CN107613022B - Content pushing method and device and computer equipment - Google Patents
Content pushing method and device and computer equipment Download PDFInfo
- Publication number
- CN107613022B CN107613022B CN201710985244.3A CN201710985244A CN107613022B CN 107613022 B CN107613022 B CN 107613022B CN 201710985244 A CN201710985244 A CN 201710985244A CN 107613022 B CN107613022 B CN 107613022B
- Authority
- CN
- China
- Prior art keywords
- user
- application program
- model
- click rate
- application
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
- H04L65/40—Support for services or applications
Abstract
The invention provides a content pushing method and device and computer equipment. The content pushing method comprises the steps of obtaining characteristics of a user who exposes an application program and clicking behaviors of the application program; generating a sample according to the characteristics of the user and the clicking behavior of the application program; inputting the samples into a clustering model, and determining a population to which the samples belong in the clustering model; inputting the sample into a click rate estimation model corresponding to the group to obtain the estimated click rate of the user to the application program; and pushing content to the user according to the estimated click rate of the application program. After the content pushing method determines the group to which the user belongs, the estimated click rate is obtained according to the click rate estimation model of the group, so that the accuracy of the estimated click rate is improved, and the content is accurately pushed to the user.
Description
Technical Field
The invention relates to the technical field of internet, in particular to a content pushing method, a content pushing device and computer equipment.
Background
The LR (Logistic Regression) model is one of the discrete choice models, and the Logistic Regression model is the earliest discrete choice model and is also the most widely used model at present. The logistic regression model is also a classification model in machine learning, and is very widely applied in practice due to the simplicity and high efficiency of the algorithm.
At present, in an application store, a model most used for estimating the click rate of an application program (APP) is an LR model, and the idea is to cross user characteristics and application characteristics approximately to be used as input characteristics of a final model and train to obtain the LR model. The method has a great disadvantage that the method is a model for all users, in many cases, the optimal LR model of the user group A is X, the optimal LR model of the user group B is Y, and X and Y are likely to conflict. I.e. on a certain feature the X-model is a positive feature and the Y-model is a negative feature. While the case of only one LR model is to generate a compromise model Z for catering both the a and B user groups, model Z is not the optimal model. Therefore, the conventional LR model is not accurate in estimating the click rate of the user to the application program, so that the user cannot be accurately pushed related content according to the conventional LR model.
Disclosure of Invention
The invention aims to provide a content pushing method, a content pushing device and computer equipment, which are used for obtaining an estimated click rate according to a click rate estimation model of a group after the group to which a user belongs is determined, so that the accuracy of the estimated click rate is improved, and related content is accurately pushed to the user.
In order to achieve the purpose, the invention provides the following technical scheme:
a content push method, comprising the steps of: acquiring the characteristics of a user who exposes the application program and the clicking behavior of the application program; generating a sample according to the characteristics of the user and the clicking behavior of the application program; inputting the samples into a clustering model, and determining a population to which the samples belong in the clustering model; inputting the sample into a click rate estimation model corresponding to the group to obtain the estimated click rate of the user to the application program; and pushing content to the user according to the estimated click rate of the application program.
In one embodiment, the pushing content to the user according to the estimated click rate of the application includes: and pushing the application program to the user according to the estimated click rate of the application program.
In one embodiment, the pushing the application to the user according to the estimated click rate of the application includes: and confirming that the estimated click rate of the application program is greater than a threshold value, and pushing the application program to the user.
In one embodiment, the pushing the application program to the user according to the estimated click rate of the application program includes calculating the estimated click rate of the user to all application programs in an application marketplace, sorting the application programs according to the estimated click rate from high to low, and pushing the application programs in the top preset number to the user.
In one embodiment, before obtaining the characteristics of the user who exposes the application and the click behavior of the application, the method further includes: acquiring the characteristics of historical users who expose the application program and the clicking behavior of the application program; generating a sample according to the characteristics of the historical user and the clicking behavior of the application program; and training a grouping model comprising a plurality of groups and a click rate estimation model corresponding to the groups by using the sample.
In one embodiment, the training of the clustering model including populations using the samples comprises: training a clustering model using the samples according to a decision tree algorithm.
In one embodiment, the training of the clustering model including a plurality of populations and the click rate estimation models corresponding to the populations by using the samples includes: and training a grouping model containing a plurality of groups by using the samples, classifying the samples according to the groups, and training a click rate estimation model of the group by using the samples classified correspondingly to the groups.
In one embodiment, the training of the click rate prediction model of the group using the samples of the corresponding classification of the group includes: and training the click rate estimation model by using the samples of the corresponding group classification and using a logistic algorithm.
In one embodiment, the click rate prediction model is a click rate prediction model obtained according to a logistic regression formula; the formula of the logistic regression is as follows:
where X represents the input variable and represents the weight vector of the different features.
In one embodiment, the training the click rate prediction model using the samples classified according to the population using a logistic algorithm includes: and training the group by using a logistic algorithm according to a gradient descent method, and training a click rate estimation model in the corresponding group.
In one embodiment, the characteristics of the user include natural attributes, social attributes, or preference attributes of the user.
In one embodiment, the generating a sample according to the characteristics of the user and the click behavior of the application includes: and taking the characteristics of the user as input variables and the clicking behavior as target variables to generate a sample.
A content pushing apparatus comprising: the first acquisition module is used for acquiring the characteristics of a user who exposes the application program and the clicking behavior of the application program; the generating module is used for generating a sample according to the characteristics of the user and the clicking behavior of the application program; the determining module is used for inputting the samples into a clustering model and determining a population to which the samples belong in the clustering model; the second acquisition module is used for inputting the sample into a click rate estimation model corresponding to the group to acquire the estimated click rate of the user on the application program; and the pushing module is used for pushing the content to the user according to the estimated click rate of the application program.
A computer apparatus, comprising: one or more processors; a memory; one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to perform the content push method of any of the embodiments described above.
