Information delivery method, device and equipment

Abstract

The embodiment of the present application provides a method, device and equipment for information delivery, which may include: when delivering information, first receiving the operation instructions including configuration parameters and / or business demand parameters input by the user on the operation interface, that is, taking into account the business demands of different personalized groups to spread, and determining the candidate group and target feature information according to the configuration parameters and / or business demand parameters input by the user, and then jointly determining the target candidate group in the candidate group according to the target feature information and seed population corresponding to each candidate user in the candidate group, effectively increasing the flexibility and humanization of population diffusion, and then delivering information to the target candidate group. Compared with the prior art of directly delivering information to the diffusion population of the seed population, the accuracy of information delivery is improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of data processing technology, and in particular to an information delivery method, device and equipment. Background Art

[0002] In Internet companies, it is usually necessary to first identify the diffusion group among the majority of users, and then deliver information to this diffusion group through advertising, coupons, text messages, etc., in order to achieve marketing goals.

[0003] In the prior art, when information is delivered, the seed population is diffused to obtain the diffusion population, and information is delivered to the diffusion population. The seed population refers to the population with the same needs and interests in products and services; the population with the same characteristics as the seed population is called the extended population, and the number of extended populations is usually several times that of the seed population. However, when the seed population is large, the number of diffusion populations will be large; and since there may be redundant or low-credibility non-high-quality seed users in the seed population, it will have a certain impact on the determination of the diffusion population, resulting in low accuracy of information delivery.

[0004] Therefore, when delivering information, how to improve the accuracy of information delivery is an urgent problem to be solved by those skilled in the art. Summary of the invention

[0005] The embodiments of the present application provide a method, apparatus and device for information delivery, which improve the accuracy of information delivery.

[0006] In a first aspect, an embodiment of the present application provides an information delivery method, which may include:

[0007] An operation instruction input on an operation interface is received, wherein the operation instruction includes configuration parameters and / or business requirement parameters.

[0008] A candidate group and target characteristic information are determined according to the configuration parameters and / or business requirement parameters; the candidate group includes a plurality of candidate users.

[0009] A seed population is obtained; the seed population includes a plurality of seed users.

[0010] A target candidate group is determined in the candidate group according to the target feature information corresponding to each candidate user in the candidate group and the seed group.

[0011] Deliver information to the target candidate group.

[0012] It can be seen that the information delivery method provided by the embodiment of the present application, when delivering information, first receives the operation instructions including configuration parameters and / or business demand parameters input by the user on the operation interface, that is, takes into account the business demands of different personalized groups of people to spread, and determines the candidate group and target feature information according to the configuration parameters and / or business demand parameters input by the user, and then determines the target candidate group in the candidate group according to the target feature information and seed population corresponding to each candidate user in the candidate group, effectively increasing the flexibility and humanization of population diffusion, and then delivering information to the target candidate group. Compared with the prior art of directly delivering information to the diffusion population of the seed population, the accuracy of information delivery is improved.

[0013] In a possible implementation, determining a target candidate group from the candidate group according to the target feature information corresponding to each candidate user in the candidate group and the seed group may include:

[0014] The multiple seed users are clustered according to the target feature information corresponding to each seed user in the seed population to obtain M clusters; wherein the center point of each cluster is used as a cluster center user; and M is an integer greater than or equal to 2.

[0015] The correlation degree corresponding to each candidate user is determined according to the target feature information corresponding to each candidate user in the candidate group and the target feature information corresponding to the M cluster center users.

[0016] In order of correlation from large to small, the candidate users corresponding to the first N correlations are determined as the target candidate group. This not only improves the accuracy of the obtained target candidate group, but also effectively saves computing resources and improves computing efficiency because only the correlation between the target feature information corresponding to each candidate user in the candidate group and the target feature information corresponding to the M cluster center users is calculated.

[0017] In a possible implementation, determining the correlation degree corresponding to each candidate user in the candidate group according to the target feature information corresponding to each candidate user and the target feature information corresponding to the M cluster center users may include:

[0018] The correlation degree between the target feature information corresponding to any candidate user in the candidate group and the target feature information corresponding to each cluster center in the M cluster center users is calculated to obtain M correlation degrees.

[0019] The mean of the M correlation degrees is determined as the correlation degree corresponding to the candidate user, thereby calculating the correlation degree corresponding to the candidate user. In this way, when calculating the correlation degree, the target feature information corresponding to the M cluster center users after clustering is used for calculation. This can solve the problem of abnormal users that may be mixed in the seed population and the problem that the seed population itself contains multiple sub-class populations to a certain extent, so that the calculation is not affected by abnormal data, and the population characteristics are strengthened, so that the feature information of the target candidate population calculated at the end is more significant, and at the same time solves the problem of large resource consumption and long time consumption when calculating between super-large matrices.

[0020] In a possible implementation, obtaining a seed population may include:

[0021] Get the initial seed population.

[0022] The initial seed population is screened according to the target characteristic information corresponding to each seed user in the initial seed population to obtain the seed population, so that outliers in the initial seed population can be eliminated, thereby obtaining a seed population with characteristic information having higher similarity.

[0023] In a possible implementation, if the operation instruction includes the configuration parameter, the candidate group includes the users corresponding to the configuration parameter; if the operation instruction includes the business requirement parameter, the candidate group includes the users corresponding to the business requirement parameter; if the operation instruction includes the configuration parameter and the business requirement parameter, the candidate group includes the intersection of the users corresponding to the configuration parameter and the users corresponding to the business requirement parameter, thereby determining the candidate group. It can be seen that the configuration parameters and / or business requirement parameters input by the user are taken into account when determining the candidate group, that is, the business demands of different personalized crowd diffusion are taken into account, which can effectively increase the flexibility and personalization of crowd diffusion.

[0024] In a possible implementation, the information delivery method may further include:

[0025] The business characteristics corresponding to each preset user in the preset population are input into the prediction model to obtain the probability that the preset user meets the business needs.

[0026] The preset users whose probabilities are greater than the preset threshold are determined as users corresponding to the business demand parameters.

[0027] In a possible implementation, determining the target characteristic information according to the configuration parameter and / or the service requirement parameter may include:

[0028] Obtain multiple attribute feature information;

[0029] The characteristic information corresponding to the configuration parameters and / or the characteristic information corresponding to the business requirement parameters, as well as the multiple attribute characteristic information, are determined as the target characteristic information, so that when the correlation is subsequently calculated based on the target characteristic information, the final target candidate group can be determined from the candidate group. Since the target characteristic information includes the configuration parameters and / or business requirement parameters input by the user, that is, the business demands of different personalized population diffusion are taken into account, the flexibility and personalization of population diffusion can be effectively increased.

[0030] In a second aspect, an embodiment of the present application further provides an information delivery device, which may include:

[0031] The receiving module is used to receive an operation instruction input on the operation interface, wherein the operation instruction includes configuration parameters and / or business requirement parameters.

[0032] The processing module is used to determine a candidate group and target feature information according to the configuration parameters and / or business requirement parameters; the candidate group includes multiple candidate users.

[0033] The acquisition module is used to acquire a seed population; the seed population includes a plurality of seed users.

[0034] The processing module is further used to determine a target candidate group in the candidate group according to the target feature information corresponding to each candidate user in the candidate group and the seed group;

[0035] The sending module is used to deliver information to the target candidate group.

[0036] In a possible implementation, the processing module is specifically used to cluster the multiple seed users according to the target feature information corresponding to each seed user in the seed population to obtain M clusters; wherein the center point of each cluster serves as a cluster center user; M is an integer greater than or equal to 2; and according to the target feature information corresponding to each candidate user in the candidate population and the target feature information corresponding to the M cluster center users, the correlation degree corresponding to each candidate user is determined; and then, in descending order of correlation degree, the candidate users corresponding to the first N correlation degrees are determined as the target candidate population.

[0037] In a possible implementation, the processing module is specifically used to calculate the correlation between the target feature information corresponding to any candidate user in the candidate group and the target feature information corresponding to each cluster center of the M cluster center users, to obtain M correlation degrees; and determine the average of the M correlation degrees as the correlation degree corresponding to the candidate user.

[0038] In a possible implementation, the acquisition module is specifically used to acquire an initial seed population; and screen the initial seed population according to target feature information corresponding to each seed user in the initial seed population to obtain the seed population.

[0039] In one possible implementation, if the operation instruction includes the configuration parameter, the candidate group includes users corresponding to the configuration parameter; if the operation instruction includes the business requirement parameter, the candidate group includes users corresponding to the business requirement parameter; if the operation instruction includes the configuration parameter and the business requirement parameter, the candidate group includes the intersection of users corresponding to the configuration parameter and users corresponding to the business requirement parameter.

[0040] In a possible implementation, the processing module is also used to input the business characteristics corresponding to each preset user in the preset population into the prediction model to obtain the probability that the preset user meets the business needs; and determine the preset users whose probability is greater than a preset threshold as users corresponding to the business demand parameters.

[0041] In a possible implementation, the processing module is specifically used to obtain multiple attribute feature information; and determine the feature information corresponding to the configuration parameters and / or the feature information corresponding to the business requirement parameters, as well as the multiple attribute feature information, as the target feature information.

[0042] In a third aspect, an embodiment of the present application further provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the information delivery method as described in any possible implementation of the first aspect described above is implemented.

[0043] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, in which computer instructions are stored. When the computer instructions are executed on an electronic device, the electronic device executes the information delivery method as described in any possible implementation of the first aspect above.

[0044] The information delivery method, device and equipment provided by the embodiments of the present application, when delivering information, first receive the operation instructions including configuration parameters and / or business demand parameters input by the user on the operation interface, that is, take into account the business demands of different personalized groups of people to spread, and determine the candidate group and target feature information according to the configuration parameters and / or business demand parameters input by the user, and then determine the target candidate group in the candidate group according to the target feature information and seed population corresponding to each candidate user in the candidate group, effectively increase the flexibility and humanization of population diffusion, and then deliver information to the target candidate group. Compared with the prior art of directly delivering information to the diffusion population of the seed population, the accuracy of information delivery is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0046] Figure 1 A flowchart of an information delivery method provided in an embodiment of the present application;

[0047] Figure 2 A schematic diagram of determining the probability of a preset user achieving a high potential order demand target provided by an embodiment of the present application;

[0048] Figure 3 A schematic diagram of a framework for determining a target candidate group provided in an embodiment of the present application;

[0049] Figure 4 A schematic diagram of a process for determining a target candidate group from a candidate group provided in an embodiment of the present application;

[0050] Figure 5 A schematic diagram of a typical pattern extraction provided in an embodiment of the present application;

[0051] Figure 6 A schematic diagram of the structure of an information delivery device provided in an embodiment of the present application;

[0052] Figure 7 A schematic diagram of the structure of an electronic device embodiment provided in an embodiment of the present application.

[0053] The above drawings show clear embodiments of the present disclosure, which will be described in more detail below. These drawings and text descriptions are not intended to limit the scope of the present disclosure in any way, but to illustrate the concepts of the present disclosure to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0054] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0055] In the embodiments of the present application, "at least one" means one or more, and "more" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. In the text description of the present application, the character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0056] The information delivery method provided in the embodiment of the present application can be applied to the scenario of information delivery or advertising marketing, and of course, it can also be applied to e-commerce scenarios or other scenarios. Taking the information delivery scenario as an example, usually, when the electronic device delivers information, it obtains the diffusion population based on the local seed population and delivers information to the diffusion population. However, when the seed population is large, the number of diffusion populations will be large; and, since there may be redundant or low-credibility non-high-quality seed users in the seed population, it will have a certain impact on the determination of the diffusion population, resulting in low accuracy of information delivery.

[0057] In order to improve the accuracy of information delivery, the embodiment of the present application provides an information delivery method. When delivering information, the operation instructions including configuration parameters and / or business demand parameters input by the user on the operation interface are first received, that is, the business demands of different personalized groups are taken into consideration, and the candidate group and target feature information are determined according to the configuration parameters and / or business demand parameters input by the user, and then the target candidate group is determined in the candidate group according to the target feature information and seed group corresponding to each candidate user in the candidate group, which effectively increases the flexibility and humanization of population diffusion, and then information is delivered to the target candidate group. Compared with the prior art of directly delivering information to the diffusion population of the seed population, the accuracy of information delivery is improved.

[0058] Taking the application of the embodiment of the present application to an e-commerce scenario as an example, the information delivered may be marketing advertisements, coupons, etc. For example, the configuration parameters may include at least one of categories, stores, values, and intentions; the business demand parameters may include at least one of high conversion potential demand, high attention potential demand, and high order potential demand. In the embodiment of the present application, when the crowd is diffused, it is precisely because the configuration parameters and / or business demand parameters input by the user are taken into account, that is, the business demands of different personalized crowd diffusion are taken into account, so that when determining the target candidate group, the flexibility and personalization of the crowd diffusion are effectively increased, thereby improving the accuracy of the acquired diffusion crowd; in this way, when delivering information to the diffusion crowd, the accuracy of information delivery is improved.

[0059] It is understandable that when the embodiments of the present application are applied to other scenarios, the corresponding configuration parameters and demand parameters can also be added or replaced with other parameters. As long as the indicators are in line with business logic and supported by data, they can be replaced during the implementation process. However, the implementation principle is similar to the implementation process in the e-commerce scenario. Here, the embodiments of the present application will not be repeated.

[0060] The following will take the e-commerce scenario as an example to explain in detail the technical solution of the present application and how the technical solution of the present application solves the above technical problems with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0061] Figure 1 This is a flow chart of an information delivery method provided by an embodiment of the present application. The information delivery method can be executed by software and / or hardware devices. For example, the hardware device can be an information delivery device, and the information delivery device can be set in an electronic device. For example, see Figure 1 As shown, the information delivery method may include:

[0062] S101, receiving an operation instruction input on an operation interface.

[0063] The operation instruction includes configuration parameters and / or business requirement parameters.

[0064] For example, before receiving the operation instructions input by the user on the operation interface, the configuration parameters and business requirement parameter list can be displayed to the user through the operation interface. The configuration parameter list can include multiple configuration parameters, such as category, store, value, and intention, etc., and the business requirement parameter list can include multiple business requirement parameters, such as high conversion potential demand, high attention potential demand, and high order potential demand, etc.; the user can select the configuration parameters and / or business requirement parameters to be used according to the configuration parameter list and business requirement parameter list displayed on the operation interface, and the configuration parameters and / or business requirement parameters can reflect the business demands of this crowd diffusion.

[0065] It is understandable that when the user selects the parameters to be used in the configuration parameter list and the business requirement parameter list, he can select only one or more parameters in the configuration parameter list, such as category; or he can select only one or more parameters in the business requirement parameter list, such as high potential demand for orders; of course, he can also select one or more parameters in the configuration parameter list and the business requirement parameter list at the same time, such as category and high potential demand for orders. The specific selection can be made according to actual needs, and this is not further limited in the embodiments of the present application.

[0066] After receiving the operation instruction including the configuration parameters and / or the business requirement parameters input by the user on the operation interface, the following S102 may be executed:

[0067] S102: Determine candidate groups and target feature information according to configuration parameters and / or business requirement parameters.

[0068] The candidate group includes multiple candidate users.

[0069] For example, when determining the candidate group according to the configuration parameters and / or business requirement parameters, if the operation instruction includes the configuration parameter, the candidate group includes the users corresponding to the configuration parameter; taking the configuration parameter as category as an example, the candidate group is the users corresponding to the category; if the operation instruction includes the business requirement parameter, the candidate group includes the users corresponding to the business requirement parameter; taking the business requirement parameter as high potential order demand as an example, the candidate group is the users who meet the high potential order demand; if the operation instruction includes the configuration parameter and the business requirement parameter, the candidate group includes the intersection of the users corresponding to the configuration parameter and the users corresponding to the business requirement parameter, taking the configuration parameter as category and the business requirement parameter as high potential order demand as an example, the candidate group is the intersection users of the users corresponding to the category and the users who meet the high potential order demand, that is, the intersection users are both the users corresponding to the category and the users who meet the high potential order demand, so that the candidate group can be determined according to the configuration parameters and / or business requirement parameters.

[0070] In the above description, when determining the user corresponding to the business demand parameter, for example, the business characteristics corresponding to each preset user in the preset population can be input into the prediction model to obtain the probability that the preset user meets the business demand; for example, the probability of each preset user achieving the business demand target in the next five days is predicted, and the preset user with a probability greater than the preset threshold is determined as the user corresponding to the business demand parameter, so that the diffusion population can be portrayed more accurately, and the business application is more directional, thereby further saving marketing costs and improving the rate of return. Among them, the preset population can be understood as a population composed of users with business characteristic behaviors in the near future. It is understandable that before the prediction model predicts the probability that the preset user meets the business demand, it is necessary to first perform model training to obtain the prediction model. For example, a supervised classification algorithm can be used; for example, a logistic regression algorithm, a random forest algorithm, GBDT, XGboost, etc. are used for model training to obtain the prediction model, and a comprehensive evaluation method can also be used to combine models with artificial experience or use a deep learning algorithm to train the prediction model. It can be set according to actual needs. Here, the embodiment of the present application does not make specific restrictions on which algorithm is used to train the prediction model.

[0071] For example, when the business demand parameter is high order potential demand, the prediction model can be a high order potential model based on user expectations. In this case, please refer to Figure 2 As shown, Figure 2 A schematic diagram of determining the probability of a preset user achieving a high order potential demand target provided in an embodiment of the present application can first input the relevant business characteristics corresponding to each preset user, such as basic attribute characteristics, search characteristics, browsing characteristics, click characteristics, add-to-cart characteristics, purchase characteristics, service characteristics, category behavior characteristics, etc. into a pre-established high order potential model to predict the probability of each preset user achieving the high order potential demand target, that is, to screen out users with a high probability of meeting the order demand as users corresponding to the high order potential demand.

[0072] After the candidate group is determined according to the configuration parameters and / or business requirement parameters, since the candidate users in the candidate group need to be further screened based on the target characteristics of each candidate user in the candidate group and the seed population to obtain the final target candidate group, the target characteristic information can also be determined according to the configuration parameters and / or business requirement parameters. For example, when determining the target characteristic information according to the configuration parameters and / or business requirement parameters, multiple attribute characteristic information can be obtained first; then the characteristic information corresponding to the configuration parameters and / or the characteristic information corresponding to the business requirement parameters are determined, and the characteristic information corresponding to the configuration parameters and / or the characteristic information corresponding to the business requirement parameters, as well as multiple attribute characteristic information, are determined as the target characteristic information.

[0073] For example, when obtaining multiple attribute feature information, you can use a combination of variance analysis, correlation analysis, missing rate, manual judgment and other methods to select multiple attribute feature information from a large number of attribute feature information including basic feature information (gender, age, education, marital status, etc.), behavioral attribute feature information (search, browse, click, add to cart, order, etc.), value attribute feature information (life cycle, user value, user purchasing power, etc.), and credit attributes (order cancellation, after-sales, risk level, etc.), so as to realize the extraction of typical attribute feature information from a large number of unsupervised feature information sets. It can be understood that in the embodiments of the present application, the purpose of extracting typical attribute feature information is: 1. When clustering or diffusion is performed through massive attribute feature information in the later stage, the stability of the diffusion of target candidate users is poor, and the interpretability of the target candidate population is weak; 2. Attribute feature information with a high missing rate and columns with very small feature changes contain little information and do not need to be retained during diffusion; 3. When homogeneous attribute feature information is used to realize population diffusion, only important variables need to be retained. If too many variables of the same category are retained, on the one hand, the calculation efficiency will be affected. On the other hand, information redundancy will not have much corrective effect on the results of population diffusion, but will hinder each other, so that the subsequent calculation of correlation, such as similarity or distance, is too greatly affected by a certain type of attribute feature information, resulting in a low accuracy of the calculated correlation.

[0074] It should be noted that the multiple attribute feature information obtained through feature screening is only attribute feature information. In order to make the crowd diffusion more in line with personalized business needs, it is also possible to obtain behavioral feature information based on configuration parameters and / or business demand parameters, and use the behavioral feature information corresponding to the configuration parameters and / or business demand parameters and the multiple attribute feature information obtained as target feature information to determine the target feature information, so that the final target candidate group can be determined from the candidate group when the correlation degree is calculated based on the target feature information. For example, taking the configuration parameter as category as an example, the feature information corresponding to the configuration parameter is the feature information corresponding to the category; taking the business demand parameter as high potential demand for orders, the feature information corresponding to the business demand parameter is the probability feature of achieving the high potential demand for orders target.

[0075] Taking the configuration parameter of cross-category as an example, when a user selects a certain category for cross-category, the user's recent behavior (click, follow, add to cart, purchase) data can be sorted by time to form a user behavior sequence, and a Word2Vec model can be constructed for embedding to calculate the similarity between categories, build a graph relationship between categories, establish a category association suitable for cross-category new users, and recommend related categories to users. In this case, the characteristics of the candidate population and the prediction model will be dynamically debugged. Candidate population: candidate users who have purchased related categories of the category in the past year or have searched, browsed, clicked, added to cart, etc. in the past three months, and the prediction model predicts that the user is marked as yes, so as to obtain a cross-category candidate population. Among them, the characteristics of the prediction model: mainly include user attribute feature information (user basic attribute feature information, user value feature information, life cycle feature information, consumption preference feature information, and full-site behavior feature information), as well as category behavior feature information, category value feature information, and category preference feature information of the category and related categories.

[0076] After determining the candidate group and target feature information according to the configuration parameters and / or business requirement parameters respectively through S102, the seed population can be obtained to determine the target candidate group in the candidate group according to the target feature information and seed population corresponding to each candidate user in the candidate group, that is, execute the following S103-S104.

[0077] S103, obtaining a seed population.

[0078] Among them, the seed population includes multiple seed users.

[0079] For example, when obtaining a seed population, the initial seed population maintained therein can be obtained locally; since the initial seed population is often mixed with many sub-category populations, if the initial seed population is directly used for subsequent correlation calculations, the characteristics of the seed population will be weakened, making the characteristics of the target candidate population ultimately determined based on the initial seed population unclear; therefore, when calculating subsequent correlation calculations, the initial seed population can be first screened according to the target feature information corresponding to each seed user in the initial seed population to obtain a seed population with feature information having a higher degree of similarity.

[0080] For example, when the initial seed population is screened according to the target feature information corresponding to each seed user in the initial seed population, the box plot method can be used to process the outliers of the initial seed users in the initial seed population. The outlier is defined as a value less than Q1 (Q1 refers to the first quartile) -1.5IQR (IQR refers to the difference between the third quartile and the first quartile), or the outlier is defined as a value greater than Q3 (Q3 refers to the third quartile) +1.5IQR, wherein a value greater than Q3+1.5IQR is replaced by Q3+1.5IQR, and a value less than Q1-1.5IQR is replaced by Q1-1.5IQR.

[0081] S104: Determine a target candidate group in the candidate group according to the target feature information and the seed population corresponding to each candidate user in the candidate group.

[0082] Among them, the target candidate group can be understood as the group of people who are ultimately needed.

[0083] When determining the target candidate group from the candidate group based on the target feature information and the seed population corresponding to each candidate user in the candidate group, for example, see Figure 3 As shown, Figure 3 A framework diagram for determining a target candidate population is provided in an embodiment of the present application. It can be seen that a seed population can be obtained first. When a population is diffused based on the seed population, a diffusion target can be determined first. The diffusion target can be achieved through configuration parameters and / or business demand parameters. The diffused population in the diffusion direction corresponding to the diffusion target is the target candidate population.

[0084] S105. Deliver information to target candidate groups.

[0085] It can be seen that the information delivery method of the embodiment of the present application, when delivering information, first receives the operation instructions including configuration parameters and / or business demand parameters input by the user on the operation interface, that is, takes into account the business demands of different personalized groups of people to spread, and determines the candidate group and target feature information according to the configuration parameters and / or business demand parameters input by the user, and then determines the target candidate group in the candidate group according to the target feature information and seed population corresponding to each candidate user in the candidate group, effectively increasing the flexibility and humanization of population diffusion, and then delivering information to the target candidate group. Compared with the prior art of directly delivering information to the diffusion population of the seed population, the accuracy of information delivery is improved.

[0086] Based on the above Figure 1 The illustrated embodiment is described in detail through the following embodiment to facilitate understanding of how to determine the target candidate group in the candidate group according to the target feature information and the seed population corresponding to each candidate user in the candidate group in the above S104.

[0087] Figure 4 A flow chart of determining a target candidate group from a candidate group provided in an embodiment of the present application is provided. For example, see Figure 4 As shown, the method may include:

[0088] S401 , clustering multiple seed users according to target feature information corresponding to each seed user in the seed population to obtain M clusters.

[0089] The center point of each cluster is regarded as a cluster center user; M is an integer greater than or equal to 2.

[0090] For example, when clustering multiple seed users, the K-means algorithm can be used to cluster the seed population. Of course, other algorithms can also be used to cluster the seed population, and each cluster center point is extracted as a typical pattern of the seed user as a virtual cluster center user, which not only saves computing resources but also makes the characteristics of the seed population more obvious. For an example, see Figure 5 As shown, Figure 5 A schematic diagram of a typical pattern extraction provided for an embodiment of the present application shows that multiple attribute feature information can be obtained through feature screening, and the behavioral feature information corresponding to the configuration parameters and / or business requirement parameters can be added as target feature information, and then combined with the seed population. In this way, multiple seed users can be clustered according to the target feature information corresponding to each seed user in the seed population, and M clusters can be obtained, and each cluster center point can be used as a typical pattern of the seed user.

[0091] S402: Determine the correlation degree corresponding to each candidate user according to the target feature information corresponding to each candidate user in the candidate group and the target feature information corresponding to the M cluster center users.

[0092] For example, when determining the correlation degree corresponding to each candidate user based on the target feature information corresponding to each candidate user in the candidate group and the target feature information corresponding to M cluster center users, since the calculation method of the correlation degree corresponding to each candidate user is similar, in order to avoid redundancy, any candidate user in the candidate group can be taken as an example. When calculating the correlation degree corresponding to the candidate user, the correlation degree between the target feature information corresponding to the candidate user and the target feature information corresponding to each cluster center of the M cluster center users can be first calculated to obtain M correlation degrees; then the average of the M correlation degrees is calculated, and the average of the M correlation degrees is determined as the correlation degree corresponding to the candidate user; in this way, the correlation degree corresponding to each candidate user in the candidate group can be obtained.

[0093] For example, the degree of association can be similarity and / or distance. When calculating the similarity between the target feature information corresponding to the candidate user and the target feature information corresponding to each cluster center of the M cluster center users, different algorithms are usually selected according to different data. The similarity algorithms that are currently more commonly used are: Pearson correlation coefficient algorithm, cosine similarity algorithm, and corrected cosine similarity algorithm. When calculating the distance between the target feature information corresponding to the candidate user and the target feature information corresponding to each cluster center of the M cluster center users, different algorithms are also selected according to different data. The distance algorithms that are currently more commonly used are: Euclidean distance, Minkowski distance, Manhattan distance, Chebyshev distance, and Mahalanobis distance.

[0094] It should be noted that when calculating the correlation degree, the target feature information corresponding to the users in the M cluster centers after clustering is used for calculation. This can, to a certain extent, solve the problem of abnormal users that may be mixed in the seed population and the problem that the seed population itself contains multiple sub-populations, so that the calculation is not affected by abnormal data, and the population characteristics are strengthened, so that the feature information of the target candidate population calculated in the end is more significant, and at the same time solves the problem of large resource consumption and long time consumption when calculating between very large matrices.

[0095] In this way, after calculating the correlation degree corresponding to each candidate user according to the target feature information corresponding to each candidate user in the candidate group and the target feature information corresponding to the M cluster center users, the following S403 can be executed:

[0096] S403 , determining the target candidate group from the candidate users corresponding to the first N relevances in descending order of relevance.

[0097] After calculating the correlation degree corresponding to each candidate user, the correlation degrees can be sorted in descending order, and the candidate users corresponding to the first N larger correlation degrees in the candidate group are determined as the target candidate group; of course, it is also possible not to sort in descending order, but directly determine the candidate users corresponding to the correlation degrees greater than the preset threshold as the target candidate group, which can be set according to actual needs, and the embodiment of the present application does not make specific restrictions here. Among them, the preset threshold can be set according to actual needs.

[0098] It can be seen that when determining the target candidate group in the candidate group according to the target feature information corresponding to each candidate user in the candidate group and the seed population, the multiple seed users in the seed population are first clustered to obtain M clusters, each cluster corresponding to a cluster center user; and the correlation degree corresponding to each candidate user is determined according to the target feature information corresponding to each candidate user in the candidate group and the target feature information corresponding to the M cluster center users, and the candidate users corresponding to the first N correlation degrees are determined as the target candidate group in order of correlation degree from large to small, which not only improves the accuracy of the obtained target candidate group, but also effectively saves computing resources and improves computing efficiency because only the correlation degree between the target feature information corresponding to each candidate user in the candidate group and the target feature information corresponding to the M cluster center users is calculated.

[0099] Figure 6 This is a structural diagram of an information delivery device 60 provided in an embodiment of the present application. For example, see Figure 6 As shown, the information delivery device 60 may include:

[0100] The receiving module 601 is used to receive an operation instruction input on an operation interface, where the operation instruction includes configuration parameters and / or business requirement parameters.

[0101] The processing module 602 is used to determine a candidate group and target characteristic information according to configuration parameters and / or business requirement parameters; the candidate group includes multiple candidate users.

[0102] The acquisition module 603 is used to acquire a seed population; the seed population includes a plurality of seed users.

[0103] The processing module 602 is further used to determine a target candidate group in the candidate group according to the target feature information corresponding to each candidate user in the candidate group and the seed population.

[0104] The sending module 604 is used to deliver information to the target candidate group.

[0105] Optionally, processing module 602 is specifically used to cluster multiple seed users according to target feature information corresponding to each seed user in the seed population to obtain M clusters; wherein the center point of each cluster is used as a cluster center user; M is an integer greater than or equal to 2; and determine the correlation degree corresponding to each candidate user according to the target feature information corresponding to each candidate user in the candidate population and the target feature information corresponding to the M cluster center users; and then determine the target candidate population according to the order of correlation from large to small for the candidate users corresponding to the first N correlation degrees.

[0106] Optionally, processing module 602 is specifically used to calculate the correlation between the target feature information corresponding to any candidate user in the candidate group and the target feature information corresponding to each cluster center of the M cluster center users, to obtain M correlations; and determine the average of the M correlations as the correlation corresponding to the candidate user.

[0107] Optionally, the acquisition module 603 is specifically used to acquire an initial seed population; and screen the initial seed population according to target feature information corresponding to each seed user in the initial seed population to obtain the seed population.

[0108] Optionally, if the operation instruction includes configuration parameters, the candidate group includes users corresponding to the configuration parameters; if the operation instruction includes business requirement parameters, the candidate group includes users corresponding to the business requirement parameters; if the operation instruction includes configuration parameters and business requirement parameters, the candidate group includes the intersection of users corresponding to the configuration parameters and users corresponding to the business requirement parameters.

[0109] Optionally, processing module 602 is also used to input the business characteristics corresponding to each preset user in the preset population into the prediction model to obtain the probability that the preset user meets the business needs; and determine the preset users whose probabilities are greater than a preset threshold as users corresponding to the business demand parameters.

[0110] Optionally, the processing module 602 is specifically used to obtain multiple attribute feature information; and determine the feature information corresponding to the configuration parameter and / or the feature information corresponding to the business requirement parameter, as well as the multiple attribute feature information, as the target feature information.

[0111] The information delivery device 60 provided in the embodiment of the present application can execute the technical solution of the information delivery method in any of the above-mentioned embodiments. Its implementation principle and beneficial effects are similar to the implementation principle and beneficial effects of the information delivery method. Please refer to the implementation principle and beneficial effects of the information delivery method, and no further details will be given here.

[0112] It should be noted that it should be understood that the division of the various modules of the above devices is only a division of logical functions. In actual implementation, they can be fully or partially integrated into one physical entity, or they can be physically separated. And these modules can all be implemented in the form of software called by processing elements; they can also be all implemented in the form of hardware; some modules can also be implemented in the form of software called by processing elements, and some modules can be implemented in the form of hardware. For example, the determination module can be a separately established processing element, or it can be integrated in a chip of the above-mentioned device. In addition, it can also be stored in the memory of the above-mentioned device in the form of program code, and called and executed by a processing element of the above-mentioned device. The implementation of other modules is similar. In addition, these modules can be fully or partially integrated together, or they can be implemented independently. The processing element described here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each module above can be completed by an integrated logic circuit of hardware in the processor element or instructions in the form of software.

[0113] For example, the above modules may be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASIC), or one or more microprocessors (digital signal processors, DSP), or one or more field programmable gate arrays (FPGA), etc. For another example, when a module above is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0114] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid state disk (SSD)), etc.

[0115] Figure 7 The following is a schematic diagram of the structure of an electronic device embodiment provided in the present application. Figure 7 As shown, the electronic device may include: a processor 701, a memory 702, and a computer program stored in the memory 702 and executable on the processor 701; of course, it may also include a communication interface 703 and a system bus 704; the memory 702 and the communication interface 703 are connected to the processor 701 via the system bus 704 and communicate with each other; the memory 702 is used to store the computer program; the communication interface 703 is used to communicate with other devices; when the processor 701 executes the computer program, it implements the technical solution of the information delivery method shown in the above embodiment; its implementation principle and beneficial effects are similar to those of the information delivery method; reference may be made to the implementation principle and beneficial effects of the information delivery method, which will not be repeated here.

[0116] In this Figure 7 In the embodiment, the processor 701 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic devices, or discrete hardware components.

[0117] The memory 702 may include a random access memory (RAM), may include a read-only memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0118] The communication interface 703 is used to implement the communication between the database access device and other devices (such as the client, the read-write library and the read-only library).

[0119] The system bus 704 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The system bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0120] Optionally, an embodiment of the present application further provides a computer-readable storage medium, in which computer instructions are stored. When the computer instructions are executed on an electronic device, the electronic device executes the technical solution of the information delivery method shown in the above embodiment. The implementation principle and beneficial effects are similar to those of the information delivery method. Please refer to the implementation principle and beneficial effects of the information delivery method, and no further details will be given here.

[0121] Optionally, an embodiment of the present application also provides a chip for running instructions, which is used to execute the technical solution of the information delivery method shown in the above embodiment. Its implementation principle and beneficial effects are similar to the implementation principle and beneficial effects of the information delivery method. Please refer to the implementation principle and beneficial effects of the information delivery method, and no further details will be given here.

[0122] An embodiment of the present application also provides a program product, which includes a computer program, which is stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium. When the at least one processor executes the computer program, the technical solution of the information delivery method shown in the above embodiment can be implemented. Its implementation principle and beneficial effects are similar to those of the information delivery method. Please refer to the implementation principle and beneficial effects of the information delivery method, and no further details will be given here.

[0123] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the application disclosed herein. This application is intended to cover any variations, uses or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0124] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for delivering information, characterized in that: include: Displaying a list of configuration parameters and business requirement parameters to a user through an operation interface, and receiving an operation instruction input on the operation interface, wherein the operation instruction includes the configuration parameters and / or business requirement parameters; Determine candidate groups and target feature information according to the configuration parameters and / or business requirement parameters; The candidate group includes multiple candidate users; Obtaining a seed population; the seed population includes a plurality of seed users; Determining a target candidate group from the candidate group according to target feature information corresponding to each candidate user in the candidate group and the seed group; delivering information to the target candidate group; Determining target characteristic information according to the configuration parameters and / or business requirement parameters includes: Obtain multiple attribute feature information; The characteristic information corresponding to the configuration parameters and / or the characteristic information corresponding to the business requirement parameters, as well as the plurality of attribute characteristic information, are determined as the target characteristic information.

2. The method according to claim 1, characterized in that The step of determining a target candidate group from the candidate group according to the target feature information corresponding to each candidate user in the candidate group and the seed group includes: Clustering the plurality of seed users according to the target feature information corresponding to each seed user in the seed population to obtain M clusters; wherein the center point of each cluster is used as a cluster center user; and M is an integer greater than or equal to 2; Determine the correlation degree corresponding to each candidate user according to the target feature information corresponding to each candidate user in the candidate group and the target feature information corresponding to the M cluster center users; According to the order of relevance from large to small, the candidate users corresponding to the first N relevances are determined as the target candidate group.

3. The method according to claim 2, characterized in that The determining the correlation degree corresponding to each candidate user according to the target feature information corresponding to each candidate user in the candidate group and the target feature information corresponding to the M cluster center users includes: Calculate the correlation between the target feature information corresponding to any candidate user in the candidate group and the target feature information corresponding to each cluster center in the M cluster center users, and obtain M correlation degrees; The average value of the M association degrees is determined as the association degree corresponding to the candidate user.

4. The method according to claim 1, characterized in that The seed obtaining population includes: Get the initial seed population; The initial seed population is screened according to target feature information corresponding to each seed user in the initial seed population to obtain the seed population.

5. The method according to any one of claims 1 to 4, characterized in that: If the operation instruction includes the configuration parameter, the candidate group includes the users corresponding to the configuration parameter; if the operation instruction includes the business requirement parameter, the candidate group includes the users corresponding to the business requirement parameter; if the operation instruction includes the configuration parameter and the business requirement parameter, the candidate group includes the intersection of the users corresponding to the configuration parameter and the users corresponding to the business requirement parameter.

6. The method according to claim 5, characterized in that The method further comprises: Inputting the service characteristics corresponding to each preset user in the preset population into the prediction model to obtain the probability that the preset user meets the service requirements; The preset users whose probabilities are greater than the preset threshold are determined as users corresponding to the business demand parameters.

7. An information delivery device, characterized in that: include: A receiving module, which displays a list of configuration parameters and business requirement parameters to a user through an operation interface, and is used to receive an operation instruction input on the operation interface, wherein the operation instruction includes the configuration parameters and / or business requirement parameters; A processing module, used to determine candidate groups and target feature information according to the configuration parameters and / or business requirement parameters; The candidate group includes multiple candidate users; An acquisition module, used for acquiring a seed population; the seed population includes a plurality of seed users; The processing module is further used to determine a target candidate group in the candidate group according to the target feature information corresponding to each candidate user in the candidate group and the seed group; A sending module, used for delivering information to the target candidate group; The processing module is specifically used to obtain multiple attribute feature information; determine the feature information corresponding to the configuration parameters and / or the feature information corresponding to the business requirement parameters, as well as the multiple attribute feature information, as the target feature information.

8. An electronic device comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the information delivery method as described in any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and when the computer instructions are executed on an electronic device, the electronic device executes the information delivery method as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the information delivery method described in any one of claims 1 to 6 is implemented.

Images