CN111047343A - Method, device, system and medium for information push - Google Patents
Method, device, system and medium for information push Download PDFInfo
- Publication number
- CN111047343A CN111047343A CN201811201106.2A CN201811201106A CN111047343A CN 111047343 A CN111047343 A CN 111047343A CN 201811201106 A CN201811201106 A CN 201811201106A CN 111047343 A CN111047343 A CN 111047343A
- Authority
- CN
- China
- Prior art keywords
- classification model
- user
- variables
- training data
- variable
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000013145 classification model Methods 0.000 claims abstract description 121
- 238000012549 training Methods 0.000 claims abstract description 104
- 238000010276 construction Methods 0.000 claims abstract description 9
- 238000012216 screening Methods 0.000 claims description 32
- 238000007477 logistic regression Methods 0.000 claims description 26
- 230000015654 memory Effects 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 238000001914 filtration Methods 0.000 claims description 3
- 238000004590 computer program Methods 0.000 description 11
- 238000010586 diagram Methods 0.000 description 11
- 239000000047 product Substances 0.000 description 11
- 230000006399 behavior Effects 0.000 description 7
- 238000004891 communication Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 238000012545 processing Methods 0.000 description 5
- 238000003066 decision tree Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 235000013305 food Nutrition 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000004806 packaging method and process Methods 0.000 description 2
- 239000000758 substrate Substances 0.000 description 2
- 108010001267 Protein Subunits Proteins 0.000 description 1
- 239000013065 commercial product Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
Landscapes
- Business, Economics & Management (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present disclosure provides a method, apparatus, system, and medium for information push. The method comprises the following steps: obtaining at least one effective characteristic variable of a first user, wherein the at least one effective characteristic variable belongs to R candidate characteristic variables which have the most important influence on the construction of a first classification model in N initial characteristic variables, the first classification model is obtained through training of a first group of training data, the first group of training data comprises the N initial characteristic variables, N and R are integers, and N is more than R and is more than or equal to 2; obtaining potential values of the first user becoming potential users based on at least one valid feature variable of the first user and a second classification model, wherein the second classification model is obtained through a second set of training data, and the second set of training data comprises the R candidate feature variables; when the potential value meets a first condition, pushing a message to the first user. Wherein the first classification model is different from the second classification model.
Description
Technical Field
The present disclosure relates to the field of internet technologies, and in particular, to a method, an apparatus, a system, and a medium for pushing information.
Background
With the development of the internet economy, information push related to the commodities and the services is directionally carried out according to the characteristics of user groups, the consumption behaviors of potential users can be effectively mined, and the operation efficiency of the commodities and the services is improved. When information push is carried out, screening out potential customer groups is very important for successful conversion of push messages. In the prior art, screening is generally performed according to a specific rule (for example, historical consumption habits or commodity browsing habits of a user), or a potential user is predicted by training a classification model through machine learning.
In carrying out the disclosed concept, the inventors have discovered that there are at least the following problems in the prior art. When a potential user is screened by using a specific rule, the historical behavior of the user may be affected by specific events such as promotion or holidays, and the predicted future behavior of the user may be very inaccurate. When the machine learning is used for training the classification model, sometimes the classification model may not contribute to predicting the behavior of the potential user due to the limitations of the acquisition path of the training data or the data dimension. For example, the present inventors have discovered that for some financial products, the acquisition of purchased users over the internet platform differs significantly from the characteristics of unpurchased users. When the behavior of the user who does not purchase is predicted by the classification model obtained from the training data, the purchase probability of most users still exists in the interval of 0-0.1 (the purchase probability is 0-10%). Thus, the classification model does not have any substantial contribution to the mining of potential users. Moreover, there may be a possibility of overfitting.
Disclosure of Invention
In view of the above, the present disclosure provides a method, an apparatus, a system, and a medium for information push, which can more effectively mine potential users for effective information push.
One aspect of the present disclosure provides a method for information push. The method comprises the steps of firstly, obtaining at least one effective characteristic variable of a first user, wherein the at least one effective characteristic variable belongs to R candidate characteristic variables which have the most important influence on the construction of a first classification model in N initial characteristic variables, the first classification model is obtained through training of a first set of training data, the first set of training data comprises the N initial characteristic variables, N and R are integers, and N is more than R and is more than or equal to 2; then, obtaining a potential value of the first user becoming a potential user based on at least one valid feature variable of the first user and a second classification model, wherein the second classification model is obtained through a second set of training data, and the second set of training data comprises the R candidate feature variables; thereafter, a message is pushed to the first user when the potential value satisfies a first condition. Wherein the first classification model is different from the second classification model.
According to an embodiment of the present disclosure, the method further comprises training the first classification model with the first set of training data, screening the R candidate feature variables from the N initial feature variables based on the first classification model, and training the second classification model with the second set of training data.
According to an embodiment of the present disclosure, the method further includes determining the at least one valid feature variable based on a correlation between an output of the second classification model and each of the R candidate feature variables during a training process.
According to an embodiment of the present disclosure, the first classification model comprises an xgboost extreme gradient ascent model, and/or the second classification model comprises a logistic regression model.
According to the embodiment of the disclosure, the R candidate feature variables are feature variables obtained through the xgb.