Compared with the prior art, the scheme of the invention has the following advantages:
according to the content pushing method provided by the invention, the sample is generated according to the characteristics of the user and the clicking behavior of the application program, and the sample is input into the clustering model to confirm that the sample belongs to the cluster in the clustering model. And then, according to the click rate estimation model in the group, acquiring the estimated click rate of the user to the application program, and then pushing corresponding content to the user according to the estimated click rate of the application program. After the content pushing method determines the group to which the user belongs, the estimated click rate is obtained according to the click rate estimation model of the group, so that the accuracy of the estimated click rate is improved, and related content is accurately pushed to the user. For example, the related content may include applications, similar applications, application presentations, and advertisements and information related to the applications. In addition, the content pushing method carries out grouping model training according to different user groups, and trains click rate estimation models for different groups in the grouping models so that different user groups correspond to different click rate estimation models, and therefore more accurate estimated click rate of the user to the application program is obtained according to the corresponding click rate estimation models, and related content is accurately pushed to the user.
Drawings
The foregoing and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a flowchart of a method for pushing content according to an embodiment of the present invention;
FIG. 2 is a flow chart of a method for pushing content according to another embodiment of the present invention;
FIG. 3 is a flowchart of a method for building a clustering logistic regression click rate prediction model according to an embodiment of the present invention;
FIG. 4 is a flowchart of a usage scenario of a model of a content push method according to an embodiment of the present invention;
FIG. 5 is a flow chart of a method for pushing content according to another embodiment of the present invention;
FIG. 6 is a flowchart of a method for pushing content according to still another embodiment of the present invention;
fig. 7 is a schematic structural diagram of a content pushing apparatus according to an embodiment of the present invention;
fig. 8 is a schematic structural diagram of a computer device according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are illustrative only and should not be construed as limiting the invention.
Those skilled in the art should understand that the content in the content push method of the present invention includes: applications, similar applications, application introductions, advertisements and information related to applications, and the like.
Fig. 1 is a flowchart of a content push method according to an embodiment of the present invention. As shown in fig. 1, a content push method of the present invention includes the steps of:
and S10, acquiring the characteristics of the user who exposes the application program and the clicking behavior of the application program.
When detecting that the user exposes the application program, the server acquires the characteristics of the user who exposes the application program and acquires the clicking behavior of the user on the application program. Here, the clicking behavior of the application by the user includes that the application is clicked by the user, and that the application is not clicked by the user. The application program can be various application programs in the application store, including game application programs, social application programs, reading application programs, system application programs, shopping application programs and the like.
The characteristics of the user include natural attributes of the user, social attributes of the user, or preference attributes of the user. The natural attributes of the user such as the age of the user, the gender of the user, etc. The social attributes of the user include the cultural level of the user, the occupation of the user, the region of the user, and the like. The preference attributes of the user are 'military fan', 'science and technology enthusiasts', 'football fan' and 'game fan', etc.
In one embodiment, the server obtains characteristics of a user of the exposure application and builds a mathematical model characterizing the user based on the characteristics of the user. The user is actually characterized by the characteristics of the user. The method for acquiring the user characteristics comprises the following methods:
the method comprises the following steps: the mathematical model of the user is characterized by the natural attributes of the user. The mathematical model of the user is characterized, for example, by the age of the user and the gender of the user. Wherein discretization is performed on the continuous quantities. Such as age dispersion into "children", "teenagers", "young", "middle-aged" and "old", etc.
The method 2 comprises the following steps: and describing the mathematical model of the user through the social attributes of the user. The mathematical model of the user is characterized, for example, according to the cultural level of the user, the occupation of the user and the region of the user.
The method 3 comprises the following steps: the mathematical model of the user is characterized by the preference attributes of the user. For example, the mathematical model of the user is characterized according to the preference of the user to articles, for example, the user likes military articles, and the user is characterized as 'military fan'. The user likes scientific articles, and the user is characterized as a scientific fan. The user likes watching football news, and the user is characterized as a football fan.
In one embodiment, the user representation is formed by aggregating the users delineated according to different methods and dimensions. The user representation is used as a mathematical model characterizing the user. For example: the characteristics of the user a comprise 'youth', 'college student', 'cantonese Guangzhou' and 'science and technology enthusiasts', and the 4 characteristics are the mathematical description model of the user a.
And S20, generating a sample according to the characteristics of the user and the clicking behavior of the application program.
And after acquiring the characteristics of the user and the clicking behavior of the user on the application program, the server generates a sample according to the characteristics of the user and the clicking behavior of the user on the application program. Specifically, the characteristics of the user are used as input variables, and the clicking behavior of the user on the application program is used as a target variable to generate a sample.
For a user, an application is exposed as a sample. And taking the characteristics of the user as input variables, taking the clicking behavior of the user as a target variable, wherein the clicked target variable of the user is 1, and the non-clicked target variable of the user is 0. For example: a certain user has features (labels) of "young" and "military fan", and the application that exposes the user is "WeChat". The user does not click on the "WeChat" application, then the recorded sample is in the form of: x is (1, 1, …) and y is 0. x represents one of the samples. x the first two 1 s respectively indicate that the user has the "youth", "military fan" tags. Labels not included in x are denoted by 0. y-0 represents a negative sample.
S30, inputting the samples into a clustering model, and determining the population to which the samples belong in the clustering model.
And after the server generates a sample according to the characteristics of the user and the clicking behavior of the user on the application program, inputting the sample into the clustering model. And simultaneously determining the group to which the sample belongs in the clustering model. Specifically, there are a plurality of groups in the clustering model, and by inputting the samples into the plurality of groups of the clustering model, result values of the samples in the plurality of groups are obtained, and the group to which the sample belongs in the clustering model can be obtained according to the result values.
In one embodiment, the clustering model may generate samples by collecting characteristics of several users, as well as click behaviors of the applications by the users. And training the clustering model according to the sample so as to obtain an accurate clustering model aiming at a plurality of user groups, and further establishing an application program click rate estimation model for different user groups. The click rate estimation models of the application programs established by different user groups can be the same click rate estimation model or different click rate estimation models.
In other embodiments, the clustering model may be a predictive clustering model. I.e. the clustering models that have been formed for different user groups.
