanation of the present invention, and have no specific meaning in itself. Thus, "module", "component" or "unit" may be used mixedly.
The marketing strategy determination terminal (called terminal, equipment or terminal equipment) in the embodiment of the invention can be a PC (personal computer), or can be mobile terminal equipment with display and data processing functions, such as a smart phone, a tablet computer, a portable computer and the like.
As shown in fig. 1, the terminal may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, a communication bus 1002. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory (e.g., a magnetic disk memory). The memory 1005 may alternatively be a storage device separate from the processor 1001.
Optionally, the terminal may further include a camera, a Radio Frequency (RF) circuit, a sensor, an audio circuit, a WiFi module, and the like. Such as light sensors, motion sensors, and other sensors. Specifically, the light sensor may include an ambient light sensor that may adjust the brightness of the display screen according to the brightness of ambient light, and a proximity sensor that may turn off the display screen and/or the backlight when the mobile terminal is moved to the ear. As one of the motion sensors, the gravity acceleration sensor can detect the magnitude of acceleration in each direction (generally, three axes), detect the magnitude and direction of gravity when the mobile terminal is stationary, and can be used for applications (such as horizontal and vertical screen switching, related games, magnetometer attitude calibration), vibration recognition related functions (such as pedometer and tapping) and the like for recognizing the attitude of the mobile terminal; of course, the mobile terminal may also be configured with other sensors such as a gyroscope, a barometer, a hygrometer, a thermometer, and an infrared sensor, which are not described herein again.
Those skilled in the art will appreciate that the terminal structure shown in fig. 1 is not intended to be limiting and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a kind of computer-readable storage medium, may include therein an operating system, a network communication module, a user interface module, and a marketing strategy determination program.
In the terminal shown in fig. 1, the network interface 1004 is mainly used for connecting to a backend server and performing data communication with the backend server; the user interface 1003 is mainly used for connecting a client (user side) and performing data communication with the client; and the processor 1001 may be configured to invoke a marketing strategy determination program stored in the memory 1005 that, when executed by the processor, implements the operations in the marketing strategy determination methods provided by the embodiments described below.
Based on the hardware structure of the equipment, the embodiment of the marketing strategy determination method is provided.
Referring to fig. 2, in a first embodiment of the marketing strategy determination method of the present invention, the marketing strategy determination method includes:
step S10, acquiring observability data of a user, and extracting basic feature information of the user from the observability data;
in this embodiment, when predicting the marketing strategy corresponding to the user, firstly, observability data of the user is obtained, where the observability data refers to data that can be observed under natural uncontrolled conditions, and includes basic feature information of the user, such as name, gender, age, occupation, and the like. The marketing strategy comprises a marketing mode and a marketing channel, and also comprises a marketing commodity type, and it is required to be noted that the conventional marketing strategy predicts the consumption tendency of a user according to the historical consumption record or the historical behavior record of the user, such as advertisement browsing and clicking behavior, and the like, and specifically recommends a specific type of marketing commodity to the user according to the consumption tendency of the user. For example, when a user clicks and enters a detailed page of a certain marketing commodity in an advertisement pushed by social software, the possibility that the user has a purchasing tendency of the commodity is predicted, and the same type of commodity can be continuously recommended to the user through different marketing modes or marketing channels, so that the purpose of accurate marketing is achieved. However, this approach ignores the structural differences between different users, which are generated based on the observability data of different users themselves, for example, a user clicks a push advertisement in social software to enter a detail page of a certain item, not necessarily for a purchase purpose, the consumption level of the push item and the consumption level of the user themselves may not match, the click behavior of the user may be a wrong click, the user clicks the advertisement to enter the detail page may also be a simple desire to view the item details, there is no purchase tendency, and the same type of item is continuously pushed through different marketing approaches or marketing channels, which also only increases the marketing cost.