According to an embodiment of the present disclosure, obtaining a potential value of the first user becoming a potential user based on the at least one valid feature variable of the first user and the second classification model includes obtaining a regression coefficient of each feature variable of the at least one valid feature variable in the logistic regression model, and obtaining the potential value based on the at least one valid feature variable of the first user and the regression coefficient.
Another aspect of the present disclosure provides an apparatus for information push. The device comprises a user characteristic acquisition module, a potential value calculation module and a push module. The user characteristic obtaining module is used for obtaining at least one effective characteristic variable of a first user, wherein the at least one effective characteristic variable belongs to R candidate characteristic variables which have the most important influence on the construction of a first classification model in N initial characteristic variables, the first classification model is obtained through training of a first set of training data, the first set of training data comprises the N initial characteristic variables, N and R are integers, and N is larger than R and is larger than or equal to 2. The potential value calculation module is used for predicting potential values of the first user to become potential users based on at least one effective characteristic variable of the first user and a second classification model, wherein the second classification model is obtained through a second set of training data, and the second set of training data comprises the R candidate characteristic variables. The pushing module is used for pushing a message to the first user when the potential value meets a first condition. Wherein the first classification model is different from the second classification model.
According to an embodiment of the present disclosure, the apparatus further includes a first training module, a first screening module, and a second training module. The first training module is to train the first classification model with the first set of training data. The first screening module is used for screening the R candidate characteristic variables from the N initial characteristic variables based on the first classification model. A second training module to train the second classification model with the second set of training data.
According to an embodiment of the present disclosure, the apparatus further comprises a second screening module. The second screening module is configured to determine the at least one valid feature variable based on a correlation between an output of the second classification model and each of the R candidate feature variables during a training process.
According to an embodiment of the present disclosure, the first classification model comprises an xgboost extreme gradient ascent model, and/or the second classification model comprises a logistic regression model.
According to the embodiment of the disclosure, the R candidate feature variables are feature variables obtained through the xgb.
According to an embodiment of the present disclosure, the potential value calculation module is specifically configured to: obtaining a regression coefficient for each of the at least one significant feature variable in the logistic regression model, and obtaining the latent value based on the at least one significant feature variable of the first user and the regression coefficient.
Another aspect of the present disclosure also provides a system for pushing information, including one or more memories storing executable instructions; and one or more processors executing the executable instructions to implement the method as described above.
Another aspect of the present disclosure also provides a computer-readable storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to perform the method as described above.
Another aspect of the disclosure provides a computer program comprising computer executable instructions for implementing the method as described above when executed.
According to the embodiment of the disclosure, the problem that potential users cannot be effectively screened when the message is pushed can be at least partially solved, and therefore, the potential users can be effectively mined, and the message pushing has the technical effect of pertinence and effectiveness.
Drawings
The above and other objects, features and advantages of the present disclosure will become more apparent from the following description of embodiments of the present disclosure with reference to the accompanying drawings, in which:
fig. 1 schematically illustrates a system architecture of a method, apparatus, system and medium for information push according to an embodiment of the present disclosure;
fig. 2 schematically shows a flowchart of a method for information push according to an embodiment of the present disclosure;
fig. 3 schematically shows a flow chart of a method for information push according to another embodiment of the present disclosure;
FIG. 4 schematically illustrates a flow chart of a method of obtaining potential values through a logistic regression model, according to an embodiment of the disclosure;
FIG. 5A schematically illustrates an application example flow diagram of a method for information push in accordance with an embodiment of the disclosure;
FIG. 5B illustrates a conceptual diagram of a training process in the application example method of FIG. 5A;
fig. 6 schematically shows a block diagram of an apparatus for information push according to an embodiment of the present disclosure;
FIG. 7 schematically shows a block diagram for an information-pushing computer system, in accordance with an embodiment of the disclosure.
Detailed Description
Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is illustrative only and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.
All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It is noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification and should not be interpreted in an idealized or overly formal sense.
Where a convention analogous to "at least one of A, B and C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B and C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). Where a convention analogous to "A, B or at least one of C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B or C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.).
The embodiment of the disclosure provides a method, a device, a system and a medium for information push. The method comprises the following steps: obtaining at least one effective characteristic variable of a first user, wherein the at least one effective characteristic variable belongs to R candidate characteristic variables which have the most important influence on the construction of a first classification model in N initial characteristic variables, the first classification model is obtained through training of a first group of training data, the first group of training data comprises the N initial characteristic variables, N and R are integers, and N is more than R and is more than or equal to 2; then obtaining a potential value of the first user becoming a potential user based on at least one effective characteristic variable of the first user and a second classification model, wherein the second classification model is obtained through a second set of training data, and the second set of training data comprises the R candidate characteristic variables; thereafter, a message is pushed to the first user when the potential value satisfies a first condition. Wherein the first classification model is different from the second classification model.
According to the embodiment of the disclosure, on the basis of a first classification model, R candidate characteristic variables which have the most important influence on the first classification model are obtained by screening from N initial characteristic variables, a second classification model is trained through the R candidate characteristic variables, and then the potential value of a potential user is predicted to be based on at least one effective characteristic variable in the second classification model and the R candidate characteristic variables, so that the potential user screening can be more efficiently realized than the single machine learning model prediction, and the information pushing is more efficiently performed.
According to the embodiment of the disclosure, through the combination of the first classification model and the second classification model and the multiple screening of the characteristic variables, the difference between different users can be enlarged, and when the popularization and marketing are needed, the probability score of the user with the larger difference is more beneficial to the screening. Moreover, through multiple screening of characteristic variables and combination of different classification models, partial overfitting influence can be reduced, and generalization capability of a prediction result is improved.
Fig. 1 schematically illustrates a system architecture 100 of a method, apparatus, system, and medium for information push according to an embodiment of the present disclosure. It should be noted that fig. 1 is only an example of a system architecture to which the embodiments of the present disclosure may be applied to help those skilled in the art understand the technical content of the present disclosure, and does not mean that the embodiments of the present disclosure may not be applied to other devices, systems, environments or scenarios.
As shown in fig. 1, the system architecture 100 according to this embodiment may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 serves as a medium for providing communication links between the terminal devices 101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few.
The user may use the terminal devices 101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The terminal devices 101, 102, 103 may have installed thereon various communication client applications, such as shopping-like applications, web browser applications, search-like applications, instant messaging tools, mailbox clients, social platform software, etc. (by way of example only).
The terminal devices 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.
The server 105 may be a server providing various services, such as a background management server (for example only) providing support for websites browsed by users using the terminal devices 101, 102, 103. The background management server may analyze and perform other processing on the received data such as the user request, and feed back a processing result (e.g., a webpage, information, or data obtained or generated according to the user request) to the terminal device.
Data information such as ordering, browsing a webpage, playing a game, watching a video, making a comment, or sharing a webpage by a user using the terminal devices 101, 102, 103 may be recorded, collected, or processed by the server 105 to obtain various feature variables describing the user. The server 105 may send push messages to the terminal devices 101, 102, 103 of the eligible users according to the triggering instructions.
It should be noted that the method for pushing information provided by the embodiment of the present disclosure may be generally performed by the server 105. Accordingly, the apparatus for pushing information provided by the embodiment of the present disclosure may be generally disposed in the server 105. The method for pushing information provided by the embodiment of the present disclosure may also be performed by a server or a server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, 103 and/or the server 105. Correspondingly, the apparatus for pushing information provided by the embodiment of the present disclosure may also be disposed in a server or a server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, 103 and/or the server 105.
It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
Fig. 2 schematically shows a flowchart of a method for pushing information according to an embodiment of the present disclosure.
As shown in fig. 2, the method for information push according to an embodiment of the present disclosure includes operations S201 to S203.
In operation S201, at least one valid feature variable of a first user is obtained, where the at least one valid feature variable belongs to R candidate feature variables that are most important for constructing a first classification model among N initial feature variables, where the first classification model is obtained by training a first set of training data, the first set of training data includes the N initial feature variables, N and R are integers, and N > R ≧ 2.
The at least one valid feature variable belongs to the R candidate feature variables, and for example, all of the R candidate feature variables may be used as the at least one valid feature variable, or for example, the at least one valid feature variable may also be obtained after the R candidate feature variables are subjected to further rule screening.
Then, in operation S202, a potential value of the first user becoming a potential user is obtained based on at least one valid feature variable of the first user and a second classification model, wherein the second classification model is obtained through a second set of training data, the second set of training data includes the R candidate feature variables, and the first classification model is different from the second classification model.
In operation S203, a message is pushed to the first user when the potential value satisfies a first condition. For example, when the potential value reaches a threshold, the first user is considered as a potential user, and a message is pushed to the first user. Or, for example, when the potential value of the first user is ordered within the number of information to be pushed in the potential values of all users, a message is pushed to the first user.
The first classification model and the second classification model may be, for example, logistic regression, random forest, xgboost, gbdt + lr, etc., as long as the first classification model and the second classification model are different classification models. Wherein, the R candidate characteristic variables are part of characteristic variables which have the most important influence on the construction of the first classification model. For the first classification model constructed in a different manner, the method of selecting the important feature variables thereof differs depending on the algorithm. For example, for the tree model, the classification purity can be screened based on the classification purity after a characteristic variable is introduced; alternatively, for regression models, the screening can be based on the residual magnitude of the regression fit after introducing a characteristic variable.
According to an embodiment of the present disclosure, the first classification model includes an xgboost (extreme gradientboosting) extreme gradient ascent model, and/or the second classification model includes a logistic regression model.
According to the embodiment of the disclosure, the R candidate feature variables are feature variables obtained through the xgb. xgboost is to iteratively build many decision trees, and each decision tree learns the error between the classification result and the true value of the previous decision tree. And each decision tree selects characteristic variables of child nodes according to a greedy algorithm when branching from a parent node. The import function may determine the importance of feature variables based on the ordering of the number of occurrences of each feature variable in all trees of the xgboost model.
According to the embodiment of the disclosure, when determining whether to push a message to a first user, a potential value of the first user is obtained according to at least one effective characteristic variable of the first user and a second classification model, and the potential value is sent when a first condition is met. And the at least one effective characteristic variable is obtained by multiple screening from N initial characteristic variables, and the second classification model is also obtained by training the screened R candidate characteristic variables based on the first classification model, so that the difference between different users can be enlarged by combining the first classification model and the second classification model and multiple screening of the characteristic variables. Thereby making pushing information more targeted. When the popularization and marketing are needed, the user probability scores with large differences are more beneficial to screening. Moreover, through multiple screening of characteristic variables and combination of different classification models, partial overfitting influence can be reduced, generalization capability of predicting potential users is improved, and conversion rate of pushed messages can be improved.
Fig. 3 schematically shows a flow chart of a method for information push according to another embodiment of the present disclosure.
As shown in fig. 3, the method for information push may include operations S301 to S303 in addition to operations S201 to S203 according to an embodiment of the present disclosure.
In operation S301, the first classification model is trained through the first set of training data.
Then, in operation S302, R candidate feature variables are screened out from the N initial feature variables based on the first classification model.
Next, in operation S303, the second classification model is trained by the second set of training data.
The following description takes the first set of training data exemplified in table 1 as an example. The data in table 1 is training data obtained for a commercial product targeted at an internet financial product (for example, a small white card in the east of kyoto).
In the target features of table 1, 1 indicates a user who opens a small white card, and 0 indicates a user who has received a push message but has not opened a small white card.
TABLE 1 first set of training data
Specifically, in the example of table 1, the N initial feature variables may be divided into seven broad categories. The method specifically comprises the following steps: 1. the consumption amount of each first-grade product of the user; the first-class items may include, for example, first-class categories of products such as foods, books, and electronic products. 2. The user part is the secondary product consumption amount. The secondary products can be, for example, books, financial periodicals, internet books, and the like, and electronic products such as mobile phones, computers, digital cameras, and the like. In practical application, because the secondary categories are various, part of the secondary categories can be screened, for example, the secondary categories which are considered to have high correlation with financial management behaviors of users can be screened, and the secondary categories with high universality are removed, so that redundant information in training data can be simplified. Subordinate secondary products such as foods, clothes, electrical appliances and the like can be eliminated, and subordinate secondary products such as books, digital products and the like which are possibly more clearly related to financial management of the user are reserved. 3. And browsing times of each primary commodity of the user. 4. And browsing times corresponding to the previously selected secondary categories. 5. The browsing times corresponding to each financial service line, such as insurance, fund, financing, crowd funding, etc. 6. Each financial business line is corresponding to the invested amount. 7. The amount of the white bar and the number of times of the white bar staging. The number of initial feature variables in the first set of training data exemplified in table 1 can often reach hundreds.
The number of samples in the first set of training data illustrated in table 1 is 57 ten thousand, and 57 ten thousand samples are sampled according to the ratio of the users who have opened cards to the users who have not opened cards in the actual users, wherein 55 ten thousand users do not open cards and twenty thousand users are users who have opened cards. For example, the users who have opened cards in table 1 are labeled a1, A2, a20000, and the users who have not opened cards are labeled B1, B2, a.
A first classification model (e.g., an xgboost model) is trained in operation S301 using a first set of training data shown in table 1.
According to the embodiment of the present disclosure, after the trained xgboost model is obtained, the probability result predicted by the xgboost model is not directly used, but a part of feature variables (i.e., R candidate feature variables) most important for constructing the xgboost model is screened from the N initial feature variables in the first set of training data of table 1 in operation S302. In particular, the xgb.
Then, in operation S303, a second classification model is trained by a second set of training data. In some embodiments, the R candidate feature variables may be sorted out for each sample in the first set of training data to obtain a corresponding second set of training data. In other embodiments, the second set of training data may be obtained from the population of users according to the ratio of the users who have opened cards to the users who have not opened cards according to the R candidate feature variables. After the second classification model is obtained, in the prediction phase, the potential value of the first user may be calculated in operation S203 according to the second classification model and the at least one valid characteristic variable of the first user.
Further, referring to fig. 3, the method for information push may further include operation S304 after operation S303. In operation S304, the at least one valid feature variable is determined based on the correlation between the output of the second classification model and each of the R candidate feature variables during the training process. According to the embodiment of the present disclosure, the R candidate feature variables may be further filtered to expect non-randomness with a larger probability between the at least one valid feature variable finally obtained and the output of the second classification model, so that the value of predicting the user behavior with the at least one valid feature variable may be improved.
For example, when the second classification model is a logistic regression model, the R candidate feature variables are further screened, and whether the p value of the regression coefficient of each feature variable in the logistic regression model satisfies a certain condition may be observed. Wherein the certain condition may be that the feature variable is determined to belong to the at least one valid feature variable when the non-randomness between each feature variable and the output of the logistic regression model satisfies a predetermined probability. The p-value of the regression coefficient (i.e., p value) can be used to verify whether the relationship between the feature variable and the target (the output of the logistic regression model) is a value of randomness. For example, a p-value of the regression coefficient may be calculated to represent the probability that the relationship between the feature variable and the target is random. If the relationship between the feature variable and the target is desired to be non-random in the case of 95% (for example only, and may be set according to actual conditions), then if p is greater than 5%, it can be considered that the relationship between the feature variable and the target has a large randomness, and the feature variable can be discarded.
FIG. 4 schematically illustrates a flow chart of a method of obtaining potential values through a logistic regression model, according to an embodiment of the disclosure.
As shown in fig. 4, when the second classification model is a logistic regression model, operation S202 may include operation S212 and operation S222.
In operation S212, a regression coefficient for each of the at least one valid characteristic variable in the logistic regression model is obtained.
In operation S222, the potential value is obtained based on the regression coefficient and at least one valid feature variable of the first user.
For example, assuming that the at least one valid characteristic variable has five valid characteristic variables, which are A, B, C, D, E respectively, and the regression coefficients corresponding to the five valid characteristic variables obtained in operation S212 are 0.3, 0.2, 0.1, 0.05, 0.02 respectively, the potential value calculation of the first user in operation S222 may be Q ═ 0.3 × a +0.2 × B +0.1 × C +0.05 × D +0.02 × E.
According to the embodiment of the disclosure, when the potential user is determined by information push, the potential value of the user can be calculated by using only the regression coefficient extracted from the logistic regression model, so that the effect of saving the operation time is achieved. Meanwhile, in daily operation, the coefficient of the logistic regression model can be updated only periodically to obtain a new regression coefficient of the characteristic variable, and the model does not need to be reused to predict all users again when user screening is carried out each time. And because only important characteristic variables are reserved, the regression coefficient of the characteristic variables is used for directly calculating the final result, so that the overfitting effect of the model is reduced, and meanwhile, the generalization capability of the model is improved.
According to the embodiment of the disclosure, when determining potential users for information push, important feature variables are selected from a large number of initial feature variables, then coefficients are added to the important feature variables through a logistic regression model, and then the important feature variables are weighted and calculated according to the coefficients, so that the effect of expanding potential value differences among users can be achieved, and the potential users can be conveniently screened compared with results directly predicted through a single classification model.
Fig. 5A schematically illustrates an application example flowchart of a method for information push according to an embodiment of the present disclosure. FIG. 5B illustrates a conceptual diagram of the training process in the application example method of FIG. 5A.
The following description will be given of the application example with the first classification model being the xgboost model and the second classification model being the logistic regression model in conjunction with fig. 5A and 5B. The application example comprises the following steps:
(1) selecting N initial characteristic variables, integrating a first group of training data, and inputting the first group of training data into an xgboost model to train the xgboost model;
(2) after the training of the xgboost model is finished, selecting a most important part of feature variables (namely R candidate feature variables) in the xgboost model through the xgb.
(3) Combining the first group of training data and the R candidate characteristic variables to obtain a second group of training data, inputting the second group of training data into a logistic regression model, and calculating to obtain a regression coefficient corresponding to each characteristic variable;
(4) removing part of characteristic variables according to the corresponding p value of each characteristic variable in the logistic regression model;
(5) for example, the at least one valid feature variable has five (A, B, C, D, E, respectively) and the regression coefficients obtained in operation S212 for the five valid feature variables are β 1, β 2, β 3, β 4, and β 5 (as shown in fig. 5B), then the potential value calculation formula for each user may be Q β 1 a + β 2B + β 3C + β 4D + β 5E;
(6) and sorting the potential values of the users from big to small, and selecting the users from front to back for information push according to the quantity of the information required to be pushed.
Fig. 6 schematically shows a block diagram of an apparatus 600 for information push according to an embodiment of the present disclosure.
As shown in fig. 6, the apparatus 600 may include a user characteristic obtaining module 610, a potential value calculating module 620, and a pushing module 630 according to an embodiment of the present disclosure. The apparatus 600 may be used to implement a method according to an embodiment of the present disclosure.
The user feature obtaining module 610 may, for example, perform operation S201, and is configured to obtain at least one valid feature variable of a first user, where the at least one valid feature variable belongs to R candidate feature variables, which are most important for constructing a first classification model, of N initial feature variables, where the first classification model is obtained through training of a first set of training data, the first set of training data includes the N initial feature variables, N and R are integers, and N > R ≧ 2.
The potential value calculation module 620 may perform operation S202, for example, to predict potential values of the first user becoming a potential user based on at least one valid feature variable of the first user and a second classification model, wherein the second classification model is obtained through a second set of training data, and the second set of training data includes the R candidate feature variables. Wherein the first classification model is different from the second classification model.
The pushing module 630 may perform operation S203, for example, to push a message to the first user when the potential value satisfies a first condition.
The apparatus 600 further includes a first training module 640, a first screening module 650, and a second training module 660 according to embodiments of the present disclosure.
The first training module 640 may, for example, perform operation S301 for training the first classification model with the first set of training data.
The first screening module 650 may, for example, perform operation S302 for screening the R candidate feature variables from the N initial feature variables based on the first classification model.
The second training module 660 may perform operation S303, for example, to train the second classification model through the second set of training data.
The apparatus 600 may also include a second screening module 670, according to embodiments of the present disclosure. The second screening module 670 may, for example, perform operation S304 for determining the at least one valid feature variable based on a correlation of an output of the second classification model and each of the R candidate feature variables during the training process.
According to an embodiment of the present disclosure, the first classification model comprises an xgboost extreme gradient ascent model, and/or the second classification model comprises a logistic regression model.
According to an embodiment of the present disclosure, the R candidate feature variables are feature variables obtained through the xgb.
According to an embodiment of the disclosure, the latent value calculating module 620 may be specifically configured to perform operations S212 and S222, including obtaining a regression coefficient of each of the at least one valid feature variable in the logistic regression model, and obtaining the latent value based on the at least one valid feature variable of the first user and the regression coefficient.
Any number of modules, sub-modules, units, sub-units, or at least part of the functionality of any number thereof according to embodiments of the present disclosure may be implemented in one module. Any one or more of the modules, sub-modules, units, and sub-units according to the embodiments of the present disclosure may be implemented by being split into a plurality of modules. Any one or more of the modules, sub-modules, units, sub-units according to embodiments of the present disclosure may be implemented at least in part as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented in any other reasonable manner of hardware or firmware by integrating or packaging a circuit, or in any one of or a suitable combination of software, hardware, and firmware implementations. Alternatively, one or more of the modules, sub-modules, units, sub-units according to embodiments of the disclosure may be at least partially implemented as a computer program module, which when executed may perform the corresponding functions.
For example, any of the user characteristic obtaining module 610, the potential value calculating module 620, the pushing module 630, the first training module 640, the first screening module 650, the second training module 660, and the second screening module 670 may be combined into one module to be implemented, or any one of them may be split into a plurality of modules. Alternatively, at least part of the functionality of one or more of these modules may be combined with at least part of the functionality of the other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the user characteristic obtaining module 610, the potential value calculating module 620, the pushing module 630, the first training module 640, the first screening module 650, the second training module 660, and the second screening module 670 may be at least partially implemented as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented by hardware or firmware in any other reasonable manner of integrating or packaging a circuit, or implemented by any one of three implementations of software, hardware, and firmware, or by a suitable combination of any of them. Alternatively, at least one of the user characteristic acquisition module 610, the potential value calculation module 620, the push module 630, the first training module 640, the first filtering module 650, the second training module 660, and the second filtering module 670 may be at least partially implemented as a computer program module that, when executed, may perform corresponding functions.
FIG. 7 schematically shows a block diagram for an information-pushing computer system 700, according to an embodiment of the disclosure. The computer system 700 shown in fig. 7 is only an example and should not bring any limitations to the functionality or scope of use of the embodiments of the present disclosure.
As shown in fig. 7, a computer system 700 according to an embodiment of the present disclosure includes a processor 701, which can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)702 or a program loaded from a storage section 708 into a Random Access Memory (RAM) 703. The processor 701 may include, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor and/or associated chipset, and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), among others. The processor 701 may also include on-board memory for caching purposes. The processor 701 may comprise a single processing unit or a plurality of processing units for performing the different actions of the method flows according to embodiments of the present disclosure.
In the RAM 703, various programs and data necessary for the operation of the system 700 are stored. The processor 701, the ROM702, and the RAM 703 are connected to each other by a bus 704. The processor 701 performs various operations of the method flows according to the embodiments of the present disclosure by executing programs in the ROM702 and/or the RAM 703. Note that the program may also be stored in one or more memories other than the ROM702 and the RAM 703. The processor 701 may also perform various operations of method flows according to embodiments of the present disclosure by executing programs stored in the one or more memories.
According to an embodiment of the present disclosure, the system 700 may also include an input/output (I/O) interface 705, the input/output (I/O) interface 705 also being connected to the bus 704. The system 700 may also include one or more of the following components connected to the I/O interface 705: an input portion 706 including a keyboard, a mouse, and the like; an output section 707 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage section 708 including a hard disk and the like; and a communication section 709 including a network interface card such as a LAN card, a modem, or the like. The communication section 709 performs communication processing via a network such as the internet. A drive 710 is also connected to the I/O interface 705 as needed. A removable medium 711 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 710 as necessary, so that a computer program read out therefrom is mounted into the storage section 708 as necessary.
According to embodiments of the present disclosure, method flows according to embodiments of the present disclosure may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable storage medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication section 709, and/or installed from the removable medium 711. The computer program, when executed by the processor 701, performs the above-described functions defined in the system of the embodiment of the present disclosure. The systems, devices, apparatuses, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the present disclosure.
The present disclosure also provides a computer-readable storage medium, which may be contained in the apparatus/device/system described in the above embodiments; or may exist separately and not be assembled into the device/apparatus/system. The computer-readable storage medium carries one or more programs which, when executed, implement the method according to an embodiment of the disclosure.
According to embodiments of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, which may include, for example but is not limited to: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. For example, according to embodiments of the present disclosure, a computer-readable storage medium may include the ROM702 and/or the RAM 703 and/or one or more memories other than the ROM702 and the RAM 703 described above.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
Those skilled in the art will appreciate that various combinations and/or combinations of features recited in the various embodiments and/or claims of the present disclosure can be made, even if such combinations or combinations are not expressly recited in the present disclosure. In particular, various combinations and/or combinations of the features recited in the various embodiments and/or claims of the present disclosure may be made without departing from the spirit or teaching of the present disclosure. All such combinations and/or associations are within the scope of the present disclosure.
The embodiments of the present disclosure have been described above. However, these examples are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although the embodiments are described separately above, this does not mean that the measures in the embodiments cannot be used in advantageous combination. The scope of the disclosure is defined by the appended claims and equivalents thereof. Various alternatives and modifications can be devised by those skilled in the art without departing from the scope of the present disclosure, and such alternatives and modifications are intended to be within the scope of the present disclosure.
Claims (14)
1. A method for information push, comprising:
obtaining at least one effective characteristic variable of a first user, wherein the at least one effective characteristic variable belongs to R candidate characteristic variables which have the most important influence on the construction of a first classification model in N initial characteristic variables, the first classification model is obtained through training of a first group of training data, the first group of training data comprises the N initial characteristic variables, N and R are integers, and N is more than R and is more than or equal to 2;
obtaining potential values of the first user becoming potential users based on at least one valid feature variable of the first user and a second classification model, wherein the second classification model is obtained through a second set of training data, and the second set of training data comprises the R candidate feature variables; and
when the potential value meets a first condition, pushing a message to the first user;
wherein:
the first classification model is different from the second classification model.
2. The method of claim 1, further comprising:
training the first classification model with the first set of training data;
screening the R candidate characteristic variables from the N initial characteristic variables based on the first classification model;
training the second classification model with the second set of training data.
3. The method of claim 2, further comprising:
and determining the at least one effective characteristic variable based on the correlation between the output of the second classification model and each of the R candidate characteristic variables in the training process.
4. The method according to any of claims 1 to 3, wherein the first classification model comprises an xgboost extreme gradient model and/or the second classification model comprises a logistic regression model.
5. The method of claim 4, wherein the R candidate feature variables are feature variables obtained by an xgb.
6. The method of claim 4, wherein obtaining potential values for the first user to become potential users based on at least one valid feature variable of the first user and a second classification model comprises:
obtaining a regression coefficient of each characteristic variable in the at least one effective characteristic variable in the logistic regression model; and
obtaining the potential value based on the at least one valid feature variable of the first user and the regression coefficient.
7. An apparatus for information push, comprising:
the user characteristic obtaining module is used for obtaining at least one effective characteristic variable of a first user, wherein the at least one effective characteristic variable belongs to R candidate characteristic variables which have the most important influence on the construction of a first classification model in N initial characteristic variables, the first classification model is obtained through training of a first group of training data, the first group of training data comprises the N initial characteristic variables, N and R are integers, and N is more than R and is more than or equal to 2;
a potential value calculation module for predicting potential values of the first user to become potential users based on at least one valid feature variable of the first user and a second classification model, wherein the second classification model is obtained through a second set of training data, and the second set of training data comprises the R candidate feature variables; and
a pushing module, configured to push a message to the first user when the potential value satisfies a first condition:
wherein:
the first classification model is different from the second classification model.
8. The apparatus of claim 7, further comprising:
a first training module to train the first classification model with the first set of training data;
a first screening module, configured to screen the R candidate feature variables from the N initial feature variables based on the first classification model;
a second training module to train the second classification model with the second set of training data.
9. The apparatus of claim 8, further comprising:
and the second screening module is used for determining the at least one effective characteristic variable based on the correlation between the output of the second classification model and each characteristic variable in the R candidate characteristic variables in the training process.
10. The apparatus of any of claims 7 to 9, wherein the first classification model comprises an xgboost extreme gradient model and/or the second classification model comprises a logistic regression model.
11. The apparatus according to claim 10, wherein the R candidate feature variables are feature variables obtained by filtering the xgb.
12. The apparatus of claim 10, wherein the potential value calculation module is specifically configured to:
obtaining a regression coefficient of each characteristic variable in the at least one effective characteristic variable in the logistic regression model; and
obtaining the potential value based on the at least one valid feature variable of the first user and the regression coefficient.
13. A system for pushing information comprises one or more memories storing executable instructions; and
one or more processors executing the executable instructions to implement the method of any one of claims 1 to 6.
14. A computer readable storage medium having stored thereon executable instructions which, when executed by a processor, cause the processor to perform the method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811201106.2A CN111047343A (en) | 2018-10-15 | 2018-10-15 | Method, device, system and medium for information push |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811201106.2A CN111047343A (en) | 2018-10-15 | 2018-10-15 | Method, device, system and medium for information push |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111047343A true CN111047343A (en) | 2020-04-21 |
Family
ID=70230453
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811201106.2A Pending CN111047343A (en) | 2018-10-15 | 2018-10-15 | Method, device, system and medium for information push |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111047343A (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001020512A2 (en) * | 1999-09-15 | 2001-03-22 | General Electric Company | Method for modeling market response rates |
| US20110004509A1 (en) * | 2009-07-06 | 2011-01-06 | Xiaoyuan Wu | Systems and methods for predicting sales of item listings |
| CN104598774A (en) * | 2015-02-04 | 2015-05-06 | 河南师范大学 | Feature gene selection method based on logistic and relevant information entropy |
| US20150310336A1 (en) * | 2014-04-29 | 2015-10-29 | Wise Athena Inc. | Predicting customer churn in a telecommunications network environment |
| JP2016031629A (en) * | 2014-07-29 | 2016-03-07 | 日本電信電話株式会社 | Feature selection device, feature selection system, feature selection method and feature selection program |
| US20160110794A1 (en) * | 2014-10-20 | 2016-04-21 | Yahoo! Inc. | E-commerce recommendation system and method |
| CN105678570A (en) * | 2015-12-31 | 2016-06-15 | 北京京东尚科信息技术有限公司 | Method and apparatus for identifying potential users of E-commerce |
| CN106127546A (en) * | 2016-06-20 | 2016-11-16 | 重庆房慧科技有限公司 | A kind of Method of Commodity Recommendation based on the big data in intelligence community |
| CN106649774A (en) * | 2016-12-27 | 2017-05-10 | 北京百度网讯科技有限公司 | Artificial intelligence-based object pushing method and apparatus |
| CN106846041A (en) * | 2016-12-26 | 2017-06-13 | 携程计算机技术（上海）有限公司 | The distribution method and system of reward voucher |
| US20170262770A1 (en) * | 2016-03-11 | 2017-09-14 | Uber Technologies, Inc. | Cascaded boosted predictive models |
| CN107895283A (en) * | 2017-11-07 | 2018-04-10 | 重庆邮电大学 | A kind of businessman's volume of the flow of passengers big data Forecasting Methodology based on Time Series |
-
2018
- 2018-10-15 CN CN201811201106.2A patent/CN111047343A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2001020512A2 (en) * | 1999-09-15 | 2001-03-22 | General Electric Company | Method for modeling market response rates |
| US20110004509A1 (en) * | 2009-07-06 | 2011-01-06 | Xiaoyuan Wu | Systems and methods for predicting sales of item listings |
| US20150310336A1 (en) * | 2014-04-29 | 2015-10-29 | Wise Athena Inc. | Predicting customer churn in a telecommunications network environment |
| JP2016031629A (en) * | 2014-07-29 | 2016-03-07 | 日本電信電話株式会社 | Feature selection device, feature selection system, feature selection method and feature selection program |
| US20160110794A1 (en) * | 2014-10-20 | 2016-04-21 | Yahoo! Inc. | E-commerce recommendation system and method |
| CN104598774A (en) * | 2015-02-04 | 2015-05-06 | 河南师范大学 | Feature gene selection method based on logistic and relevant information entropy |
| CN105678570A (en) * | 2015-12-31 | 2016-06-15 | 北京京东尚科信息技术有限公司 | Method and apparatus for identifying potential users of E-commerce |
| US20170262770A1 (en) * | 2016-03-11 | 2017-09-14 | Uber Technologies, Inc. | Cascaded boosted predictive models |
| CN106127546A (en) * | 2016-06-20 | 2016-11-16 | 重庆房慧科技有限公司 | A kind of Method of Commodity Recommendation based on the big data in intelligence community |
| CN106846041A (en) * | 2016-12-26 | 2017-06-13 | 携程计算机技术（上海）有限公司 | The distribution method and system of reward voucher |
| CN106649774A (en) * | 2016-12-27 | 2017-05-10 | 北京百度网讯科技有限公司 | Artificial intelligence-based object pushing method and apparatus |
| CN107895283A (en) * | 2017-11-07 | 2018-04-10 | 重庆邮电大学 | A kind of businessman's volume of the flow of passengers big data Forecasting Methodology based on Time Series |
Non-Patent Citations (1)
| Title |
|---|
| 张凌: "《能时代的银行知识管理》", 31 May 2018, 武汉大学出版社, pages: 283 - 284 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20220351016A1 (en) | Presentation module for webinterface production and deployment system | |
| Henry et al. | Emergence of segregation in evolving social networks | |
| US11087403B2 (en) | Risk quantification for insurance process management employing an advanced decision platform | |
| US20210314234A1 (en) | Systems and methods for prediction of anomalies | |
| US10402906B2 (en) | Quantification for investment vehicle management employing an advanced decision platform | |
| US20210385251A1 (en) | System and methods for integrating datasets and automating transformation workflows using a distributed computational graph | |
| AU2014287551A1 (en) | Network-aware product rollout in online social networks | |
| US20210136122A1 (en) | Crowdsourced innovation laboratory and process implementation system | |
| US11546380B2 (en) | System and method for creation and implementation of data processing workflows using a distributed computational graph | |
| CN107908662B (en) | Method and device for realizing search system | |
| CN113159934A (en) | Method and system for predicting passenger flow of network, electronic equipment and storage medium | |
| US20190294645A1 (en) | Systems and methods for optimizing the selection and display of electronic content | |
| US20210349920A1 (en) | Method and apparatus for outputting information | |
| CN112749323A (en) | Method and device for constructing user portrait | |
| CN109978594B (en) | Order processing method, device and medium | |
| CN114297278A (en) | Method, system and device for quickly writing batch data | |
| CN113761350A (en) | Data recommendation method, related device and data recommendation system | |
| US20180130077A1 (en) | Automated selection and processing of financial models | |
| CN111047343A (en) | Method, device, system and medium for information push | |
| CN113391988A (en) | Method and device for losing user retention, electronic equipment and storage medium | |
| CN112860999A (en) | Information recommendation method, device, equipment and storage medium | |
| CN112799797A (en) | Task management method and device | |
| CN112016790A (en) | User policy distribution method and device and electronic equipment | |
| CN113570467B (en) | Method and device for pushing special resource sharing information and electronic equipment | |
| CN113360765B (en) | Event information processing method and device, electronic equipment and medium |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: Room 221, 2 / F, block C, 18 Kechuang 11th Street, Daxing District, Beijing, 100176 Applicant after: Jingdong Technology Holding Co.,Ltd. Address before: Room 221, 2 / F, block C, 18 Kechuang 11th Street, Daxing District, Beijing, 100176 Applicant before: JINGDONG DIGITAL TECHNOLOGY HOLDINGS Co.,Ltd. |