S40, inputting the sample into the click rate estimation model corresponding to the group, and obtaining the estimated click rate of the user to the application program.
After the server obtains that the sample belongs to the group to which the grouping model belongs, the sample is input into the click rate estimation model corresponding to the group, and therefore the estimated click rate of the user on the application program is obtained. In the clustering model, the click rate estimation model corresponding to each cluster is a click rate estimation model obtained by a standard formula of logistic regression. In other embodiments, the click rate estimation model corresponding to each group may also be a click rate estimation model obtained by other formulas. In addition, the click rate estimation models corresponding to each group in the grouping model may not be completely the same.
And S50, pushing the content to the user according to the estimated click rate of the application program.
In this embodiment, according to steps S10 to S40, the server may obtain the estimated click rate of the user for each application program in the application marketplace. And pushing corresponding content to the user according to the estimated click rate of the user to each application program.
The content herein may include the application itself and content related to the application. For example, similar applications and application presentations may also include corresponding advertising and information content obtained from the application.
In one embodiment, according to steps S10 to S40, the server may obtain the estimated click rate of the user for each application program in the application marketplace. And pushing the application program to the user according to the estimated click rate of the user to each application program.
According to the content pushing method, the sample is generated according to the characteristics of the user and the clicking behavior of the application program, and the sample is input into the clustering model to confirm that the sample belongs to the cluster in the clustering model. And then, obtaining the estimated click rate of the user to the application program according to the click rate estimation model in the group. And finally, pushing the content to the user according to the estimated click rate of the user to the application program. After the content pushing method determines the group to which the user belongs, the estimated click rate is obtained according to the click rate estimation model of the group, so that the accuracy of the estimated click rate is improved, and the corresponding content can be accurately pushed to the user according to the estimated click rate of the application program. For example, an application can be pushed to a user based accurately on its estimated click rate.
Fig. 2 is a flowchart of a content push method according to another embodiment of the present invention. As shown in fig. 2, before step S10, the method further includes:
s101, acquiring characteristics of historical users who expose the application program and clicking behaviors of the application program.
The server obtains the characteristics of all users who expose the history of the application program and the clicking behavior of each user on the application program. Here, the clicking behavior of the application by the user includes that the application is clicked by the user, and that the application is not clicked by the user. The application program can be various application programs in the application store, including game application programs, social application programs, reading application programs, system application programs, shopping application programs and the like.
The characteristics of the user include natural attributes of the user, social attributes of the user, or preference attributes of the user. The natural attributes of the user such as the age of the user, the gender of the user, etc. The social attributes of the user include the cultural level of the user, the occupation of the user, the region of the user, and the like. The preference attributes of the user are 'military fan', 'science and technology enthusiasts', 'football fan' and 'game fan', etc.
In one embodiment, the server obtains characteristics of a user of the exposure application and builds a mathematical model characterizing the user based on the characteristics of the user. The user is actually characterized by the characteristics of the user. The method for acquiring the user characteristics comprises the following methods:
the method comprises the following steps: the mathematical model of the user is characterized by the natural attributes of the user. The mathematical model of the user is characterized, for example, by the age of the user and the gender of the user. Wherein discretization is performed on the continuous quantities. Such as age dispersion into "children", "teenagers", "young", "middle-aged" and "old", etc.
The method 2 comprises the following steps: and describing the mathematical model of the user through the social attributes of the user. The mathematical model of the user is characterized, for example, according to the cultural level of the user, the occupation of the user and the region of the user.
The method 3 comprises the following steps: the mathematical model of the user is characterized by the preference attributes of the user. For example, the mathematical model of the user is characterized according to the preference of the user to articles, for example, the user likes military articles, and the user is characterized as 'military fan'. The user likes scientific articles, and the user is characterized as a scientific fan. The user likes watching football news, and the user is characterized as a football fan.
In one embodiment, the user representation is formed by aggregating the users delineated according to different methods and dimensions. The user representation is used as a mathematical model characterizing the user. For example: the characteristics of the user a comprise 'youth', 'college student', 'cantonese Guangzhou' and 'science and technology enthusiasts', and the 4 characteristics are the mathematical description model of the user a.
And S103, generating a sample according to the characteristics of the historical user and the clicking behavior of the application program.
And after acquiring the characteristics of all historical users and the clicking behavior of each user on the application program, the server generates a sample according to the characteristics of each user and the clicking behavior of each user on the application program. Specifically, the characteristics of the user are used as input variables, and the clicking behavior of the user on the application program is used as a target variable to generate a sample.
For a user, an application is exposed as a sample. And taking the characteristics of the user as input variables, taking the clicking behavior of the user as a target variable, wherein the clicked target variable of the user is 1, and the non-clicked target variable of the user is 0. For example: a certain user has features (labels) of "young" and "military fan", and the application that exposes the user is "WeChat". The user does not click on the "WeChat" application, then the recorded sample is in the form of: x is (1, 1, …) and y is 0. x represents one of the samples. x the first two 1 s respectively indicate that the user has the "youth", "military fan" tags. Labels not included in x are denoted by 0. y-0 represents a negative sample.
S105, training a grouping model comprising a plurality of groups and a click rate estimation model corresponding to the groups by using the sample.
In the present embodiment, the sample corresponds to a sample generated from the characteristics of all users and the click behavior of the application program in the history. From the sample, a clustering model containing a number of populations may be trained. Meanwhile, according to the sample, a click rate estimation model corresponding to each group in the group model can be trained. Specifically, a grouping model containing a plurality of groups is trained by using samples, the samples are classified according to the groups, and the samples classified correspondingly to the groups are used for training the click rate estimation model of the groups.
The method for training the click rate estimation model of the group by using the samples classified correspondingly to the group comprises the following steps: and training a click rate estimation model by using the samples classified correspondingly by the groups and utilizing a logistic algorithm according to a gradient descent method. The click rate prediction model is obtained according to a standard formula of the logistic regression, and the standard formula of the logistic regression is as follows:
where X denotes the input variable and β denotes the weight vector for the different features.