The marketing strategy determination method in the embodiment is different in that when the marketing strategy corresponding to the user is predicted, observability data of the user is obtained, and basic characteristic information of the user is extracted from the observability data, so that the users with similar guest group structures can be found conveniently, structural differences among different users are eliminated, the marketing strategies borne by different users are independently and uniformly distributed, and the prediction accuracy of the marketing strategies corresponding to different users is improved.
Step S20, inputting the basic feature information into a pre-trained target classification model, classifying the user based on the basic feature information, and obtaining a decision result of the user;
further, after extracting the basic feature information of the user from the obtained observability data, inputting the extracted basic feature information into a pre-trained target classification model, and classifying the user based on the extracted basic feature information to obtain a decision result of the user. The target classification model at least includes a decision tree model and a random forest model, and the following description will take the random forest model as the target classification model as an example. On one hand, because the representation capability of the decision tree is weaker than that of the random forest, when the feature dimension is higher, a better classification effect can be obtained by adopting the random forest model. On the other hand, the extracted basic feature information of the user comprises a plurality of features, for different users, the completeness of the obtained observability data is different, and the number of the extracted features is different, so that a better classification effect can be obtained by adopting a random forest model based on the completeness.
Further, when classifying users based on the basic feature information of the users, the extracted basic feature information may be used to weight data such as historical consumption records and historical behavior records of the users, and the weighted data may be used as classification conditions for classification, or the basic feature information of the users may be directly used as classification conditions for classification, or the basic feature information of the users may be used as classification conditions for classification, and then the users are further classified based on the historical consumption records and the historical behavior records of the users, so as to obtain the classification decision results of the users.
Before step S20, the method further includes:
a1, acquiring observability data of each user in a user set to form an observability data set, and constructing a sample data set based on the observability data set;
and A2, pre-training a preset basic classification model by using the sample data set to obtain a target classification model.
In this embodiment, before determining the target marketing strategy corresponding to the user, a pre-trained basic classification model needs to be pre-trained to obtain a pre-trained target classification model. When pre-training is carried out, firstly, observability data of each user in a user set is obtained to form an observability data set, a sample data set is constructed on the basis of the observability data set, wherein the user set comprises a plurality of users, when the sample data set is constructed, each user and corresponding basic feature information of each user are extracted, a preset basic classification model is pre-trained by utilizing the constructed sample data set, and a target classification model is obtained and comprises a random forest model.
When the target classification model is a random forest model, the pre-training process specifically includes that firstly, a random forest model is established by taking which marketing strategy the user accepts as a target, and L (x) is used for representing the leaf node number of sample data corresponding to the user, wherein the leaf node number falls into the sample data. L (x) constitutes a disjoint partition of the feature space, and users with the same feature are "similar". When the decision tree classification effect of the random forest is good, the marketing modes T borne by the users in the same leaf node can be considered to be independent and distributed, and then the users approximately form a random test sample. And calculating samples under the same leaf node so as to determine the optimal marketing strategy corresponding to each user. A sample corresponding to each user falls into a unique leaf node in each decision tree of the random forest, and the results of the trees are integrated to obtain a leaf node vector L (x) ═ L (L)1(x),...,Ln(x) The leaf node vector is the decision result for the user, where L isi(x) Indicating the leaf node number in which the sample falls in the ith tree.
Further, when model pre-training is performed, the constructed sample data set includes basic feature information of each user in the user set, and the pre-training step further includes:
step B1, obtaining model training parameters, wherein the model training parameters comprise a depth threshold and a sample threshold;
and step B2, splitting the basic characteristic information of each user in the user set as a classification condition to generate a random forest model, wherein the random forest model comprises a plurality of decision trees, the maximum depth of each decision tree is smaller than the depth threshold, leaf nodes of each decision tree are candidate marketing strategies, and the number of samples corresponding to each candidate marketing strategy is not smaller than the sample threshold.
When model pre-training is carried out, firstly, model training parameters are obtained, the model training parameters comprise a depth threshold and a sample threshold, pre-pruning is carried out on a random forest, and overfitting of the model is prevented. The depth threshold is used for limiting the maximum depth of each decision tree in the random forest, and the sample threshold is used for limiting the sample number of each leaf node (namely the number of users falling into the same leaf node) after each decision tree is split according to the classification condition. Splitting is carried out by taking the basic characteristic information of each user in the sample data set as a classification condition, a random forest model is generated, the generated random forest model comprises a plurality of decision trees, the maximum depth of each decision tree is smaller than a set depth threshold, leaf nodes of each decision tree are candidate marketing strategies, the number of samples after the leaf nodes of each decision tree are split is not smaller than the set sample threshold, namely the number of users falling into each candidate marketing strategy is not smaller than the set sample threshold, if the number of samples of the leaf sub-nodes after the splitting is smaller than the set sample threshold, the splitting is not carried out, and in the splitting process, when the maximum depth of the decision trees exceeds the set depth threshold, the splitting is stopped, and overfitting of the model is prevented.
In the pre-training process, the depth threshold and the sample threshold are adjustable, and can be set in a user-defined manner according to the actual training requirement and the scale of the sample data set, which is not specifically limited herein.
And step S30, determining a target marketing strategy corresponding to the user according to the decision result.
And determining a target marketing strategy corresponding to the user according to a decision result of the user, and selecting the marketing strategy with the highest marketing benefit from the leaf node vector corresponding to the user as the target marketing strategy corresponding to the user. Further, further calculation may be performed based on the leaf node vector corresponding to the user, for example, a marketing manner and a marketing channel corresponding to each marketing strategy, a commodity type in each marketing manner and marketing channel, and the like are subdivided, so as to generate a corresponding target marketing strategy. Further, the profit promotion value of the candidate marketing strategy corresponding to each leaf node can be calculated based on the leaf node vector corresponding to the user, and the marketing strategy with the maximum profit promotion value is used as the target marketing strategy. Specifically, according to the marketing cost and the profit value of the marketing strategy currently corresponding to the user, if the marketing strategy is changed into each candidate marketing strategy, the variation of the marketing cost and the variation of the profit value are calculated, the difference between the increment of the profit value and the increment of the marketing cost is the profit promotion value, and the candidate marketing strategy with the maximum profit promotion value is used as the target marketing strategy.
In this embodiment, the mode of determining the target marketing strategy corresponding to the user may also be to use the profit variation as a leaf node value of the random forest decision tree, and specifically, between two different marketing strategies, the profit variation may be calculated by the following formula 1:
Diff(x)＝E(Y|T＝1,X＝x)-E(Y|T＝0,X＝x) (1)
wherein E (· |) represents a conditional expectation; y represents a profit value, and the meaning of the profit value is determined according to the actual applied service scene, for example, for an internet information platform, the profit is the click rate of a user on a webpage or an APP; for the E-commerce platform, the income is the amount or times of purchasing commodities by the user; x is available basic characteristic information of the user such as gender, age, occupation, etc.; t denotes a marketing strategy, T ═ 0 denotes a first marketing strategy, and T ═ 1 denotes a second marketing strategy.
And taking the calculated income difference value as a leaf node value of the random forest decision tree so as to intuitively obtain the income variation of each candidate marketing strategy, comparing the value of each leaf node with the corresponding cost variation, and calculating an income improvement value so as to determine the target marketing strategy corresponding to the user. For example, suppose there is a data set (X)i,Ti,Yi) And i is 0,1,2,., n, X represents the basic characteristic information of the user, Y represents the income generated by the user, T represents the marketing strategy suffered by the user, and T belongs to {0,1} without loss of generality. For a random forest model generated by taking T as a dependent variable and X as an independent variable, the calculation of the Diff estimation value of the leaf node of the mth decision tree is shown in the following formula 2:
in the generated decision tree of the original random forest model, the leaf values of the leaf nodes of the decision trees are probability values of the users falling into the leaf nodes, and the original leaf values of the decision trees are replaced by the profit variation, so that the target marketing strategy corresponding to the users can be determined more intuitively and quickly.
In the embodiment, basic feature information of a user is extracted from observability data by acquiring the observability data of the user; inputting the basic feature information into a pre-trained target classification model, classifying the user based on the basic feature information, and obtaining a decision result of the user; and determining a target marketing strategy corresponding to the user according to the decision result. The users are classified based on the observability data of the users, structural differences do not exist among user groups of the same category, the influence of the tendency of the user groups based on self objective conditions can be eliminated through the evaluation of marketing strategies, the influence of different marketing strategies on the user behaviors can be accurately estimated, the selected marketing strategies are more accurate, the matching degree with the users is higher, the promotion of marketing benefits is facilitated, and personalized marketing services are performed on different users.
Further, in the embodiment, the modeling target is converted, and the random forest decision tree is reconstructed, so that the target marketing strategy corresponding to the user can be determined more intuitively and quickly, the income change amount can be quantized, and the decision result can be evaluated.
Further, on the basis of the above embodiment of the present invention, a second embodiment of the marketing strategy determination method of the present invention is provided.
The present embodiment is a step of step S20 and step S30 refinement in the first embodiment, and based on the above embodiment, the refinement of step S20 includes:
step S201, inputting the basic feature information into a pre-trained random forest model, and generating a plurality of decision trees by taking the basic feature information as a classification condition, wherein the maximum depth of each decision tree is less than a preset depth threshold;
step S202, classifying the users based on the decision trees, and determining target leaf nodes corresponding to the users in the decision trees to obtain decision results for the users, wherein the leaf nodes of the decision trees are candidate marketing strategies.
Based on the foregoing embodiment, in this embodiment, when classifying users based on the basic feature information of the users to obtain a decision result for the users, the basic feature information of the users is input into a pre-trained random forest model, and each feature in the basic feature information of the users is used as a classification condition to generate a plurality of decision trees. The maximum depth of each decision tree is smaller than a preset depth threshold, and leaf nodes in each decision tree are candidate marketing strategies. Classifying users based on each decision tree, determining target leaf nodes which the users fall into in each decision tree, wherein the target leaf nodes comprise a plurality of leaf nodes, integrating the leaf nodes which the users fall into in each decision tree to obtain corresponding leaf node vectors, and the leaf node vectors are decision results of the users.
Further, the decision result for the user includes a plurality of candidate marketing strategies, and after the decision result for the user is obtained, the refinement of the target marketing strategy corresponding to the user is determined according to the decision result, which includes:
step S301, calculating a first income value corresponding to each candidate marketing strategy in the decision result;
step S302, obtaining a second profit value corresponding to the user at present, and calculating a target difference value between each first profit value and each second profit value;
step S303, determining a target marketing strategy corresponding to the user from the marketing strategies according to the target difference value.
When a target marketing strategy corresponding to a user is determined based on a decision result of the user, first profit values corresponding to candidate marketing strategies in the decision result are calculated, then second profit values corresponding to the user are obtained, a target difference value between each first profit value and each second profit value is calculated, and the target marketing strategy corresponding to the user is determined based on the target difference value. It should be noted that the second profit value currently corresponding to the user may be a profit value in a natural state where no marketing strategy is accepted, or may be a profit value when the current marketing strategy is accepted.
Further, in step S303, determining refinement of the target marketing strategy corresponding to the user based on the calculated target difference between each first profit value and each second profit value, further includes:
step C1, calculating a marketing cost value for converting the current marketing strategy of the user into each candidate marketing strategy;
step C2, calculating a profit improvement value for the user based on the marketing cost value and the target difference value;
step C3, determining the target marketing strategy with the maximum profit margin from the plurality of candidate marketing strategies.
Therefore, the change of the marketing strategy brings the corresponding change of the marketing cost, and in the practical application, if the increase of the income value brought by the transformation of the marketing strategy is not enough to cover the increase of the marketing cost, the transformation of the marketing strategy is not needed. That is, when determining the target marketing strategy corresponding to the user according to the calculated target difference, the variation of the marketing cost needs to be considered at the same time. Specifically, a marketing cost value when the current marketing strategy of the user is converted into each candidate marketing strategy in the decision result is calculated, the marketing cost value is a cost added value brought by the conversion of the marketing strategy, and a profit increase value corresponding to the user is obtained by subtracting the marketing cost value from a target difference value of marketing profits. That is, in the present embodiment, the profit margin improvement value by the conversion of the marketing strategy to the user is calculated by subtracting the marketing cost value added by the conversion of the marketing strategy. And selecting the marketing strategy with the maximum profit promotion value from the candidate marketing strategies as a target marketing strategy corresponding to the user.
Further, after the step C2, the method further includes:
and step C4, if the income improvement value is negative, setting the current marketing strategy of the user as a target marketing strategy.
It can be understood that, in the calculation decision result, the revenue improvement value corresponding to each candidate marketing strategy is negative, that is, the revenue increase caused by the conversion of each marketing strategy is not enough to cover the increase of the marketing cost, so that the current marketing strategy of the user may not be converted to the target marketing strategy.
It should be noted that, in the embodiments of the present invention, structural differences of the guest groups are solved by converting modeling targets and combining with characteristics of the tree models, and based on the same ideas and methods, through other classification models and in combination with specific service scenarios, as long as users with the same basic feature information can be classified, the same effect as that of the embodiments of the present invention can be achieved, and details are not described here.
In this embodiment, when the target marketing strategy corresponding to the user is determined, the leaf values of the leaf nodes of the decision tree are replaced, and meanwhile, the income improvement value corresponding to each candidate marketing strategy can be accurately measured in consideration of the increase of marketing cost brought by the marketing strategy conversion, so that the target marketing strategy corresponding to the user, which can maximize the marketing income, is accurately determined, and the accuracy of marketing strategy selection is improved.
In addition, referring to fig. 3, an embodiment of the present invention further provides an apparatus, where the apparatus includes:
the characteristic extraction module 10 is configured to acquire observability data of a user and extract basic characteristic information of the user from the observability data;
a classification decision module 20, configured to input the basic feature information into a pre-trained target classification model, and classify the user by using the basic feature information as a classification condition to obtain a decision result for the user;
and the strategy selecting module 30 is configured to determine a target marketing strategy corresponding to the user according to the decision result.
Optionally, the decision result includes a plurality of candidate marketing strategies, and the strategy selection module 30 is further configured to:
calculating a first profit value corresponding to each candidate marketing strategy in the decision result;
acquiring a second profit value corresponding to the user at present, and calculating a target difference between each first profit value and each second profit value;
and determining a target marketing strategy corresponding to the user from the plurality of marketing strategies according to the target difference.
Optionally, the policy selecting module 30 is further configured to:
calculating a marketing cost value for converting the current marketing strategy of the user into each candidate marketing strategy;
calculating a profit-improvement value for the user based on the marketing cost value and the target difference value;
determining the target marketing strategy with the maximum profit increase value from the plurality of candidate marketing strategies.
Optionally, the policy selecting module 30 is further configured to:
and if the income improvement value is negative, setting the current marketing strategy of the user as a target marketing strategy.
Optionally, the target classification model includes a random forest model, and the classification decision module 20 is further configured to:
inputting the basic feature information into a pre-trained random forest model, and generating a plurality of decision trees by taking the basic feature information as a classification condition, wherein the maximum depth of each decision tree is smaller than a preset depth threshold;
classifying the users based on the decision trees, determining target leaf nodes corresponding to the users in the decision trees, and obtaining decision results of the users, wherein the leaf nodes of the decision trees are candidate marketing strategies.
Optionally, the marketing strategy determination apparatus further comprises a model pre-training module, configured to:
acquiring observability data of each user in a user set to form an observability data set, and constructing a sample data set based on the observability data set;
and pre-training a preset basic classification model by using the sample data set to obtain a target classification model.
Optionally, the target classification model includes a random forest model, the sample data set includes basic feature information of each user in the user set, and the model pre-training module is further configured to:
obtaining model training parameters, wherein the model training parameters comprise a depth threshold and a sample threshold;
and splitting by taking the basic characteristic information of each user in the user set as a classification condition to generate a random forest model, wherein the random forest model comprises a plurality of decision trees, the maximum depth of each decision tree is smaller than the depth threshold, leaf nodes of each decision tree are candidate marketing strategies, and the number of samples corresponding to each candidate marketing strategy is not smaller than the sample threshold.
In addition, an embodiment of the present invention further provides a computer-readable storage medium, where a marketing strategy determination program is stored, and when executed by a processor, the marketing strategy determination program implements operations in the marketing strategy determination method provided in the foregoing embodiment.
In addition, an embodiment of the present invention further provides a computer program product, which includes a computer program, and when executed by a processor, the computer program implements the operations in the marketing strategy determination method provided in the foregoing embodiment.
The embodiments of the apparatus, the computer program product, and the computer-readable storage medium of the present invention may refer to the embodiments of the marketing strategy determination method of the present invention, and are not described herein again.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity/action/object from another entity/action/object without necessarily requiring or implying any actual such relationship or order between such entities/actions/objects; the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
For the apparatus embodiment, since it is substantially similar to the method embodiment, it is described relatively simply, and reference may be made to some descriptions of the method embodiment for relevant points. The above-described apparatus embodiments are merely illustrative, in that elements described as separate components may or may not be physically separate. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of the invention. One of ordinary skill in the art can understand and implement it without inventive effort.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention may be substantially or partially embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) as described above and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the marketing strategy determination method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A marketing strategy determination method is characterized by comprising the following steps:
acquiring observability data of a user, and extracting basic feature information of the user from the observability data;
inputting the basic feature information into a pre-trained target classification model, classifying the user based on the basic feature information, and obtaining a decision result of the user;
and determining a target marketing strategy corresponding to the user according to the decision result.
2. The marketing strategy determination method of claim 1, wherein the decision result comprises a plurality of candidate marketing strategies, and the step of determining the target marketing strategy corresponding to the user according to the decision result comprises:
calculating a first profit value corresponding to each candidate marketing strategy in the decision result;
acquiring a second profit value corresponding to the user at present, and calculating a target difference between each first profit value and each second profit value;
and determining a target marketing strategy corresponding to the user from the plurality of marketing strategies according to the target difference.
3. The marketing strategy determination method of claim 2, wherein the step of determining the targeted marketing strategy corresponding to the user from the plurality of marketing strategies according to the targeted difference comprises:
calculating a marketing cost value for converting the current marketing strategy of the user into each candidate marketing strategy;
calculating a profit-improvement value for the user based on the marketing cost value and the target difference value;
determining the target marketing strategy with the maximum profit increase value from the plurality of candidate marketing strategies.
4. The marketing strategy of claim 3, wherein after the step of calculating the revenue enhancement value for the user based on the marketing cost value and the target difference value, further comprising:
and if the income improvement value is negative, setting the current marketing strategy of the user as a target marketing strategy.
5. The marketing strategy determination method of claim 1, wherein the target classification model comprises a random forest model, and the step of inputting the basic feature information into a pre-trained target classification model, classifying the user based on the basic feature information, and obtaining the decision result for the user comprises:
inputting the basic feature information into a pre-trained random forest model, and generating a plurality of decision trees by taking the basic feature information as a classification condition, wherein the maximum depth of each decision tree is smaller than a preset depth threshold;
classifying the users based on the decision trees, determining target leaf nodes corresponding to the users in the decision trees, and obtaining decision results of the users, wherein the leaf nodes of the decision trees are candidate marketing strategies.
6. The marketing strategy of claim 1, wherein the step of inputting the basic feature information into a pre-trained classification model, classifying the user using the basic feature information as a classification condition, and obtaining the decision result of the user further comprises:
acquiring observability data of each user in a user set to form an observability data set, and constructing a sample data set based on the observability data set;
and pre-training a preset basic classification model by using the sample data set to obtain a target classification model.
7. The marketing strategy determination method of claim 6, wherein the target classification model comprises a random forest model, the sample data set comprises basic feature information of each user in the user set, and the step of pre-training a preset basic classification model by using the sample data set to obtain the target classification model comprises:
obtaining model training parameters, wherein the model training parameters comprise a depth threshold and a sample threshold;
and splitting by taking the basic characteristic information of each user in the user set as a classification condition to generate a random forest model, wherein the random forest model comprises a plurality of decision trees, the maximum depth of each decision tree is smaller than the depth threshold, leaf nodes of each decision tree are candidate marketing strategies, and the number of samples corresponding to each candidate marketing strategy is not smaller than the sample threshold.
8. A marketing strategy determination device, characterized in that the marketing strategy determination device comprises:
the characteristic extraction module is used for acquiring observability data of a user and extracting basic characteristic information of the user from the observability data;
the classification decision module is used for inputting the basic characteristic information into a pre-trained target classification model, classifying the user by taking the basic characteristic information as a classification condition, and obtaining a decision result of the user;
and the strategy selection module is used for determining a target marketing strategy corresponding to the user according to the decision result.
9. A terminal device, characterized in that the terminal device comprises: a memory, a processor, and a marketing strategy determination program stored on the memory and executable on the processor, the marketing strategy determination program when executed by the processor implementing the steps of the marketing strategy determination method of any of claims 1-7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a marketing strategy determination program, which when executed by a processor, implements the steps of the marketing strategy determination method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111606916.8A CN114266601A (en) | 2021-12-24 | 2021-12-24 | Marketing strategy determination method and device, terminal equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111606916.8A CN114266601A (en) | 2021-12-24 | 2021-12-24 | Marketing strategy determination method and device, terminal equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114266601A true CN114266601A (en) | 2022-04-01 |
Family
ID=80829938
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111606916.8A Pending CN114266601A (en) | 2021-12-24 | 2021-12-24 | Marketing strategy determination method and device, terminal equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114266601A (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106682949A (en) * | 2016-12-31 | 2017-05-17 | 中国移动通信集团江苏有限公司 | Service recommending method and service information receiving method, device and system |
| CN107679915A (en) * | 2017-09-30 | 2018-02-09 | 平安科技（深圳）有限公司 | Management method, device, equipment and the computer-readable storage medium of marketing activity |
| CN108734297A (en) * | 2017-04-24 | 2018-11-02 | 微软技术授权有限责任公司 | The machine learning commending system of the performance optimization of electronic content items for network transmission |
| CN109598542A (en) * | 2018-11-20 | 2019-04-09 | 阿里巴巴集团控股有限公司 | A kind of put-on method, device and the electronic equipment of equity of marketing |
| CN110717806A (en) * | 2019-09-06 | 2020-01-21 | 中国平安财产保险股份有限公司 | Product information pushing method, device, equipment and storage medium |
| CN110728317A (en) * | 2019-09-30 | 2020-01-24 | 腾讯科技（深圳）有限公司 | Training method and system of decision tree model, storage medium and prediction method |
| CN111292122A (en) * | 2020-01-16 | 2020-06-16 | 支付宝(杭州)信息技术有限公司 | Method and apparatus for facilitating user to perform target behavior for target object |
| CN111598616A (en) * | 2020-05-11 | 2020-08-28 | 百度在线网络技术（北京）有限公司 | Method, device and equipment for realizing object set screening and storage medium |
| CN112150179A (en) * | 2019-06-28 | 2020-12-29 | 京东数字科技控股有限公司 | Information pushing method and device |
| CN113222649A (en) * | 2021-04-28 | 2021-08-06 | 北京三快在线科技有限公司 | Method and device for recommending service execution mode |
| CN113362185A (en) * | 2021-06-07 | 2021-09-07 | 中国工商银行股份有限公司 | Investment portfolio data processing method and device |
-
2021
- 2021-12-24 CN CN202111606916.8A patent/CN114266601A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106682949A (en) * | 2016-12-31 | 2017-05-17 | 中国移动通信集团江苏有限公司 | Service recommending method and service information receiving method, device and system |
| CN108734297A (en) * | 2017-04-24 | 2018-11-02 | 微软技术授权有限责任公司 | The machine learning commending system of the performance optimization of electronic content items for network transmission |
| CN107679915A (en) * | 2017-09-30 | 2018-02-09 | 平安科技（深圳）有限公司 | Management method, device, equipment and the computer-readable storage medium of marketing activity |
| CN109598542A (en) * | 2018-11-20 | 2019-04-09 | 阿里巴巴集团控股有限公司 | A kind of put-on method, device and the electronic equipment of equity of marketing |
| CN112150179A (en) * | 2019-06-28 | 2020-12-29 | 京东数字科技控股有限公司 | Information pushing method and device |
| CN110717806A (en) * | 2019-09-06 | 2020-01-21 | 中国平安财产保险股份有限公司 | Product information pushing method, device, equipment and storage medium |
| CN110728317A (en) * | 2019-09-30 | 2020-01-24 | 腾讯科技（深圳）有限公司 | Training method and system of decision tree model, storage medium and prediction method |
| CN111292122A (en) * | 2020-01-16 | 2020-06-16 | 支付宝(杭州)信息技术有限公司 | Method and apparatus for facilitating user to perform target behavior for target object |
| CN111598616A (en) * | 2020-05-11 | 2020-08-28 | 百度在线网络技术（北京）有限公司 | Method, device and equipment for realizing object set screening and storage medium |
| CN113222649A (en) * | 2021-04-28 | 2021-08-06 | 北京三快在线科技有限公司 | Method and device for recommending service execution mode |
| CN113362185A (en) * | 2021-06-07 | 2021-09-07 | 中国工商银行股份有限公司 | Investment portfolio data processing method and device |
Non-Patent Citations (2)
| Title |
|---|
| 施颖;: "基于目标定位的煤炭企业营销策略研究", 煤炭经济研究, no. 10, 28 October 2016 (2016-10-28) * |
| 汤毅;: ""互联网+"时代下新媒体营销策略的探讨", 营销界, no. 08, 21 February 2020 (2020-02-21) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR20200123015A (en) | Information recommendation method, apparatus, device and medium | |
| CN109615060B (en) | CTR estimation method, CTR estimation device and computer-readable storage medium | |
| CN109993627B (en) | Recommendation method, recommendation model training device and storage medium | |
| CN110162359B (en) | Method, device and system for pushing novice guiding information | |
| CN111061979B (en) | User tag pushing method and device, electronic equipment and medium | |
| CN115917535A (en) | Recommendation model training method, recommendation device and computer readable medium | |
| US20110117537A1 (en) | Usage estimation device | |
| CN113688310B (en) | Content recommendation method, device, equipment and storage medium | |
| WO2021155691A1 (en) | User portrait generating method and apparatus, storage medium, and device | |
| CN111667024B (en) | Content pushing method, device, computer equipment and storage medium | |
| CN111611496A (en) | Product recommendation method and device | |
| CN111626767B (en) | Resource data issuing method, device and equipment | |
| CN113407854A (en) | Application recommendation method, device and equipment and computer readable storage medium | |
| CN110264093B (en) | Credit model establishing method, device, equipment and readable storage medium | |
| CN113657087B (en) | Information matching method and device | |
| CN115222433A (en) | Information recommendation method and device and storage medium | |
| CN112381236A (en) | Data processing method, device, equipment and storage medium for federal transfer learning | |
| CN112632275B (en) | Crowd clustering data processing method, device and equipment based on personal text information | |
| CN116204709A (en) | Data processing method and related device | |
| CN111768218A (en) | Method and device for processing user interaction information | |
| CN111611469A (en) | Identification information determination method and device, electronic equipment and storage medium | |
| CN114266601A (en) | Marketing strategy determination method and device, terminal equipment and storage medium | |
| CN113792952A (en) | Method and apparatus for generating a model | |
| CN114398960A (en) | Target user determination method and device, electronic equipment and storage medium | |
| CN114430504A (en) | Recommendation method and related device for media content |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |