Marketing strategy determination method and device, storage medium and processor

Abstract

The invention discloses a marketing strategy determination method and device, a storage medium and a processor. In the scheme, user information of a target user is acquired; inputting the user information into a complex network prediction model, determining the similarity between the target user and each user node by the complex network prediction model according to the user information, and predicting a marketing difficulty level corresponding to the target user based on the similarity; wherein the complex network prediction model is a network model constructed based on the user information of M users and the actual marketing difficulty level corresponding to each user, and the users are in one-to-one correspondence with the user nodes; and determining a target marketing strategy matched with the target user from a plurality of marketing strategies based on the marketing difficulty level corresponding to the target user. According to the technical scheme, the problem that the marketing effect is poor due to the fact that the marketing strategy matched with the user cannot be accurately determined in the prior art is solved, and the marketing effect is improved.

Description

Technical Field

[0001] The present application relates to the field of financial technology, and in particular to a method, device, storage medium and processor for determining a marketing strategy. Background Art

[0002] In the process of promoting financial products, financial institutions usually adopt some marketing strategies to increase the attractiveness of financial products to users. Taking the marketing strategy for farmer loans as an example, it can adopt low interest rates, flexible repayment methods, fast approval and other methods to enhance the attractiveness and stickiness of users to financial products, so as to help financial institutions improve user satisfaction and loyalty.

[0003] At present, the existing methods for determining marketing strategies usually adopt the mainstream method of mining target groups, using the user's contract information, payment information, transaction records and other data as the main parameters to build a machine learning model, identify potential users through the machine learning model, and determine the marketing strategy that meets the user's needs and preferences based on the characteristics of the potential users. However, when using the mainstream method of mining target groups to determine marketing strategies, it is difficult to accurately identify potential users due to incomplete sample data, resulting in the inability to accurately formulate marketing strategies that match user needs and preferences, resulting in poor marketing results. Summary of the invention

[0004] Based on the above problems, the present application provides a method, device, storage medium and processor for determining a marketing strategy, with the aim of solving the problem that the prior art cannot accurately determine a marketing strategy that matches the user, resulting in poor marketing results, and improving the marketing results.

[0005] The embodiments of the present application disclose the following technical solutions:

[0006] In a first aspect, the present application provides a method for determining a marketing strategy, the method comprising:

[0007] Obtain user information of target users; user information includes basic attribute information and financial information;

[0008] The user information is input into the complex network prediction model, which determines the similarity between the target user and each user node based on the user information, and predicts the marketing difficulty level corresponding to the target user based on the similarity; the complex network prediction model is a network model constructed based on the user information of M users and the actual marketing difficulty level corresponding to each user, where users correspond to user nodes one by one, and M is a positive integer;

[0009] Based on the marketing difficulty level corresponding to the target users, determine the target marketing strategy that matches the target users from multiple marketing strategies.

[0010] In an optional implementation, the complex network prediction model determines the similarity between the target user and each user node based on the user information, including:

[0011] The complex network prediction model is used to classify the user information of the target user to obtain a classification result; the classification result includes N first attribute groups, each of which includes multiple sub-attributes corresponding to the target user and the value of each sub-attribute, and there is an association relationship between the multiple sub-attributes, and N is a positive integer;

[0012] The complex network prediction model obtains N second attribute groups corresponding to each user from each user node according to the type of each first attribute group; each second attribute group includes multiple sub-attributes corresponding to the user and the value of each sub-attribute;

[0013] The complex network prediction model calculates the similarity between the first target attribute group and the second target attribute group corresponding to each user based on the first target attribute group corresponding to the target user and the second target attribute group corresponding to each user; the first target attribute group is any attribute group among the N first attribute groups, the second target attribute group is any attribute group among the N second attribute groups, and the first target attribute group and the second target attribute group are attribute groups of the same type;

[0014] Based on the similarity between the first target attribute group and the second target attribute group corresponding to each user, and the weight value of each second attribute group, the similarity between the target user and each user node is calculated; the weight value of each second attribute group is used to characterize the importance of the second attribute group to evaluating the marketing difficulty level corresponding to the user.

[0015] In an optional implementation, the complex network prediction model calculates the similarity between the first target attribute group and the second target attribute group corresponding to each user based on the first target attribute group corresponding to the target user and the second target attribute group corresponding to each user, including:

[0016] The weight value of each sub-attribute in each second attribute group is obtained by the complex network prediction model; the weight value of each sub-attribute is used to characterize the importance of the sub-attribute in evaluating the marketing difficulty level corresponding to the user;

[0017] The complex network prediction model calculates the similarity between the first target attribute group and the second target attribute group corresponding to each user based on the values ​​of the sub-attributes in the first target attribute group, the values ​​of the sub-attributes in the second target attribute group corresponding to each user, and the weight value of each sub-attribute.

[0018] In an optional implementation, predicting the marketing difficulty level corresponding to the target user based on the similarity includes:

[0019] Based on the similarity between the target user and each user node, determine the user node with the highest similarity to the target user from the M user nodes;

[0020] The marketing difficulty level corresponding to the user node with the highest similarity is used as the marketing difficulty level corresponding to the target user.

[0021] In an optional implementation, the complex network prediction model is obtained through the following process:

[0022] Obtain a sample data set; the sample data set includes user information of M users and an actual marketing difficulty level corresponding to each user;

[0023] According to the actual marketing difficulty level corresponding to each user, the data in the sample data set is divided into multiple communities; each community includes multiple attribute groups corresponding to L users, each attribute group includes multiple sub-attributes and the value of each sub-attribute; L users correspond to the same marketing difficulty level; L is a positive integer, and L is less than M;

[0024] Treat each user in the community as a user node;

[0025] Add edges between each user node in each community to build a complex network;

[0026] Construct a complex network prediction model based on complex networks.

[0027] In an optional implementation, a complex network prediction model is constructed based on the complex network, including:

[0028] The degree of certainty corresponding to each sub-attribute is determined based on the value of the sub-attribute; the degree of certainty is used to characterize the degree of influence of the sub-attribute on the marketing difficulty level corresponding to the evaluation user;

[0029] Determine the weight value of each sub-attribute according to the corresponding degree of certainty of each sub-attribute;

[0030] Based on the weight value corresponding to each sub-attribute, determine the weight value of the attribute group corresponding to the sub-attribute;

[0031] A complex network prediction model is constructed based on the weight values ​​corresponding to each attribute group in each community, the weight values ​​of each sub-attribute in each attribute group, and the complex network.

[0032] In a second aspect of the present application, a device for determining a marketing strategy is provided, the device comprising:

[0033] The acquisition module is used to obtain the user information of the target user; the user information includes basic attribute information and financial information;

[0034] The prediction module is used to input user information into the complex network prediction model, and the complex network prediction model determines the similarity between the target user and each user node based on the user information, and predicts the marketing difficulty level corresponding to the target user based on the similarity; the complex network prediction model is a network model constructed based on the user information of M users and the actual marketing difficulty level corresponding to each user, and the user corresponds to the user node one by one, and M is a positive integer;

[0035] The determination module is used to determine a target marketing strategy that matches the target user from multiple marketing strategies based on the marketing difficulty level corresponding to the target user.

[0036] Optionally, the prediction module includes: a classification unit, which is used to classify the user information of the target user by a complex network prediction model to obtain a classification result; the classification result includes N first attribute groups, each of which includes multiple sub-attributes corresponding to the target user and a value of each sub-attribute, and there is an association relationship between the multiple sub-attributes, and N is a positive integer;

[0037] A data acquisition unit, used to acquire N second attribute groups corresponding to each user from each user node according to the type of each first attribute group by a complex network prediction model; each second attribute group includes multiple sub-attributes corresponding to the user and a value of each sub-attribute;

[0038] A first calculation unit is used to calculate the similarity between the first target attribute group and the second target attribute group corresponding to each user according to the first target attribute group corresponding to the target user and the second target attribute group corresponding to each user by the complex network prediction model; the first target attribute group is any one attribute group among the N first attribute groups, the second target attribute group is any one attribute group among the N second attribute groups, and the first target attribute group and the second target attribute group are attribute groups of the same type;

[0039] The second calculation unit is used to calculate the similarity between the target user and each user node based on the similarity between the first target attribute group and the second target attribute group corresponding to each user, and the weight value of each second attribute group; the weight value of each second attribute group is used to characterize the importance of the second attribute group for evaluating the marketing difficulty level corresponding to the user.

[0040] According to a third aspect of the present application, a computer-readable storage medium is provided, in which a computer program is stored. When the computer program is executed by a processor, the above-mentioned method for determining the marketing strategy is implemented.

[0041] In a fourth aspect of the present application, a processor is provided for running a computer program, and the computer program executes the above-mentioned method for determining the marketing strategy when running.

[0042] Compared with the prior art, this application has the following beneficial effects:

[0043] In the technical solution of the present application, firstly, the user information of the target user is obtained, and the user information includes basic attribute information and financial information; secondly, the user information is input into the complex network prediction model, and the complex network prediction model determines the similarity between the target user and each user node based on the user information, and predicts the marketing difficulty level corresponding to the target user based on the similarity, and then determines the target marketing strategy matching the target user from multiple marketing strategies based on the marketing difficulty level corresponding to the target user. In the technical solution of the present application, since the complex network prediction model is a network model constructed based on the user information of M users and the actual marketing difficulty level corresponding to each user, the user corresponds to the user node in the network model one by one, and thus the complex network prediction model can determine the similarity between the target user and each user who has completed the marketing based on the user information of the target user and the similarity between the user information of the M users who have completed the marketing, and accurately predict the marketing difficulty level corresponding to the target user based on the similarity; thus, based on the accurate marketing difficulty level, the target marketing strategy with the highest matching degree with the target user can be determined from multiple marketing strategies, which solves the problem that the existing technology cannot accurately determine the marketing strategy matching the user, resulting in poor marketing effect, and improves the marketing effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0045] Figure 1 A flowchart of a method for determining a marketing strategy provided in an embodiment of the present application;

[0046] Figure 2 A flowchart of a process for determining the similarity between a target user and each user node provided in an embodiment of the present application;

[0047] Figure 3 A flowchart of a process for determining a similarity between a first target attribute group and a second target attribute group provided in an embodiment of the present application;

[0048] Figure 4 A schematic diagram of a complex network provided in an embodiment of the present application;

[0049] Figure 5A schematic diagram of a weight discrete constraint provided in an embodiment of the present application;

[0050] Figure 6 A flowchart of another method for determining a marketing strategy provided in an embodiment of the present application;

[0051] Figure 7 A schematic diagram of the structure of a device for determining a marketing strategy provided in an embodiment of the present application. DETAILED DESCRIPTION

[0052] As described above, existing technologies usually adopt the mainstream target group mining method, using the user's credit information, contract information, payment information, transaction records and other data as the main parameters to build a machine learning model, identify potential users through the machine learning model, and determine the marketing strategy that meets the user's needs and preferences based on the characteristics of the potential users. However, the use of the mainstream target group mining method to determine the marketing strategy is prone to the inability to accurately identify potential users due to inaccurate and incomplete sample data, resulting in low accuracy of the determined marketing strategy, making it difficult to achieve precision marketing for users.

[0053] After research, the inventor proposed a method for determining a marketing strategy. The method first obtains user information of the target user, which includes basic attribute information and financial information; then the user information is input into a complex network prediction model, and the complex network prediction model determines the similarity between the target user and each user node based on the user information, and predicts the marketing difficulty level corresponding to the target user based on the similarity, and determines the target marketing strategy that matches the target user from multiple marketing strategies based on the marketing difficulty level corresponding to the target user. In the present application, since the complex network prediction model is a network model constructed based on the user information of M users and the actual marketing difficulty level corresponding to each user, the users correspond one-to-one to the user nodes in the network model. Therefore, the complex network prediction model can determine the similarity between the target user and each user who has completed the marketing based on the user information of the target user and the similarity between the user information of M users who have completed the marketing, and based on the similarity, the marketing difficulty level corresponding to the target user can be accurately predicted; thus, based on the accurate marketing difficulty level, the target marketing strategy with the highest matching degree with the target user can be determined from multiple marketing strategies, thereby solving the problem that the prior art cannot accurately determine the marketing strategy that matches the user, resulting in poor marketing effect, thereby improving the marketing effect.

[0054] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0055] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards in the relevant regions.

[0056] It is understandable that in the specific implementation of this application, related data such as user information is involved. When the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data need to comply with relevant laws, regulations and standards of the relevant regions.

[0057] Keyword definitions:

[0058] Complex network: A network with some or all of the properties of self-organization, self-similarity, small world, and scale-free. Complex networks aim to understand the interactions between nodes (vertices) and edges (connections) in the network and their impact on the overall behavior of the network.

[0059] Community: It is a grouping of nodes in a complex network, where the connections within the group are relatively tight and the connections between groups are relatively sparse.

[0060] Method Embodiment

[0061] The present application provides an embodiment of a method for determining a marketing strategy. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions. In addition, although the logical order is shown in the flowchart, it is only an example order provided. In some cases, the steps shown or described can be executed in a different order than here.

[0062] See also Figure 1 , which is a flow chart of a method for determining a marketing strategy provided by an embodiment of the present application, such as Figure 1 As shown, the method comprises the following steps:

[0063] Step S101, obtaining user information of a target user.

[0064] In an optional embodiment, a marketing strategy determination system can be used as the execution subject of the marketing strategy determination method of the embodiment of the present application. For the convenience of description, the marketing strategy determination system is referred to as the system below.

[0065] In step S101, the user information includes basic attribute information and financial information. For example, the user information may include basic personal information, family income and expenditure information, business information, asset information, liability information, etc. Among them, the basic personal information includes user type, business entity type, permanent address, occupation, education level and marital status, etc., the family income and expenditure information includes the total family population, labor force, dependent population, family annual income, family annual expenditure, etc., the business information includes the planting business status, the breeding business status, the private business status (such as small shops, small workshops, family farms, etc. operated by individuals), etc., the asset and liability information includes deposit status, real estate value information, automobile value information, agricultural machinery value, biological asset value, etc., and the liability information includes loan status, guarantee status, private loan status, other liabilities, etc.

[0066] Step S102, inputting the user information into the complex network prediction model, the complex network prediction model determines the similarity between the target user and each user node based on the user information, and predicts the marketing difficulty level corresponding to the target user based on the similarity.

[0067] In step S102, the complex network prediction model is a network model constructed based on the user information of M users and the actual marketing difficulty level corresponding to each user, where users correspond to user nodes one by one, and M is a positive integer. The actual marketing difficulty level can be specifically shown in Table 1.

[0068] Table 1

[0069] Actual Marketing Difficulty Level Difficulty of marketing to users Level 1 easy Level 2 generally Level 3 Difficult Level 4 difficulty Level 5 Marketing Failure

[0070] It can be seen from the contents in Table 1 that the higher the user's marketing difficulty level, the higher the marketing difficulty for the user, and the lower the success rate of marketing to the user.

[0071] In order to accurately predict the marketing difficulty level corresponding to the target user, in this embodiment, when the system inputs the user information of the target user into the complex network prediction model, the complex network prediction model can add the target user as a new user node to the complex network, and determine the similarity between the target user and each user who has completed the marketing based on the similarity between the user information of the target user and the user information of the M users who have completed the marketing, and determine the marketing difficulty level corresponding to the target user based on the similarity. Specifically, see Figure 2 , Figure 2A flowchart of a process for determining the similarity between a target user and each user node provided in an embodiment of the present application, the process comprising the following steps:

[0072] Step S201 , classifying the user information of the target user by a complex network prediction model to obtain a classification result; the classification result includes N first attribute groups.

[0073] In step S201, each first attribute group includes multiple sub-attributes corresponding to the target user and the value of each sub-attribute, and there is an association relationship between the multiple sub-attributes, and N is a positive integer. The data in the first attribute group can be specifically shown in Table 2.

[0074] Table 2

[0075]

[0076] It should be noted that, in this embodiment, the divisions of the first attribute group shown in Table 2 may be increased or decreased according to actual conditions, and the sub-attributes included in the first attribute group may also be increased or decreased accordingly.

[0077] It should be noted that by classifying the user information of the target user, a more detailed analysis of the user information can be performed to reveal the relationship between the user information of different users, thereby providing data preparation for the subsequent determination of the similarity between the target user and each user node.

[0078] Step S202 : The complex network prediction model obtains N second attribute groups corresponding to each user from each user node according to the type of each first attribute group.

[0079] In step S202, each second attribute group includes multiple sub-attributes corresponding to the user and the value of each sub-attribute. For example, when the second attribute group is personal basic status, the sub-attributes are: total family population, labor force, dependent population, annual family income, annual family expenditure, etc.; when the second attribute group is debt status, the sub-attributes are: loan status, guarantee status, private loan status, other debts, etc.

[0080] In this embodiment, each user node stores N second attribute groups of the user corresponding to the user node. For example, when the type of the first attribute group is personal basic status, family income and expenditure status, business status, asset status and liability status, the system can obtain the attribute groups such as personal basic status, family income and expenditure status, business status, asset status and liability status corresponding to each user from each user node through the complex network prediction model.

[0081] Step S203 : The complex network prediction model calculates the similarity between the first target attribute group and the second target attribute group corresponding to each user based on the first target attribute group corresponding to the target user and the second target attribute group corresponding to each user.

[0082] In step S203, the first target attribute group is any one of the N first attribute groups, the second target attribute group is any one of the N second attribute groups, and the first target attribute group and the second target attribute group are attribute groups of the same type.

[0083] For example, when the type of the first target attribute group and the second target attribute group is personal basic conditions, the system can use a complex network prediction model to calculate the similarity between the personal basic conditions corresponding to the target user and the personal basic conditions corresponding to each user based on the data in the personal basic conditions corresponding to the target user and the data in the personal basic conditions corresponding to each user. For another example, when the type of the first target attribute group and the second target attribute group is asset conditions, the system can use a complex network prediction model to calculate the similarity between the asset conditions corresponding to the target user and the asset conditions corresponding to each user based on the data in the asset conditions corresponding to the target user and the data in the asset conditions corresponding to each user.

[0084] Specifically, see Figure 3 , Figure 3 A flowchart of a process for determining a similarity between a first target attribute group and a second target attribute group provided in an embodiment of the present application, the process comprising the following steps:

[0085] Step S2031: Obtain the weight value of each sub-attribute in each second attribute group by using the complex network prediction model.

[0086] In step S2031, the weight value of each sub-attribute is used to characterize the importance of the sub-attribute in evaluating the marketing difficulty level corresponding to the user.

[0087] Step S2032: The complex network prediction model calculates the similarity between the first target attribute group and the second target attribute group corresponding to each user based on the values ​​of the sub-attributes in the first target attribute group, the values ​​of the sub-attributes in the second target attribute group corresponding to each user, and the weight value of each sub-attribute.

[0088] In this embodiment, the higher the similarity between the sub-attributes in the first target attribute group and the sub-attributes in the second target attribute group, the higher the similarity between the first target attribute group and the second target attribute group corresponding to each user. The system can calculate the similarity between the first target attribute group and the second target attribute group corresponding to each user by formula (1) and formula (2):

[0089]

[0090] Wherein, q is the user node corresponding to the target user, p is the user node corresponding to any user; k is the identifier of the community to which any user belongs, K=1 indicates that the user belongs to the first community, and K=2 indicates that the user belongs to the second community; m is the number of sub-attributes in the attribute group (i.e., the first target attribute group and the second target attribute group); is the weight value of the i-th sub-attribute in the attribute group; is the i-th sub-attribute in the attribute group corresponding to the target user, is the i-th sub-attribute in the attribute group corresponding to the user; The value range is (0,1].

[0091] It should be noted that, since the weight value of each sub-attribute can be used to characterize the importance of the sub-attribute to evaluating the marketing difficulty level corresponding to the user, the similarity between the first target attribute group and the second target attribute group corresponding to each user is calculated based on the values ​​of the sub-attributes in the first target attribute group, the values ​​of the sub-attributes in the second target attribute group corresponding to each user, and the weight value of each sub-attribute. This can avoid the problem of being unable to accurately calculate the similarity between the first target attribute group and the second target attribute group corresponding to each user due to missing data, thereby improving the accuracy of the calculated similarity.

[0092] In order to accurately predict the marketing difficulty level corresponding to the target user, thereby improving the accuracy of determining the marketing strategy, in this embodiment, the system can construct a complex network prediction model based on the complex network, based on the user information of M users, and the actual marketing difficulty level corresponding to each user.

[0093] Specifically, the system can obtain a sample data set; the sample data set includes user information of M users and the actual marketing difficulty level corresponding to each user; then the system can divide the data in the sample data set into multiple communities according to the actual marketing difficulty level corresponding to each user; each community includes multiple attribute groups corresponding to L users, each attribute group includes multiple sub-attributes and the value of each sub-attribute; L users correspond to the same marketing difficulty level; L is a positive integer, and L is less than M; then the users in each community are taken as a user node, and edges are added between the user nodes in each community to construct a complex network; finally, the system can construct a complex network prediction model based on the complex network.

[0094] In this embodiment, the system can use users as user nodes, user information as node attributes, similar marketing difficulty levels as edge standards, multiple users with the same marketing difficulty level as a community, and add edges between user nodes in the same community to construct Figure 4 The complex network shown in FIG. 1 is used to construct a complex network prediction model based on the complex network. The complex network may include Figure 4 In the community 1, community 2, community 3 and community 4, the user nodes in each community are connected by edges, wherein the system can determine the edge probability by the similarity between two user nodes, and determine whether the two user nodes are connected based on the edge probability. Among them, each community corresponds to a marketing difficulty level, for example, the marketing difficulty level corresponding to community 1 is level 1, the marketing difficulty level corresponding to community 2 is level 2, the marketing difficulty level corresponding to community 3 is level 3, and the marketing difficulty level corresponding to community 4 is level 4. Each user node in each community contains the same attribute group, and the sub-attributes in each attribute group are the same.

[0095] In order to improve the model accuracy of the complex network prediction model, in this embodiment, the system can build a complex network prediction model based on the complex network, the weight values ​​of each attribute group, and the weight values ​​of each sub-attribute in each attribute group. The process includes the following steps:

[0096] Step 1: The system may determine the degree of certainty corresponding to each sub-attribute according to the value of the sub-attribute; the degree of certainty is used to characterize the degree of influence of the sub-attribute on the marketing difficulty level corresponding to the evaluation user;

[0097] In step 1, the system can determine the probability of occurrence of various conditions of each sub-attribute in the community to which it belongs based on the value of each sub-attribute, and determine the degree of certainty corresponding to the sub-attribute based on the probability of occurrence corresponding to the sub-attribute. The probability of occurrence corresponding to the sub-attribute is used to characterize the possibility of the sub-attribute appearing in the community to which it belongs. For example, one of the sub-attributes in community 1 is marital status. Based on the marital status of all user nodes in the community (e.g., married, unmarried, divorced), the system determines the possibility of the marital status being married in the community, the possibility of the marital status being unmarried in the community, and the possibility of the marital status being divorced in the community.

[0098] In this embodiment, the system can determine the degree of certainty corresponding to the sub-attribute through formula (3).

[0099]

[0100] in, is the certainty of the th sub-attribute, k is the identifier of the community to which any user belongs; i is a positive integer, and i∈(0,h], h represents the classification of the sub-attribute; is the occurrence probability corresponding to the sub-attribute.

[0101] Step 2: Determine the weight value of each sub-attribute according to the certainty of each sub-attribute.

[0102] In step 2, the system may determine the weight value of each sub-attribute using formula (4).

[0103]

[0104] in, is the certainty of each sub-attribute, The weight value for each sub-attribute.

[0105] Optionally, in order to improve the prediction accuracy of the complex network prediction model and avoid the model being overfitted to the same result, in this embodiment, the system can perform a year-on-year amplification processing on the weight value of each sub-attribute so that the minimum value of the weight value is 1, and set an upper limit value of the discrete degree (variance) of the weight ratio of the attribute group. (the lower limit is 0), such as Figure 5 As shown, when the weight value exceeds When (such as Figure 5 The weight value shown by the red dot in the figure is calculated. The maximum weight of the multiple sub-attributes of the attribute group is reduced by 1, the minimum weight is increased by 1, and the discrete degree (variance) of the attribute group weight is recalculated until the discrete degree of the attribute group weight is less than (like Figure 5 The weight value is shown by the black dot in the figure).

[0106] Step three: determine the weight value of the attribute group corresponding to each sub-attribute based on the weight value corresponding to the sub-attribute.

[0107] In step 3, the system may calculate the weight value of the attribute group using a normalized weighted sum method. Specifically, the system may first add up the weight values ​​of all sub-attributes in the attribute group to obtain a total weight value corresponding to the attribute group, and then normalize the total weight value to obtain a weight value corresponding to the attribute group.

[0108] Step 4: construct a complex network prediction model based on the weight values ​​of the attribute groups corresponding to each user node in each community, the weight values ​​of each sub-attribute in each attribute group, and the complex network.

[0109] Optionally, the system can adjust the weight values ​​of different attribute groups and the upper limit value of the discrete degree L k j , the cross-validation method is used to calculate the prediction accuracy of the complex network prediction model, and this is used as an indicator to obtain the ideal parameter combination of attribute group weights and edge number thresholds, and finally determine the optimal prediction model.

[0110] It should be noted that the above steps 1 to 4 are not shown in the figure.

[0111] Step S204: Calculate the similarity between the target user and each user node based on the similarity between each first target attribute group and each second target attribute group, and the weight value of each second attribute group.

[0112] In step S204, the weight value of each second attribute group is used to represent the importance of the second attribute group to evaluating the marketing difficulty level corresponding to the user.

[0113] In this embodiment, the system can calculate the similarity S between the target user and each user node through formula (5): k (q,p).

[0114]

[0115] Among them, q is the user node corresponding to the target user, p is the user node corresponding to any user; k is the identifier of the community to which any user belongs; is the weight value of the i-th attribute group corresponding to node p, is the similarity between the i-th attribute group corresponding to node q and the i-th attribute group corresponding to node p. k (q,p) ranges from [0,1], S k The larger the (q,p) value is, the easier it is for nodes q to be connected to nodes p. That is, the higher the similarity between nodes q and p, the greater the probability that nodes q and p belong to community k.

[0116] It should be noted that since the weight value of the second attribute group is used to characterize the importance of the second attribute group to evaluating the marketing difficulty level corresponding to the user, based on the similarity between each first target attribute group and each second target attribute group, and the weight value of each second attribute group, the similarity between the target user and each user node can be accurately calculated, avoiding the problem of being unable to accurately calculate the similarity between the target user and each user node due to missing data.

[0117] In order to further accurately determine the marketing difficulty level corresponding to the target user, in this embodiment, the system can determine the user node with the highest similarity to the target user from M user nodes based on the similarity between the target user and each user node; then the system can use the marketing difficulty level of the user corresponding to the user node with the highest similarity as the marketing difficulty level corresponding to the target user.

[0118] For example, when M=5, the similarities between the target user and user node 1, user node 2, user node 3, user node 4 and user node 5 are 50%, 60%, 78%, 88% and 90% respectively, and the marketing difficulty levels of user node 1 to user node 5 are level one, level two, level three, level two and level five respectively, the system may use the marketing difficulty level of user node 5 with a similarity of 90%: level five as the marketing difficulty level corresponding to the target user.

[0119] Step S103, determining a target marketing strategy that matches the target user from multiple marketing strategies based on the marketing difficulty level corresponding to the target user.

[0120] In this embodiment, different marketing difficulty levels correspond to different marketing strategies. For example, when the marketing difficulty level is level 5, the marketing personnel configuration in the corresponding marketing strategy is the highest-level personnel configuration; when the marketing difficulty level is level 1, the marketing personnel configuration in the corresponding marketing strategy is the ordinary-level personnel configuration.

[0121] Optionally, after determining the target marketing strategy that matches the target user, the system can obtain the marketing results of marketing to the target user based on the target marketing strategy, judge the actual marketing difficulty level of the target user based on the marketing results, and update the complex network prediction model based on the actual marketing difficulty level of the target user and user information.

[0122] In the application scenario where the users are farmers, Figure 6 As shown, the system can acquire data, obtain user information (such as personal basic information, family income and expenditure status, debt status, etc.) and actual marketing difficulty levels of multiple farmers who have completed marketing, and then the system can build a complex network based on the actual marketing difficulty levels of multiple farmers, and determine weight values ​​based on the user information of multiple farmers, and then the system can build a complex network prediction model based on the complex network and the weight values. When determining the marketing strategy corresponding to the farmers to be marketed, the system can obtain the user information (such as personal basic information, family income and expenditure status, debt status, etc.) of the farmers to be marketed, and input the user information of the farmers to be marketed into the complex network prediction model, and the complex network prediction model predicts the marketing difficulty level of the farmers to be marketed, and then the system can determine the target marketing strategy with the highest matching degree with the farmers to be marketed based on the marketing difficulty level of the farmers to be marketed, thereby improving the accuracy of determining the marketing strategy, thereby realizing the precision marketing of farmers' loans.

[0123] Through the method for determining the marketing strategy provided in this embodiment, a complex network prediction model is constructed based on the user information of M users and the actual marketing difficulty level corresponding to each user, so that the users correspond one-to-one with the user nodes in the model. Therefore, through the complex network prediction model, the similarity between the target user and each user who has completed the marketing can be determined based on the similarity between the user information of the target user and the user information of M users who have completed the marketing, and the marketing difficulty level corresponding to the target user can be accurately predicted based on the similarity; therefore, based on the accurate marketing difficulty level, the target marketing strategy with the highest matching degree with the target user can be determined from multiple marketing strategies, which solves the problem of poor marketing effect caused by the inability to accurately determine the marketing strategy that matches the user in the prior art, thereby improving the marketing effect.

[0124] The method for determining the marketing strategy provided in this embodiment realizes the reconstruction of data from a network perspective based on user basic attribute data. The collected user data is networked by constructing a complex network. The self-organization, self-similarity, small-world and scale-free characteristics of complex networks are utilized to construct attribute similarity link predictions (i.e., complex network prediction models) for the same attributes of users. Edges are established through the similarity between nodes. The existing similar user data is used to divide the community structure, strip off the characteristics of multiple user communities, highlight the differences in behavioral characteristics between community groups, and improve the accuracy of determining the user marketing difficulty level.

[0125] Device Embodiment

[0126] The embodiment of the present application provides a device for determining a marketing strategy, wherein: Figure 7 A schematic diagram of a marketing strategy determination device provided in an embodiment of the present application is shown in FIG. Figure 7 As shown, the device includes: an acquisition module 11, a prediction module 12 and a determination module 13. Figure 7 You can see the connection relationship between several modules.

[0127] The acquisition module 11 is used to acquire the user information of the target user; the user information includes basic attribute information and financial information;

[0128] The prediction module 12 is used to input the user information into the complex network prediction model, and the complex network prediction model determines the similarity between the target user and each user node based on the user information, and predicts the marketing difficulty level corresponding to the target user based on the similarity; the complex network prediction model is a network model constructed based on the user information of M users and the actual marketing difficulty level corresponding to each user, and the user corresponds to the user node one by one, and M is a positive integer;

[0129] The determination module 13 is used to determine a target marketing strategy that matches the target user from multiple marketing strategies based on the marketing difficulty level corresponding to the target user.

[0130] Optionally, the prediction module includes: a classification unit, a data acquisition unit, a first calculation unit and a second calculation unit.

[0131] The classification unit is used to classify the user information of the target user by the complex network prediction model to obtain a classification result; the classification result includes N first attribute groups, each of which includes multiple sub-attributes corresponding to the target user and the value of each sub-attribute, and there is an association relationship between the multiple sub-attributes, and N is a positive integer;

[0132] A data acquisition unit, used to acquire N second attribute groups corresponding to each user from each user node according to the type of each first attribute group by a complex network prediction model; each second attribute group includes multiple sub-attributes corresponding to the user and a value of each sub-attribute;

[0133] A first calculation unit is used to calculate the similarity between the first target attribute group and the second target attribute group corresponding to each user according to the first target attribute group corresponding to the target user and the second target attribute group corresponding to each user by the complex network prediction model; the first target attribute group is any one attribute group among the N first attribute groups, the second target attribute group is any one attribute group among the N second attribute groups, and the first target attribute group and the second target attribute group are attribute groups of the same type;

[0134] The second calculation unit is used to calculate the similarity between the target user and each user node based on the similarity between the first target attribute group and the second target attribute group corresponding to each user, and the weight value of each second attribute group; the weight value of each second attribute group is used to characterize the importance of the second attribute group for evaluating the marketing difficulty level corresponding to the user.

[0135] Optionally, the first calculation unit includes: a first acquisition unit and a third calculation unit.

[0136] The first acquisition unit is used to acquire the weight value of each sub-attribute in each second attribute group by the complex network prediction model; the weight value of each sub-attribute is used to characterize the importance of the sub-attribute in evaluating the marketing difficulty level corresponding to the user;

[0137] The third calculation unit is used to calculate the similarity between the first target attribute group and the second target attribute group corresponding to each user based on the values ​​of the sub-attributes in the first target attribute group, the values ​​of the sub-attributes in the second target attribute group corresponding to each user, and the weight value of each sub-attribute by the complex network prediction model.

[0138] Optionally, the prediction module further includes: a first determination unit and a second determination unit.

[0139] The first determining unit is used to determine the user node with the highest similarity to the target user from the M user nodes based on the similarity between the target user and each user node;

[0140] The second determining unit is used to use the marketing difficulty level corresponding to the user node with the highest similarity as the marketing difficulty level corresponding to the target user.

[0141] Optionally, the marketing strategy determination device further includes: a second acquisition unit, a data division unit, a third determination unit, a first construction unit and a second construction unit.

[0142] The second acquisition unit is used to acquire a sample data set; the sample data set includes user information of M users and an actual marketing difficulty level corresponding to each user;

[0143] The data partitioning unit is used to partition the data in the sample data set into a plurality of communities according to the actual marketing difficulty level corresponding to each user; each community includes a plurality of attribute groups corresponding to L users, each attribute group includes a plurality of sub-attributes and a value of each sub-attribute; the L users correspond to the same marketing difficulty level; L is a positive integer, and L is less than M;

[0144] A third determining unit is used to regard the users in each community as a user node;

[0145] The first construction unit is used to add edges between each user node in each community to construct a complex network;

[0146] The second construction unit is used to construct a complex network prediction model based on the complex network.

[0147] Optionally, the second constructing unit includes: a fourth determining unit, a fifth determining unit, a sixth determining unit and a third constructing unit.

[0148] The fourth determination unit is used to determine the degree of certainty corresponding to each sub-attribute according to the value of the sub-attribute; the degree of certainty is used to characterize the influence of the sub-attribute on the marketing difficulty level corresponding to the evaluation user;

[0149] A fifth determining unit, configured to determine a weight value of each sub-attribute according to a degree of certainty corresponding to each sub-attribute;

[0150] a sixth determining unit, configured to determine, based on the weight value corresponding to each sub-attribute, a weight value of the attribute group corresponding to the sub-attribute;

[0151] The third construction unit is used to construct a complex network prediction model based on the weight values ​​corresponding to each attribute group in each community, the weight values ​​of each sub-attribute in each attribute group, and the complex network.

[0152] Storage Medium Embodiments

[0153] The embodiment of the present application provides a computer-readable storage medium, on which a program is stored, wherein when the program is executed by a processor, some or all of the steps in the method for determining the marketing strategy described in the aforementioned method embodiment of the present application are implemented. The storage medium can be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0154] Processor Embodiments

[0155] An embodiment of the present application provides a processor for running a program, wherein part or all of the steps in the method for determining a marketing strategy introduced in the aforementioned method embodiment are executed when the program is running.

[0156] It should be noted that each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The device embodiment described above is merely schematic, in which the unit described as a separate component may or may not be physically separated, and the component prompted as a unit may or may not be a physical unit, that is, it may be located in one place, or it may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative work.

[0157] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for determining a marketing strategy, characterized in that: include: Get the user information of the target user; The user information includes basic attribute information and financial information; The user information is input into a complex network prediction model, and the complex network prediction model determines the similarity between the target user and each user node based on the user information, and predicts the marketing difficulty level corresponding to the target user based on the similarity; the complex network prediction model is a network model constructed based on the user information of M users and the actual marketing difficulty level corresponding to each of the users, the users correspond to the user nodes one by one, and M is a positive integer; Based on the marketing difficulty level corresponding to the target user, a target marketing strategy matching the target user is determined from multiple marketing strategies.

2. The method according to claim 1, characterized in that The determining, by the complex network prediction model, the similarity between the target user and each user node according to the user information includes: The complex network prediction model performs data classification on the user information of the target user to obtain a classification result; the classification result includes N first attribute groups, each of which includes multiple sub-attributes corresponding to the target user and a value of each sub-attribute, and there is an association relationship between the multiple sub-attributes, and N is a positive integer; The complex network prediction model obtains N second attribute groups corresponding to each user from each user node according to the type of each first attribute group; each second attribute group includes multiple sub-attributes corresponding to the user and the value of each sub-attribute; The complex network prediction model calculates the similarity between the first target attribute group corresponding to the target user and the second target attribute group corresponding to each of the users based on the first target attribute group corresponding to the target user and the second target attribute group corresponding to each of the users; the first target attribute group is any one of the N first attribute groups, the second target attribute group is any one of the N second attribute groups, and the first target attribute group and the second target attribute group are attribute groups of the same type; Based on the similarity between the first target attribute group and the second target attribute group corresponding to each of the users, and the weight value of each of the second attribute groups, the similarity between the target user and each of the user nodes is calculated; the weight value of each of the second attribute groups is used to characterize the importance of the second attribute group in evaluating the marketing difficulty level corresponding to the user.

3. The method according to claim 2, characterized in that The calculating, by the complex network prediction model, the similarity between the first target attribute group corresponding to the target user and the second target attribute group corresponding to each of the users, includes: The weight value of each of the sub-attributes in each of the second attribute groups is obtained by the complex network prediction model; the weight value of each of the sub-attributes is used to characterize the importance of the sub-attribute in evaluating the marketing difficulty level corresponding to the user; The complex network prediction model calculates the similarity between the first target attribute group and the second target attribute group corresponding to each of the users based on the values ​​of the sub-attributes in the first target attribute group, the values ​​of the sub-attributes in the second target attribute group corresponding to each of the users, and the weight value of each of the sub-attributes.

4. The method according to claim 1, characterized in that: The predicting the marketing difficulty level corresponding to the target user based on the similarity includes: Based on the similarity between the target user and each user node, determining a user node with the highest similarity to the target user from the M user nodes; The marketing difficulty level corresponding to the user node with the highest similarity is used as the marketing difficulty level corresponding to the target user.

5. The method according to claim 1, characterized in that The complex network prediction model is obtained through the following process: Acquire a sample data set; the sample data set includes user information of M users and an actual marketing difficulty level corresponding to each user; According to the actual marketing difficulty level corresponding to each user, the data in the sample data set is divided into a plurality of communities; each community includes a plurality of attribute groups corresponding to L users, each attribute group includes a plurality of sub-attributes and a value of each sub-attribute; the L users correspond to the same marketing difficulty level; L is a positive integer, and L is less than M; Taking each user in the community as a user node; Adding edges between the user nodes in each community to construct a complex network; The complex network prediction model is constructed based on the complex network.

6. The method according to claim 5, characterized in that The constructing the complex network prediction model according to the complex network includes: Determining the degree of certainty corresponding to each of the sub-attributes according to the value of the sub-attribute; the degree of certainty is used to characterize the degree of influence of the sub-attribute on evaluating the marketing difficulty level corresponding to the user; Determining a weight value of each of the sub-attributes according to the degree of certainty corresponding to each of the sub-attributes; Based on the weight value corresponding to each of the sub-attributes, determine the weight value of the attribute group corresponding to the sub-attribute; The complex network prediction model is constructed based on the weight values ​​corresponding to each of the attribute groups in each of the communities, the weight values ​​of each of the sub-attributes in each of the attribute groups, and the complex network.

7. A device for determining a marketing strategy, characterized in that: include: An acquisition module is used to obtain user information of a target user; The user information includes basic attribute information and financial information; A prediction module, used for inputting the user information into a complex network prediction model, wherein the complex network prediction model determines the similarity between the target user and each user node based on the user information, and predicts the marketing difficulty level corresponding to the target user based on the similarity; the complex network prediction model is a network model constructed based on the user information of M users and the actual marketing difficulty level corresponding to each of the users, wherein the users correspond to the user nodes one by one, and M is a positive integer; The determination module is used to determine a target marketing strategy that matches the target user from multiple marketing strategies based on the marketing difficulty level corresponding to the target user.

8. The device according to claim 7, characterized in that The prediction module comprises: A classification unit, configured to classify the user information of the target user by using the complex network prediction model to obtain a classification result; the classification result includes N first attribute groups, each of which includes a plurality of sub-attributes corresponding to the target user and a value of each of the sub-attributes, and there is an association relationship between the plurality of sub-attributes, and N is a positive integer; A data acquisition unit, configured to acquire, by the complex network prediction model, N second attribute groups corresponding to each user from each user node according to the type of each first attribute group; each second attribute group includes a plurality of sub-attributes corresponding to the user and a value of each sub-attribute; A first calculation unit is configured to calculate, by the complex network prediction model, a similarity between the first target attribute group corresponding to the target user and the second target attribute group corresponding to each of the users; the first target attribute group is any one of the N first attribute groups, the second target attribute group is any one of the N second attribute groups, and the first target attribute group and the second target attribute group are attribute groups of the same type; The second calculation unit is used to calculate the similarity between the target user and each of the user nodes based on the similarity between the first target attribute group and the second target attribute group corresponding to each of the users, and the weight value of each of the second attribute groups; the weight value of each of the second attribute groups is used to characterize the importance of the second attribute group for evaluating the marketing difficulty level corresponding to the user.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for determining a marketing strategy according to any one of claims 1 to 6 is implemented.

10. A processor, characterized in that: Used to run a computer program, which, when running, executes the method for determining a marketing strategy according to any one of claims 1 to 6.

Images