The method for obtaining the click rate estimation model of the group by adopting the logistic algorithm is only one method for obtaining the click rate estimation model, and besides the logistic algorithm, other algorithms can be used for obtaining the click rate estimation model of the group.
In this embodiment, the clustering model is trained using samples according to a decision tree algorithm. In other embodiments, the clustering model may also be trained using samples according to a naive bayes algorithm, a SVM (Support Vector Machines) algorithm, or a neural network algorithm. That is, the algorithm for training the clustering model using the samples is not limited, and the present solution merely provides several algorithms among them as an explanation.
In summary, in the present embodiment, the sample is generated by collecting the characteristics of the historical users and the click behaviors of the corresponding users on the application. And training a grouping model comprising a plurality of groups according to the sample, and training a click rate estimation model corresponding to each group in the grouping model, so as to generate click rate estimation models for different user groups, further acquiring the estimated click rate of the corresponding user to the application program according to the click rate estimation models of the different user groups, and accurately pushing corresponding contents to the user according to the estimated click rate. For example, the corresponding application program can be accurately pushed to the user according to the estimated click rate.
The following provides a specific embodiment for establishing a clustering model:
all exposure events of the last day are used for generating input variables of training samples according to the method, and for each sample, each sample is attributed to a group in each iteration process. For example: the number of clusters 3, sample x is a negative sample, the value calculated using the LR model of "cluster 1" during a certain iteration is 0.3, the value calculated using the LR model of "cluster 2" is 0.2, and the value calculated using the LR model of "cluster 3" is 0.1, so the LR model error of "cluster 3" is minimal, and the class of sample x is "cluster 3". And establishing a clustering model by using a decision tree algorithm.
C(X)＝Card(X)。
Wherein Card (X) is a decision tree algorithm, and the training of the method uses a model Card classification algorithm commonly used in the industry, which is not described herein again.
In one embodiment, as shown in fig. 3, the method for building a clustering logistic regression click rate estimation model includes:
s301, initialization: the samples are randomly divided into n groups, the iteration number i of a group LR (Logistic Regression) is 0, the maximum iteration number is m, and the iteration error threshold is c.
In the present embodiment, samples of all users in the history are randomly divided into n groups. The number of iterations of the cluster LR is i. At initialization, the number of iterations i is 0. Here, assume that the maximum number of iterations is m and the iteration error threshold is c.
S303, train an LR model for the samples of each cluster (positive samples are 1, negative samples are 0).
In this embodiment, a corresponding LR model is trained from the samples of each cluster. Wherein the positive sample is 1 and the negative sample is 0.
S305, each sample is substituted into each LR model to predict the click probability, and the sample is classified into a group closest to the target value of the sample (negative sample is closest to 0, and positive sample is closest to 1).
And randomly clustering according to the samples, training an LR model corresponding to each group according to the samples of each group, and then respectively substituting each sample in the samples into the LR model of each group to respectively obtain the predicted click rate, which is obtained according to the LR model, of each group corresponding to each sample. And comparing each sample according to the predicted click rate obtained by each group, and classifying the sample into the group closest to the target value of the sample (wherein negative samples are closest to 0, and positive samples are closest to 1).
S307, the error of all sample predictions is calculated. The error is Msm (i). Msm (i) ═ sum (| prediction of sample in this group LR model — true value of sample |).
In this embodiment, the error value between the predicted click rate of all samples in each cluster and the true click rate of the sample is calculated. Assuming that the error value is Msm (i), Msm (i) is sum (| sample is the predicted value of the group LR model-sample true |).
S309, a judgment condition is established, i.e., | Msm (i) -Msm (i-1) | < c Or > -m.
In the present embodiment, a judgment condition is established for the iterative process. The judgment condition is | Msm (i) -Msm (i-1) | < c or i > -m. When the condition does not satisfy the judgment condition, the execution returns to step S303. When the condition satisfies the determination condition, the following step S311 is performed.
S311, training a clustering model for predicting which cluster an unknown sample belongs to: regarding each group as a class, wherein n classes exist in total, obtaining the group to which each sample belongs in the steps, taking the group as a target variable, taking the characteristics of each sample as input variables of classification, substituting the input variables into a classification algorithm, and training a clustering model.
In the present embodiment, when the condition satisfies the determination condition in step 309, that is, all samples are classified into the belonging group. Further, a clustering model is trained for predicting the population to which the unknown sample belongs. Specifically, each group is regarded as one class, assuming that n classes are total, which group each sample belongs to is obtained in the above steps, and the group is used as a target variable, and each sample feature is used as an input variable of classification and is substituted into a classification algorithm to train a clustering model. In this embodiment, the classification algorithm selects a decision tree algorithm. In other embodiments, a naive bayes algorithm, an SVM algorithm, or a neural network algorithm may also be selected.
Fig. 4 provides a usage scenario of a model of a content push method. The content here is a sample. Sample x may be an application. As shown in fig. 4, the method comprises the following steps:
s401, sorting input characteristic data of the unknown sample x.
And S403, substituting the characteristic data characteristics of the unknown sample x into the clustering model to obtain the unknown sample x belonging to the Kth cluster.
S405, substituting the feature data of the unknown sample x into the LR model of the K group to obtain the estimated click rate of the unknown sample x.
That is, the input feature data of the unknown sample x is obtained first. And substituting the characteristic data of the unknown sample x into the clustering model, thereby obtaining that the unknown sample x belongs to a population (K-th population) in the clustering model. And substituting the characteristic data of the unknown sample x into a click rate estimation model (LR model) of the group to which the characteristic data belongs to obtain the estimated click rate of the unknown sample x.
S407, pushing the sample x to the user according to the estimated click rate of the sample x.
According to steps S401 to S405, the estimated click rate of the user for the sample x can be obtained. And pushing the sample x to the user according to the estimated click rate of the user on the sample x.
In one embodiment, step S50 includes step a: and pushing the application program to the user according to the estimated click rate of the application program. Specifically, as shown in fig. 5, step a includes: s501, confirming that the estimated click rate of the application program is larger than a threshold value, and pushing the application program to the user.
The server may obtain the estimated click rate of the user for each application program in the application store according to steps S10 to S40. When the estimated click rate of the user to each application program in the application store is larger than a threshold value (which is set in advance according to actual requirements), the server pushes the application program to the user. Therefore, the application programs which are interested in the user in the application store can be pushed to the user.
In one embodiment, as shown in fig. 6, step a includes the steps of:
s503, calculating the estimated click rate of the user to all the application programs in the application store, sequencing the application programs according to the estimated click rate from high to low, and pushing the application programs in the top preset number to the user.
The server respectively obtains estimated click rates of the user for the plurality of application programs in the application store according to the steps S10 to S40. Thus, the server may calculate the user's estimated click-through rate for all applications in the application store. And sequencing the application programs from high to low according to the estimated click rate of each application program, thereby acquiring the application programs in the top preset number and pushing the application programs in the preset number to the user. For example, in an application store, 8 applications need to be pushed to the user. The server acquires the estimated click rate of all the application programs of the application store according to the scheme of the invention. And sequencing all the application programs from high to low according to the estimated click rate of each application program. Further, the 8 application programs ranked at the top are acquired, and the 8 application programs are pushed to the user in the application store.
The application program pushing method provided by the embodiment can screen the application programs of the mobile phone application store according to the estimation of the click rate of the application programs so as to push the application programs interested by the user to the user.
The invention also provides a content pushing device, as shown in fig. 7. The content pushing device comprises a first obtaining module 701, a generating module 703, a determining module 705, a second obtaining module 707 and a pushing module 709.
The first obtaining module 701 is used for obtaining features of a user who exposes an application program and click behaviors of the application program. When detecting that the user exposes the application program, the server acquires the characteristics of the user who exposes the application program and acquires the clicking behavior of the user on the application program. Here, the clicking behavior of the application by the user includes that the application is clicked by the user, and that the application is not clicked by the user. The application program can be various application programs in the application store, including game application programs, social application programs, reading application programs, system application programs, shopping application programs and the like.
The characteristics of the user include natural attributes of the user, social attributes of the user, or preference attributes of the user. The natural attributes of the user such as the age of the user, the gender of the user, etc. The social attributes of the user include the cultural level of the user, the occupation of the user, the region of the user, and the like. The preference attributes of the user are 'military fan', 'science and technology enthusiasts', 'football fan' and 'game fan', etc.
In one embodiment, the server obtains characteristics of a user of the exposure application and builds a mathematical model characterizing the user based on the characteristics of the user. The user is actually characterized by the characteristics of the user. The method for acquiring the user characteristics comprises the following methods:
the method comprises the following steps: the mathematical model of the user is characterized by the natural attributes of the user. The mathematical model of the user is characterized, for example, by the age of the user and the gender of the user. Wherein discretization is performed on the continuous quantities. Such as age dispersion into "children", "teenagers", "young", "middle-aged" and "old", etc.
The method 2 comprises the following steps: and describing the mathematical model of the user through the social attributes of the user. The mathematical model of the user is characterized, for example, according to the cultural level of the user, the occupation of the user and the region of the user.
The method 3 comprises the following steps: the mathematical model of the user is characterized by the preference attributes of the user. For example, the mathematical model of the user is characterized according to the preference of the user to articles, for example, the user likes military articles, and the user is characterized as 'military fan'. The user likes scientific articles, and the user is characterized as a scientific fan. The user likes watching football news, and the user is characterized as a football fan.
In one embodiment, the user representation is formed by aggregating the users delineated according to different methods and dimensions. The user representation is used as a mathematical model characterizing the user. For example: the characteristics of the user a comprise 'youth', 'college student', 'cantonese Guangzhou' and 'science and technology enthusiasts', and the 4 characteristics are the mathematical description model of the user a.
The generating module 703 is configured to generate a sample according to the characteristics of the user and the click behavior of the application. And after acquiring the characteristics of the user and the clicking behavior of the user on the application program, the server generates a sample according to the characteristics of the user and the clicking behavior of the user on the application program. Specifically, the characteristics of the user are used as input variables, and the clicking behavior of the user on the application program is used as a target variable to generate a sample.
For a user, an application is exposed as a sample. And taking the characteristics of the user as input variables, taking the clicking behavior of the user as a target variable, wherein the clicked target variable of the user is 1, and the non-clicked target variable of the user is 0. For example: a certain user has features (labels) of "young" and "military fan", and the application that exposes the user is "WeChat". The user does not click on the "WeChat" application, then the recorded sample is in the form of: x is (1, 1, …) and y is 0. x represents one of the samples. x the first two 1 s respectively indicate that the user has the "youth", "military fan" tags. Labels not included in x are denoted by 0. y-0 represents a negative sample.
The determining module 705 is configured to input the sample into a clustering model, and determine a population to which the sample belongs in the clustering model. And after the server generates a sample according to the characteristics of the user and the clicking behavior of the user on the application program, inputting the sample into the clustering model. And simultaneously determining the group to which the sample belongs in the clustering model. Specifically, there are a plurality of groups in the clustering model, and by inputting the samples into the plurality of groups of the clustering model, result values of the samples in the plurality of groups are obtained, and the group to which the sample belongs in the clustering model can be obtained according to the result values.
In one embodiment, the clustering model may generate samples by collecting characteristics of several users, as well as click behaviors of the applications by the users. And training the clustering model according to the sample so as to obtain an accurate clustering model aiming at a plurality of user groups, and further establishing an application program click rate estimation model for different user groups. The click rate estimation models of the application programs established by different user groups can be the same click rate estimation model or different click rate estimation models.
In other embodiments, the clustering model may be a predictive clustering model. I.e. the clustering models that have been formed for different user groups.
The second obtaining module 707 is configured to input the sample into a click rate estimation model corresponding to the group, and obtain an estimated click rate of the user on the application program. After the server obtains that the sample belongs to the group to which the grouping model belongs, the sample is input into the click rate estimation model corresponding to the group, and therefore the estimated click rate of the user on the application program is obtained. In the clustering model, the click rate estimation model corresponding to each cluster is a click rate estimation model obtained by a standard formula of logistic regression. In other embodiments, the click rate estimation model corresponding to each group may also be a click rate estimation model obtained by other formulas. In addition, the click rate estimation models corresponding to each group in the grouping model may not be completely the same.
The pushing module 709 is configured to push content to the user according to the estimated click rate of the application program. In this embodiment, according to steps S10 to S40, the server may obtain the estimated click rate of the user for each application program in the application marketplace. And pushing corresponding content to the user according to the estimated click rate of the user to each application program.
The content herein may include the application itself and content related to the application. For example, similar applications and application presentations may also include corresponding advertising and information content obtained from the application.
In one embodiment, according to steps S10 to S40, the server may obtain the estimated click rate of the user for each application program in the application marketplace. And pushing the application program to the user according to the estimated click rate of the user to each application program.
The invention also provides computer equipment. The computer device includes one or more processors, memory, and one or more application programs. Wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to perform the content push method of any of the above embodiments.
Fig. 8 is a schematic structural diagram of a computer device according to an embodiment of the present invention. Such as servers, personal computers, and network appliances. As shown in fig. 8, the apparatus includes a processor 803, a memory 805, an input unit 807, and a display unit 809. Those skilled in the art will appreciate that the device configuration means shown in fig. 8 do not constitute a limitation of all devices and may include more or less components than those shown, or some components in combination. The memory 805 may be used to store the application program 801 and various functional modules, and the processor 803 executes the application program 801 stored in the memory 805, thereby performing various functional applications of the device and data processing. The memory may be internal or external memory, or include both internal and external memory. The internal memory may include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), flash memory, or random access memory. The external memory may include a hard disk, a floppy disk, a ZIP disk, a usb-disk, a magnetic tape, etc. The disclosed memory includes, but is not limited to, these types of memory. The disclosed memory is by way of example only and not by way of limitation.
The input unit 807 is used to receive input of signals and keywords input by a user. The input unit 807 may include a touch panel and other input devices. The touch panel can collect touch operations of a user on or near the touch panel (for example, operations of the user on or near the touch panel by using any suitable object or accessory such as a finger, a stylus and the like) and drive the corresponding connecting device according to a preset program; other input devices may include, but are not limited to, one or more of a physical keyboard, function keys (e.g., play control keys, switch keys, etc.), a trackball, a mouse, a joystick, and the like. The display unit 809 may be used to display information input by the user or information provided to the user and various menus of the computer apparatus. The display unit 809 may take the form of a liquid crystal display, an organic light emitting diode, or the like. The processor 803 is a control center of a computer device, connects various parts of the entire computer using various interfaces and lines, and performs various functions and processes data by operating or executing software programs and/or modules stored in the memory 803 and calling data stored in the memory.
In one embodiment, the computer device includes one or more processors 803, as well as one or more memories 805, one or more applications 801. Wherein the one or more application programs 801 are stored in the memory 805 and configured to be executed by the one or more processors 803, the one or more application programs 801 configured to perform the content push method described in the above embodiments.
In addition, functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.
Those skilled in the art will appreciate that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, and the program may be stored in a computer readable storage medium, and the storage medium may include a memory, a magnetic disk, an optical disk, or the like.
The foregoing is only a partial embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Claims (13)
1. A content push method, comprising the steps of:
acquiring the characteristics of a user who exposes the application program and the clicking behavior of the application program;
generating a sample according to the characteristics of the user and the clicking behavior of the application program; the method comprises the following steps: taking the characteristics of the user as input variables and the clicking behavior as target variables to generate a sample;
inputting the samples into a clustering model, and determining a population to which the samples belong in the clustering model;
inputting the sample into a click rate estimation model corresponding to the group to obtain the estimated click rate of the user to the application program;
and pushing content to the user according to the estimated click rate of the application program.
2. The content pushing method according to claim 1, wherein the pushing content to the user according to the estimated click-through rate of the application comprises:
and pushing the application program to the user according to the estimated click rate of the application program.
3. The content pushing method according to claim 2, wherein the pushing the application to the user according to the estimated click-through rate of the application comprises:
and confirming that the estimated click rate of the application program is greater than a threshold value, and pushing the application program to the user.
4. The content pushing method according to claim 2, wherein the pushing the application to the user according to the estimated click-through rate of the application comprises:
and calculating the estimated click rate of the user to all the application programs in the application market, sequencing the application programs according to the estimated click rate from high to low, and pushing the application programs in the top preset number to the user.
5. The content pushing method according to claim 1, wherein before the obtaining of the feature of the user who exposes the application and the click action on the application, the method further comprises:
acquiring the characteristics of historical users who expose the application program and the clicking behavior of the application program;
generating a sample according to the characteristics of the historical user and the clicking behavior of the application program;
and training a grouping model comprising a plurality of groups and a click rate estimation model corresponding to the groups by using the sample.
6. The content push method according to claim 5, wherein said training a clustering model including populations using said samples comprises:
training a clustering model using the samples according to a decision tree algorithm.
7. The method according to claim 5, wherein the training of the clustering model including a plurality of groups and the click-through rate prediction model corresponding to the groups using the samples comprises:
and training a grouping model containing a plurality of groups by using the samples, classifying the samples according to the groups, and training a click rate estimation model of the group by using the samples classified correspondingly to the groups.
8. The method of claim 7, wherein the training of the click-through rate prediction model of the group using the samples classified according to the group comprises:
and training the click rate estimation model by using the samples of the corresponding group classification and using a logistic algorithm.
9. The content push method according to claim 8, wherein the click-through rate prediction model is a click-through rate prediction model obtained according to a formula of logistic regression; the formula of the logistic regression is as follows:
where X denotes the input variable and β denotes the weight vector for the different features.
10. The method of claim 8, wherein the training the click through rate prediction model using the population-to-population classified samples using a logistic algorithm comprises:
and training the group by using a logistic algorithm according to a gradient descent method, and training a click rate estimation model in the corresponding group.
11. The content pushing method according to claim 1, wherein the characteristics of the user include a natural attribute, a social attribute, or a preference attribute of the user.
12. A content pushing apparatus, comprising:
the first acquisition module is used for acquiring the characteristics of a user who exposes the application program and the clicking behavior of the application program;
the generating module is used for generating a sample according to the characteristics of the user and the clicking behavior of the application program; the method comprises the following steps: taking the characteristics of the user as input variables and the clicking behavior as target variables to generate a sample;
the determining module is used for inputting the samples into a clustering model and determining a population to which the samples belong in the clustering model;
the second acquisition module is used for inputting the sample into a click rate estimation model corresponding to the group to acquire the estimated click rate of the user on the application program;
and the pushing module is used for pushing the content to the user according to the estimated click rate of the application program.
13. A computer device, comprising:
one or more processors;
a memory;
one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications configured to perform the content push method of any of claims 1 to 11.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710985244.3A CN107613022B (en) | 2017-10-20 | 2017-10-20 | Content pushing method and device and computer equipment |
| PCT/CN2018/105810 WO2019076173A1 (en) | 2017-10-20 | 2018-09-14 | Content pushing method and apparatus, and computer device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710985244.3A CN107613022B (en) | 2017-10-20 | 2017-10-20 | Content pushing method and device and computer equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107613022A CN107613022A (en) | 2018-01-19 |
| CN107613022B true CN107613022B (en) | 2020-10-16 |
Family
ID=61078736
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710985244.3A Active CN107613022B (en) | 2017-10-20 | 2017-10-20 | Content pushing method and device and computer equipment |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN107613022B (en) |
| WO (1) | WO2019076173A1 (en) |
Families Citing this family (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107613022B (en) * | 2017-10-20 | 2020-10-16 | 阿里巴巴（中国）有限公司 | Content pushing method and device and computer equipment |
| CN107766580A (en) * | 2017-11-20 | 2018-03-06 | 北京奇虎科技有限公司 | The method for pushing and device of message |
| CN108427708B (en) * | 2018-01-25 | 2021-06-25 | 腾讯科技（深圳）有限公司 | Data processing method, data processing apparatus, storage medium, and electronic apparatus |
| CN108510352A (en) * | 2018-02-09 | 2018-09-07 | 广州优视网络科技有限公司 | Application issue recommends method, apparatus and computer equipment |
| CN110210884B (en) * | 2018-05-29 | 2023-05-05 | 腾讯科技（深圳）有限公司 | Method, device, computer equipment and storage medium for determining user characteristic data |
| CN109190791A (en) * | 2018-07-25 | 2019-01-11 | 广州优视网络科技有限公司 | Using the appraisal procedure of recommended models, device and electronic equipment |
| CN109495552A (en) * | 2018-10-31 | 2019-03-19 | 北京字节跳动网络技术有限公司 | Method and apparatus for updating clicking rate prediction model |
| CN109408723A (en) * | 2018-11-06 | 2019-03-01 | 北京奇艺世纪科技有限公司 | A kind of method for pushing and device |
| CN109359247B (en) * | 2018-12-07 | 2021-07-06 | 广州市百果园信息技术有限公司 | Content pushing method, storage medium and computer equipment |
| CN110008399B (en) * | 2019-01-30 | 2023-04-07 | 创新先进技术有限公司 | Recommendation model training method and device, and recommendation method and device |
| CN110046952B (en) * | 2019-01-30 | 2021-12-10 | 创新先进技术有限公司 | Recommendation model training method and device, and recommendation method and device |
| CN110033342A (en) * | 2019-01-30 | 2019-07-19 | 阿里巴巴集团控股有限公司 | A kind of training method and device, a kind of recommended method and device of recommended models |
| CN110264276A (en) * | 2019-06-25 | 2019-09-20 | 广州视源电子科技股份有限公司 | Determine method, apparatus, equipment and the storage medium of push user group |
| CN111160961B (en) * | 2019-12-20 | 2023-10-17 | 深圳前海微众银行股份有限公司 | Information recommendation method and device |
| CN115004667B (en) * | 2020-04-24 | 2024-04-02 | 深圳市欢太科技有限公司 | Information pushing method, device, electronic equipment and computer readable medium |
| CN112669078A (en) * | 2020-12-30 | 2021-04-16 | 上海众源网络有限公司 | Behavior prediction model training method, device, equipment and storage medium |
| CN113743643B (en) * | 2021-02-05 | 2023-11-03 | 北京京东振世信息技术有限公司 | Method, device, equipment and medium for determining commodity data prediction accuracy |
| CN116821512B (en) * | 2023-08-25 | 2024-02-20 | 深圳唯爱智云科技有限公司 | Recommendation model training method and device, recommendation method and device |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102346899A (en) * | 2011-10-08 | 2012-02-08 | 亿赞普(北京)科技有限公司 | Method and device for predicting advertisement click rate based on user behaviors |
| CN102521767A (en) * | 2011-12-13 | 2012-06-27 | 亿赞普(北京)科技有限公司 | Method and system for publishing network advertising information |
| CN103345512A (en) * | 2013-07-06 | 2013-10-09 | 北京品友互动信息技术有限公司 | Online advertising click-through rate forecasting method and device based on user attribute |
| CN105631711A (en) * | 2015-12-30 | 2016-06-01 | 合一网络技术(北京)有限公司 | Advertisement putting method and apparatus |
| CN105701191A (en) * | 2016-01-08 | 2016-06-22 | 腾讯科技（深圳）有限公司 | Push information click rate estimation method and device |
| CN105809462A (en) * | 2014-12-30 | 2016-07-27 | Tcl集团股份有限公司 | Estimation method and device for estimating advertisement click rate |
| CN105897847A (en) * | 2015-12-15 | 2016-08-24 | 乐视网信息技术（北京）股份有限公司 | Information push method and device |
| WO2016169427A1 (en) * | 2015-04-21 | 2016-10-27 | 腾讯科技（深圳）有限公司 | Correction method for advertisement click-through rate and advertisement delivery server |
| CN106096629A (en) * | 2016-06-01 | 2016-11-09 | 南京邮电大学 | A kind of based on user between the ad click rate Forecasting Methodology of similarity relation |
| CN106570718A (en) * | 2015-10-13 | 2017-04-19 | 深圳市腾讯计算机系统有限公司 | Information releasing method and releasing system |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110238496A1 (en) * | 2010-02-23 | 2011-09-29 | Vishal Gurbuxani | Systems and Methods for Generating Data from Mobile Applications and Dynamically Delivering Advertising Based on Generated Data |
| CN104572962A (en) * | 2014-12-31 | 2015-04-29 | 浙江大学 | APP (Application) recommendation method and system |
| US20170004524A1 (en) * | 2015-06-30 | 2017-01-05 | Yahoo! Inc. | Systems and Methods For Mobile Campaign Optimization Without Knowing User Identity |
| CN105678587B (en) * | 2016-01-12 | 2020-11-24 | 腾讯科技（深圳）有限公司 | Recommendation feature determination method, information recommendation method and device |
| CN107613022B (en) * | 2017-10-20 | 2020-10-16 | 阿里巴巴（中国）有限公司 | Content pushing method and device and computer equipment |
-
2017
- 2017-10-20 CN CN201710985244.3A patent/CN107613022B/en active Active
-
2018
- 2018-09-14 WO PCT/CN2018/105810 patent/WO2019076173A1/en active Application Filing
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102346899A (en) * | 2011-10-08 | 2012-02-08 | 亿赞普(北京)科技有限公司 | Method and device for predicting advertisement click rate based on user behaviors |
| CN102521767A (en) * | 2011-12-13 | 2012-06-27 | 亿赞普(北京)科技有限公司 | Method and system for publishing network advertising information |
| CN103345512A (en) * | 2013-07-06 | 2013-10-09 | 北京品友互动信息技术有限公司 | Online advertising click-through rate forecasting method and device based on user attribute |
| CN105809462A (en) * | 2014-12-30 | 2016-07-27 | Tcl集团股份有限公司 | Estimation method and device for estimating advertisement click rate |
| WO2016169427A1 (en) * | 2015-04-21 | 2016-10-27 | 腾讯科技（深圳）有限公司 | Correction method for advertisement click-through rate and advertisement delivery server |
| CN106570718A (en) * | 2015-10-13 | 2017-04-19 | 深圳市腾讯计算机系统有限公司 | Information releasing method and releasing system |
| CN105897847A (en) * | 2015-12-15 | 2016-08-24 | 乐视网信息技术（北京）股份有限公司 | Information push method and device |
| CN105631711A (en) * | 2015-12-30 | 2016-06-01 | 合一网络技术(北京)有限公司 | Advertisement putting method and apparatus |
| CN105701191A (en) * | 2016-01-08 | 2016-06-22 | 腾讯科技（深圳）有限公司 | Push information click rate estimation method and device |
| CN106096629A (en) * | 2016-06-01 | 2016-11-09 | 南京邮电大学 | A kind of based on user between the ad click rate Forecasting Methodology of similarity relation |
Non-Patent Citations (2)
| Title |
|---|
| 基于多维特征组合逻辑回归模型的广告点击率预测;伊雯雯;《通信技术》;20160930;第49卷(第9期);全文 * |
| 基于用户相似度和特征分化的广告点击率预测研究;潘书敏;《计算机科学》;20170215;第44卷(第2期);第284-285，287页，图2 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107613022A (en) | 2018-01-19 |
| WO2019076173A1 (en) | 2019-04-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107613022B (en) | Content pushing method and device and computer equipment | |
| US10958748B2 (en) | Resource push method and apparatus | |
| US10671933B2 (en) | Method and apparatus for evaluating predictive model | |
| CN111984689B (en) | Information retrieval method, device, equipment and storage medium | |
| US20190354810A1 (en) | Active learning to reduce noise in labels | |
| CN108921221A (en) | Generation method, device, equipment and the storage medium of user characteristics | |
| US10511681B2 (en) | Establishing and utilizing behavioral data thresholds for deep learning and other models to identify users across digital space | |
| JP6911603B2 (en) | How to generate predictive models for the categories of facilities visited by users, programs, server equipment, and processing equipment | |
| US11146580B2 (en) | Script and command line exploitation detection | |
| CN109165975B (en) | Label recommending method, device, computer equipment and storage medium | |
| CN108769125B (en) | Application recommendation method and device, storage medium and computer equipment | |
| WO2019061664A1 (en) | Electronic device, user's internet surfing data-based product recommendation method, and storage medium | |
| CN107862551B (en) | Method and device for predicting network application promotion effect and terminal equipment | |
| WO2017075980A1 (en) | Information pushing method and apparatus | |
| CN111460384A (en) | Policy evaluation method, device and equipment | |
| CN111275205A (en) | Virtual sample generation method, terminal device and storage medium | |
| CN113076416A (en) | Information heat evaluation method and device and electronic equipment | |
| CN111563198B (en) | Material recall method, device, equipment and storage medium | |
| CN108520034B (en) | Application recommendation method and device and computer equipment | |
| US11102161B2 (en) | Social networking service content sharing | |
| CN107786376B (en) | Content pushing method and device and computer equipment | |
| CN117897738A (en) | Artificial intelligence assisted live sports data quality assurance | |
| KR102242042B1 (en) | Method, apparatus and computer program for data labeling | |
| CN111191142B (en) | Electronic resource recommendation method and device and readable medium | |
| Giannakopoulos et al. | Revisiting the effect of history on learning performance: the problem of the demanding lord |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200508 Address after: 310052 room 508, floor 5, building 4, No. 699, Wangshang Road, Changhe street, Binjiang District, Hangzhou City, Zhejiang Province Applicant after: Alibaba (China) Co.,Ltd. Address before: 510640 Guangdong city of Guangzhou province Whampoa Tianhe District Road No. 163 Xiping Yun Lu Yun Ping square B radio tower 15 layer self unit 02 Applicant before: GUANGZHOU UC NETWORK TECHNOLOGY Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |