CN110070399A - A kind of discount coupon method for pushing and device - Google Patents
A kind of discount coupon method for pushing and device Download PDFInfo
- Publication number
- CN110070399A CN110070399A CN201910335996.4A CN201910335996A CN110070399A CN 110070399 A CN110070399 A CN 110070399A CN 201910335996 A CN201910335996 A CN 201910335996A CN 110070399 A CN110070399 A CN 110070399A
- Authority
- CN
- China
- Prior art keywords
- discount coupon
- user
- pushed
- probability
- target user
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000012549 training Methods 0.000 claims description 177
- 230000006399 behavior Effects 0.000 claims description 69
- 238000003066 decision tree Methods 0.000 claims description 35
- 238000004590 computer program Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 13
- 238000012417 linear regression Methods 0.000 claims description 11
- 230000008901 benefit Effects 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000011156 evaluation Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 description 16
- 238000010586 diagram Methods 0.000 description 13
- 230000006870 function Effects 0.000 description 13
- 230000006872 improvement Effects 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 3
- 230000014759 maintenance of location Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000003780 insertion Methods 0.000 description 2
- 230000037431 insertion Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 229920001296 polysiloxane Polymers 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 239000010979 ruby Substances 0.000 description 1
- 229910001750 ruby Inorganic materials 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- 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
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0219—Discounts or incentives, e.g. coupons or rebates based on funds or budget
-
- 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
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0222—During e-commerce, i.e. online transactions
-
- 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
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0224—Discounts or incentives, e.g. coupons or rebates based on user history
-
- 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
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0239—Online discounts or incentives
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Game Theory and Decision Science (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Subject description discloses a kind of discount coupon method for pushing and devices, first according to the historical behavior data of target user, determine the user characteristics of target user, later for each discount coupon to be pushed, according to the discount coupon feature of the user characteristics of target user and the discount coupon to be pushed, pass through model trained in advance, under the different service conditions to the discount coupon to be pushed, the probability of target user execution buying behavior, then according to the cost for the probability and the push discount coupon determined, determine the return value that the discount coupon to be pushed is pushed to target user, finally, according to the return value of respectively discount coupon to be pushed, from respectively wait push determining targeted coupon in discount coupon and be pushed to target user.
Description
Technical field
This application involves Internet technical field more particularly to a kind of discount coupon method for pushing and device.
Background technique
With the development of internet technology and mature, more and more users' habit is consumed by e-commerce platform to be purchased
Object.And largely occur in e-commerce platform and under homogeneous serious situation, only enough any active ues just can guarantee
The normal operation of platform, therefore how to improve important process one of of user's retention ratio as each e-commerce platform.And it is a kind of
The common method for improving user's retention ratio is: to its interested information of user's pushed information, using platform to increase user
Frequency, improve retention ratio.Wherein, pushed information includes: advertisement, discount coupon, interactive event, electronic ticket etc..
Since user usually only browses oneself interested information, and the demand of different user is not fully identical, therefore
In order to improve the accuracy rate of pushed information, in the prior art, by obtaining the portrait information of user, according to the use of training completion
In the prediction model of prediction user preferences, the information of true directional user's push.
Also, since electronic ticket or this category information of discount coupon are more obvious for the effect for attracting user to retain,
More common prediction model is for predicting that user is complete in different expense deduction and exemption value (that is, preferential amounts that discount coupon provides)
At the probability of default transaction.Which kind of discount coupon what then last platform determined pushes to user.
Function in the prior art using each expense deduction and exemption value of fitting and corresponding each transaction probability, based on quasi-
Function and preset several parameters are closed, if user group is divided into Ganlei, and determines the corresponding push of every a kind of user
Discount coupon (that is, which kind of discount coupon pushed).But the granularity of prediction model is larger, cannot really accomplish to be directed to each user
Prediction and discount coupon push are carried out, causes the accuracy rate for pushing discount coupon lower.In addition, due to the cost of discount coupon push,
Discount coupon itself is further comprised in the expense deduction and exemption value used.But in the prior art, there is no consider discount coupon actual use
When caused by cost, be easy to cause push discount coupon cost it is excessively high, put the cart before the horse.
Then aiming at the problems existing in the prior art, this specification provides a kind of new discount coupon method for pushing and dress
It sets.
Summary of the invention
This specification embodiment provides a kind of discount coupon method for pushing and device, for solving the skill of existing push discount coupon
Cost is not considered in art, and the granularity of prediction model is larger, cause to push the low problem of discount coupon accuracy.
This specification embodiment adopts the following technical solutions:
A kind of discount coupon method for pushing that this specification provides, comprising:
According to the historical behavior data of target user, the user characteristics of the target user are determined；
For respectively discount coupon to be pushed, according to the user characteristics of the target user and it is somebody's turn to do the preferential of discount coupon to be pushed
Certificate feature determines the target user under the different service conditions of discount coupon to be pushed to this by model trained in advance
Execute the probability of buying behavior；
According to the cost of the probability and the discount coupon to be pushed, determines and the discount coupon to be pushed is pushed to the mesh
Mark the return value of user；
According to for return value that respectively discount coupon to be pushed is determined respectively, from respectively wait push in discount coupon, determining target
Discount coupon is simultaneously pushed to the target user.
Optionally, preparatory training pattern, specifically includes:
According to the user information of each user, several user types are determined；
Training is determined according to each discount coupon and each user of the user type for each user type determined
Sample, and training is for predicting that the user of the user type executes the general of buying behavior under the different service conditions to discount coupon
The model of rate；
Wherein, for each training sample, which includes a user and a discount coupon, if the training sample
In include user hold the discount coupon for including in the training sample, then the training sample be positive example, if being wrapped in the training sample
The user contained does not hold the discount coupon for including in the training sample, then the training sample is counter-example.
Optionally, preparatory training pattern, specifically includes:
Purchase row is executed according to the counter-example in the training sample, when training is obtained for predicting that user does not hold discount coupon
For probability the first model；
According to the positive example and the first probability for being calculated by first model of each positive example in the training sample,
The second model of the probability of buying behavior is executed when training is obtained for predicting that user holds discount coupon；
According to positive example in the training sample, training is obtained for predicting that user executes the general of buying behavior using discount coupon
The third model of rate.
Optionally, according to the counter-example in the training sample, training is held when obtaining for predicting that user does not hold discount coupon
First model of the probability of row buying behavior, specifically includes:
According to the counter-example and grad enhancement decision tree GBDT in the training sample, training is for handling user characteristics
GBDT decision-tree model；
According to the GBDT decision-tree model that training is completed, the user characteristics of each user in counter-example are handled, and by processing result
As the input of linear regression LR model, the probability of buying behavior is executed when with training for predicting that user does not hold discount coupon
LR model, as first model.
Optionally, according in the training sample positive example and each positive example be calculated by first model
One probability executes the second model of the probability of buying behavior, specifically includes when training is obtained for predicting that user holds discount coupon:
For each positive example in the training sample, is calculated in the positive example and wrapped according to first model that training is completed
The user contained executes the probability of buying behavior, the first probability as positive example when not holding discount coupon；
According to the first probability of positive example and positive example in the training sample, training is for predicting that user holds discount coupon
The LR model of the probability of Shi Zhihang buying behavior, as second model.
Optionally, according to positive example in the training sample, training is obtained for predicting that user executes purchase using discount coupon
The third model of the probability of behavior, specifically includes:
According to the positive example and grad enhancement decision tree GBDT in the training sample, training for handle user characteristics with
And the GBDT decision-tree model of discount coupon feature；
According to the GBDT decision-tree model that training is completed, the excellent of the user characteristics and discount coupon of user in each positive example is handled
Favour certificate feature, and using processing result as the input of linear regression LR model, with training for predicting that user is held using discount coupon
The LR model of the probability of row buying behavior, as the third model.
Optionally, according to the discount coupon feature of the user characteristics of the target user and the discount coupon to be pushed, pass through
Trained model in advance determines that the target user executes purchase row under the different service conditions to the discount coupon to be pushed
For probability, specifically include:
According to the user information of the target user, the corresponding user type of the target user is determined；
According to the user characteristics of the target user, by being directed to the corresponding user type training of the target user in advance
The first model prediction described in target user the first probability；
According to the discount coupon feature of the user characteristics of the target user and the discount coupon to be pushed, by being directed in advance
The second probability of target user described in second model prediction of the corresponding user type training of the target user；
According to the discount coupon feature of the user characteristics of the target user and the discount coupon to be pushed, by being directed in advance
The third probability of target user described in the third model prediction of the corresponding user type training of the target user.
Optionally, it according to the cost of the probability and the discount coupon to be pushed, determines and pushes the discount coupon to be pushed
To the return value of the target user, specifically include:
According to the difference and the corresponding user type of the target user of second probability and first probability
Number of users determines the income that the discount coupon to be pushed is pushed to the target user；
According to the number of users of the third probability, first probability, the corresponding user type of the target user with
And it is somebody's turn to do the preferential cost of discount coupon to be pushed, determine the push cost that the discount coupon to be pushed is pushed to the target user；
According to the income and the push cost, determines and the discount coupon to be pushed is pushed to returning for the target user
Report value.
Optionally, according to for return value that respectively discount coupon to be pushed is determined respectively, from respectively wait push in discount coupon, really
It sets the goal and discount coupon and is pushed to the target user, specifically include:
It is pushed to the sequence of the return value of the target user from big to small according to the discount coupon respectively to be pushed determined, really
The discount coupon to be pushed for determining specified quantity is pushed to the target user as targeted coupon.
This specification provides a kind of push coupon device, comprising:
User characteristics determining module determines that the user of the target user is special according to the historical behavior data of target user
Sign；
Probability evaluation entity according to the user characteristics of the target user and is somebody's turn to do wait push away for respectively discount coupon to be pushed
The discount coupon feature for sending discount coupon determines the target user to the discount coupon to be pushed by model trained in advance
The probability of buying behavior is executed under different service conditions；
Return computing module, according to the probability and should discount coupon be pushed cost, determine by this wait pushing it is preferential
Certificate is pushed to the return value of the target user；
Pushing module, according to each return value determined, from respectively wait push in discount coupon, determining targeted coupon and push
To the target user.
A kind of computer readable storage medium that this specification provides, which is characterized in that the storage medium is stored with meter
Calculation machine program, the computer program realize above-mentioned discount coupon method for pushing when being executed by processor.
The a kind of electronic equipment that this specification provides, including memory, processor and storage on a memory and can located
The computer program run on reason device, which is characterized in that the processor realizes above-mentioned discount coupon push when executing described program
Method.
This specification embodiment use at least one above-mentioned technical solution can reach it is following the utility model has the advantages that
Determining that the first historical behavior data according to target user determine that target is used when pushing discount coupon to target user
The user characteristics at family according to the user characteristics of target user and are somebody's turn to do preferential wait push later for each discount coupon to be pushed
The discount coupon feature of certificate passes through model trained in advance, under the different service conditions to the discount coupon to be pushed, the target
User executes the probability of buying behavior, then according to the cost for the probability and the push discount coupon determined, true directional aim
User pushes the return value of the discount coupon to be pushed, finally, according to the return value of respectively discount coupon to be pushed, from respectively preferential wait push
Targeted coupon is determined in certificate and is pushed to target user.Since return value is the cost determination based on discount coupon to be pushed,
It is thus determined that having the influence of cost during being pushed to the targeted coupon of target user, and each return value is all based on the mesh
What the feature calculation of mark user obtained, therefore it is pushed to the targeted coupon of target user, it is true individually for the target user
Fixed.It solves problems of the prior art, improves the accuracy of push discount coupon.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present application, constitutes part of this application, this Shen
Illustrative embodiments and their description please are not constituted an undue limitation on the present application for explaining the application.In the accompanying drawings:
Fig. 1 is that a kind of discount coupon that this specification provides pushes process；
Fig. 2 is the schematic diagram for the first probability of output that this specification provides；
Fig. 3 is a kind of structural schematic diagram for discount coupon driving means that this specification embodiment provides；
Fig. 4 is the electronic equipment schematic diagram corresponding to Fig. 1 that this specification embodiment provides.
Specific embodiment
To keep the purposes, technical schemes and advantages of this specification clearer, it is embodied below in conjunction with this specification
Technical scheme is clearly and completely described in example and corresponding attached drawing.Obviously, described embodiment is only this Shen
Please a part of the embodiment, instead of all the embodiments.The embodiment of base in this manual, those of ordinary skill in the art exist
Every other embodiment obtained under the premise of creative work is not made, shall fall in the protection scope of this application.
Below in conjunction with attached drawing, the technical scheme provided by various embodiments of the present application will be described in detail.
Fig. 1 is a kind of push discount coupon process that this specification embodiment provides, and specifically be can comprise the following steps that
S102: according to the historical behavior data of target user, the user characteristics of the target user are determined.
In the present specification, the server that specifically can be e-commerce platform for pushing discount coupon, when server determines
It needs to when some user of platform login account pushes discount coupon, which is target user.Then, server is for needle
Discount coupon is more accurately pushed to the target user, server can first obtain the historical behavior data of the target user, further according to
Preset user characteristics dimension and the historical behavior data of acquisition, determine the characteristic value of each user characteristics of the target user,
As the basis for executing subsequent step.
Specifically, the user's history behavioral data includes access record, transaction record, upload of the user in the platform
User's original content (User Generated Content, UGC), user click advertisement behavior using discount coupon behavior, user
Etc. data.Certainly, the business tine and service content being related to due to different e-commerce platforms are not quite identical, at this
Which data is the specification historical behavior data specifically include, and this specification without limitation, can be set as needed.
Similarly, user characteristics dimension preset in the present specification can be set as needed, then server can correspond to each use
Family characteristic dimension obtains the historical behavior data needed, which this specification which specifically has without limitation, can basis
It needs to be arranged.
For example, preset user characteristics dimension 1 are as follows: daily user accesses the number of the A class commodity page, then is visited according to user
It asks record, can determine the characteristic value of the user characteristics 1 of the target user.
S104: for respectively discount coupon to be pushed, according to the user characteristics of the target user and it is somebody's turn to do discount coupon to be pushed
Discount coupon feature determine that the target user makes in the difference to the discount coupon to be pushed by model trained in advance
The probability of buying behavior is executed in the case of.
In this specification embodiment, server is being sent to the target user for determining discount coupon each to be pushed
Afterwards, the data such as validity and the income of discount coupon to be pushed are somebody's turn to do, it can be by model trained in advance, for each excellent wait push
Favour certificate, which calculates separately, determines that the target user under the different service conditions of discount coupon to be pushed, executes the probability of buying behavior,
Which, so that discount coupon to be pushed selected according to the determine the probability determined in the next steps, pushed as targeted coupon.
In the present specification, the different service conditions for being somebody's turn to do discount coupon to be pushed include at least: holding and using discount coupon
Situation holds discount coupon and unused situation and one of the case where do not hold discount coupon.
Also, it is somebody's turn to do the discount coupon feature of discount coupon to be pushed, it is similar with preset user characteristics, it can match as needed
It sets, this specification is with no restrictions.The discount coupon feature can include: the usage mode of discount coupon is (for example, completely subtracting mode, limiting quotient
Product usage mode shares red packet mode etc.) and discount coupon be provided to the preferential amount of money of user (e.g., user use preferential
The amount of money of certificate rear platform actual expenses or the amount of money of profit reduction).
Specifically, the server can be according to the use of the target user in the present specification for each discount coupon to be pushed
Family feature, the first model completed by training calculate and determine the target user the case where not holding the discount coupon to be pushed
The lower probability for executing buying behavior, as first probability for being somebody's turn to do discount coupon to be pushed.First probability can be considered as this
Target user buys the underlying probabilities of commodity, that is, regardless of whether receiving the discount coupon of push, which can all disappear
The probability taken.
According to the discount coupon feature of the user characteristics of the target user and the discount coupon to be pushed, pass through what training was completed
Second model calculates and determines that the target user in the probability for holding this when pushing discount coupon and executing buying behavior, waits for as this
Push the second probability of discount coupon.
According to the discount coupon feature of the user characteristics of the target user and the discount coupon to be pushed, pass through what training was completed
Third model calculates the probability for determining that the target user executes buying behavior using the discount coupon to be pushed, waits pushing as this
The third probability of discount coupon.The third probability can be considered as the probability that the discount coupon conversion to be pushed is buying behavior.
In addition, server, can be according to each in each model of training in order to which the model for completing the upper training is more acurrate
The user information of user determines several user types.And for each user type point, model is respectively trained.It is, root
According to user information, training sample is divided into several groups, the corresponding one group of training sample of each user type, and use the corresponding use
The training sample of family type, the corresponding each model of the training user type.
Wherein, for each training sample, which includes a user and a discount coupon, if the training sample
In include user hold the discount coupon for including in the training sample, then the training sample be positive example, if being wrapped in the training sample
The user contained does not hold the discount coupon for including in the training sample, then the training sample is counter-example.
Further, for the server in the first model of training and third model, can be used will be terraced in the present specification
Algorithm is as prime for degree enhancing decision tree (Gradient Boosting Decision Tree, GBDT), by linear regression
(Linear Regression, LR) algorithm is as rear class, and prime is for handling feature, after rear class is handled according to prime
Feature export prediction result, be respectively trained to obtain the first model and third model.Wherein, training when the first model of training
Sample uses counter-example, and due to it is trained be for predicting that user does not hold discount coupon when execute the probability of buying behavior, because
Discount coupon feature is not involved in training in this training sample, that is, not using discount coupon feature as section when the building of GBDT decision tree
Point.And training sample when third model is trained to use user characteristics and discount coupon when GBDT decision tree constructs using positive example
Feature is as node.
The progress of linear regression (Linear Regression, LR) algorithm can be used in the second model of training in the server
Training.Also, when using the first probability of positive example and positive example training LR model, the first probability of the positive example, to pass through instruction
Practice when the user for including in the calculated each positive example of first model completed does not hold discount coupon and executes the general of buying behavior
Rate.
For the first probability for calculating the discount coupon to be pushed, if server can train completion according to GBDT is passed through
Dry GBDT decision-tree model, is encoded (that is, handling each user characteristics) to the feature of the target user, determines target
The feature of user combines, then combines this feature determined as input, exports the first probability, such as Fig. 2 by the first model
It is shown.
Fig. 2 is the schematic diagram for the first probability of output that this specification provides, wherein the user characteristics inputted are target user
User characteristics, several GBDT decision-tree models completed by GBDT training, to the features of the user characteristics of target user into
Row coding obtains feature combination, such as 0 and 1 combined ordered series of numbers in figure.By feature combination again input training completion LR model (that is,
First model), the first probability exported.
Similarly, for the third probability for calculating the discount coupon to be pushed, GBDT decision when due to third model training
The building of tree use user characteristics and discount coupon feature, therefore input be also the discount coupon to be pushed discount coupon feature with
And the user characteristics of target user, several GBDT decision-tree models completed by GBDT training, to the user characteristics of input with
And discount coupon feature is encoded, and determines that feature combines, then combines the feature determined as input, it is defeated by third model
Third probability out.
S106: according to the cost of the probability and the discount coupon to be pushed, which is pushed to by determination
The return value of the target user.
In this specification embodiment, when server is for each discount coupon to be pushed, determining should discount coupon pair be pushed
After the first probability, the second probability and the third probability answered, the discount coupon to be pushed can be calculated and correspond to the target user's
Return value.So which discount coupon to be pushed is given the target user by subsequent determination.
Specifically, the server can determine that the target user is receiving according to the difference of the second probability and the first probability
After should be wait push discount coupon, possible increased purchase probability corresponds to the number of users of user type according to the target user later
Amount, determines the number of transaction of the actual gain of the discount coupon to be pushed.
And the target user can be determined using should according to corresponding second probability of the discount coupon to be pushed and third probability
The probability (that is, probability that buying behavior is executed using the discount coupon to be pushed) that the content of discount coupon to be pushed is done shopping, it
It is used afterwards according to the cost (e.g., being somebody's turn to do the corresponding preferential amount of money of content of discount coupon to be pushed) and the target of being somebody's turn to do discount coupon to be pushed
Family corresponds to the number of users of user type, when determining that this is done shopping using the content wait pushing discount coupon, required for platform
The cost of cost.
It finally according to the ratio of number of transaction and cost, determines for the target user, is somebody's turn to do discount coupon to be pushed and facilitates list
The cost of transaction.It is, the push discount coupon can facilitate the mesh when the target user receives this after pushing discount coupon
The cost that mark user is consumed.And the return value of the target user is corresponded to using this as the discount coupon to be pushed.
S110: according to each return value determined, from respectively wait push in discount coupon, determining targeted coupon and be pushed to institute
State target user.
In this specification embodiment, server, can after determining the corresponding return value of discount coupon respectively to be pushed
According to the sequence of return value from big to small, determines the discount coupon to be pushed of specified quantity, as targeted coupon and be pushed to this
Target user.Wherein, specified quantity can be set as needed, for example, number of executions be 1 when, server can be highest by return value
Discount coupon to be pushed is pushed to the target user as targeted coupon.
Specifically can be when specified quantity be 1, the process of step S106 to step S108 can use formula:It indicates.Wherein Max (ROI)I, jRepresentative function takes ROI
It is worth maximum discount coupon to be pushed as targeted coupon, i indicates i-th of user (that is, target user), and j-th of j expression wait push away
Discount coupon is sent, ROI indicates return value, therefore (ROI)I, jIt indicates for target user i, the return value of discount coupon j to be pushed.P
(model2)I, jIndicate target user i the second probability of discount coupon j to be pushed being calculated by the second model, similarly P
(model3)I, jIndicate the target user i discount coupon j third probability to be pushed being calculated by third model, P (model1)i
Indicate that the first probability of target user i being calculated by the first model, N indicate the use of the corresponding user type of target user i
Amount amount, ValuejIndicate the preferential cost of the discount coupon to be pushed.
Based on push Coupon Method shown in FIG. 1, according to the corresponding user type of target user (e.g., user draws a portrait),
Execution buying behavior of the target user under the different service conditions for treating push discount coupon is calculated by trained model
Probability.The not only simple probability that discount coupon to be pushed is used according to calculated user, which to determine waited pushing
Discount coupon is pushed to user as targeted coupon, but according to the execution buying behavior in the above-mentioned various situations determined
Probability, determine that using which discount coupon to be pushed as targeted coupon to be pushed to the target user more suitable.Also, each probability
And be calculated according to the user characteristics of the target user of input, therefore be also to be determined individually for the target user.
The accuracy for improving push discount coupon increases push discount coupon transfer efficiency.
In addition, being mentioned in step S104 in the training process of this specification in order to make training sample closer to the mesh
User is marked, the model for completing training is more acurrate, and user group can be divided into several user types in advance by server.Specifically
, server can be based on user information, and user group is divided, determines several user types.For example, by 12 to 15 years old
Male, occupation in 15 to 18 years old is the male of student, and occupation in 15 to 18 years old is not that the male of student is respectively divided into difference
User type, etc..
In addition, the training for the first model.
In the present specification, server can be determined and be used first according to the historical behavior data of other users in the user type
User characteristics in each training sample of training pattern.
Later, according to the counter-example and GBDT in training sample, the GBDT decision tree mould for handling user characteristics is trained
Type.Also, since the purpose of training managing user characteristics is that user characteristics is made to be more suitable for predicting, the GBDT decision
The prediction target of tree can be the probability for executing buying behavior for the discount coupon of not holding of prediction user.Wherein, each counter-example can be made
It is input by GBDT, training obtains several GBDT decision trees.Certainly, without limitation for training termination condition this specification,
It can be and determine that training is completed when accuracy rate reaches threshold value, or determination has been trained when frequency of training reaches preset times
At can specifically be set as needed.
Then, according to the corresponding feature of node each in the GBDT decision-tree model of training completion, linear regression LR mould is determined
Each feature of type.Since preset user characteristics quantity is more, but for the first model for predicting user's purchase probability,
The role of each user characteristics is not exactly the same, therefore can be screened user characteristics by GDBT, and training can be improved
The forecasting accuracy of the LR model of completion.Also, there is GDBT it is confirmed that several decision trees, and each node of tree construction is all point
Branch, therefore can be the feature of discrete values by the Feature Conversion of consecutive variations value, it can further improve the training effect of LR model.
For example, this user characteristics of daily user's online duration, the online duration of different user is personalized, and number
Value is more mixed and disorderly, such as 1 hour, 5 minutes 1 hour, 45 minutes, 30 minutes 3 hours, it is further assumed that trained by GDBT
To some decision tree in a node are as follows: whether daily online duration more than 1 hour, then the user characteristics are by " per daily
Family online duration " is converted for user characteristics " whether daily user's online duration is more than 1 hour ", is that then characteristic value is 1, otherwise
Characteristic value is 0, and the time numerical value of the online duration of consecutive variations is converted for 0 and 1 two kind of characteristic value.
Finally, the training LR model obtains first model according to the counter-example in the training sample.With existing skill
The training process of LR model is consistent in art, and server inputs in the LR model to be trained using each counter-example as input, training
The LR model for not receiving the purchase probability of coupon user to be pushed to prediction, as the first model.
Further, for the training of the second model.
In the present specification, the first model which can first complete according to training calculates and determines that each positive example includes
User executes the probability of buying behavior when not holding discount coupon, the first probability as positive example.And it is general by the first of the positive example
Rate, also as a kind of feature of training sample.Later, according to include in each training sample user characteristics, discount coupon feature
And the first probability of the positive example, determine that each feature of LR model to be trained is unpaid.Later, using each positive example as input, instruction
Practice LR model and obtains the second model.
It should be noted that discount coupon feature is also contained due to containing user characteristics in determining positive example,
First probability of positive example, the user for including for the positive example execute the general of buying behavior when not holding the discount coupon that the positive example includes
Rate.It is repeated no more certainly for training process this specification of LR model.
Further, for the training of third model.
In the present specification, for the training third model, which is also to first pass through GBDT to determine LR model
Feature, then determined by training LR model.It is different from the decision tree completed in the first model by GBDT training, due to
The third model is related to calculating user using the probability of discount coupon execution Shopping Behaviors, therefore constructs the GBDT decision tree and be somebody's turn to do
It also include discount coupon feature in the feature of LR model.
Specifically, server can according to the positive example in training sample, by GBDT training for handle user characteristics and
The GBDT decision-tree model of discount coupon feature.Similarly, since the purpose of training managing user characteristics is to keep user characteristics more suitable
It shares in prediction, therefore the prediction target of the GBDT decision tree can execute the general of buying behavior using discount coupon for prediction user
Rate.Also, the corresponding feature of node of each GBDT decision-tree model determines each feature of LR model to be trained.Finally, according to
Counter-example in training sample and the respectively corresponding characteristic value of discount coupon to be pushed, the training LR model obtain the third model.It should
Training process can refer to the description during the first model training, and this specification repeats no more this.
It should be noted that the first model that with good grounds training is completed calculates due in the feature of the second model of training
Feature (that is, first probability of positive example), therefore the training of second model be the first model training completion after carry out.
In addition the process of push targeted coupon described in this specification is also applied for push other information, then calculates
Probability is the probability for certain behavior that the information is facilitated.For example, being executed when other information is advertising information using information to be pushed
The probability of buying behavior, the conversion ratio that can be advertisement (that is, after user receives the advertisement, buy the general of commodity in advertisement
Rate).Alternatively, executing the probability of buying behavior using information to be pushed when other information is action message, user's reception can be
To after the action message, the movable probability is participated in.
Based on push Coupon Method shown in FIG. 1, this specification embodiment is also corresponding to provide a kind of push discount coupon dress
The structural schematic diagram set, as shown in Figure 3.
Fig. 3 is a kind of structural schematic diagram for push coupon device that this specification embodiment provides, and described device includes:
User characteristics determining module 202 determines the user of the target user according to the historical behavior data of target user
Feature；
Probability evaluation entity 204, for respectively discount coupon to be pushed, according to the user characteristics of the target user and should be to
The discount coupon feature for pushing discount coupon determines the target user to the discount coupon to be pushed by model trained in advance
Different service conditions under execute buying behavior probability；
Return computing module 206, according to the probability and should discount coupon be pushed cost, determine by this wait pushing it is excellent
Favour certificate is pushed to the return value of the target user；
Pushing module 208, according to each return value determined, from respectively wait push in discount coupon, determining targeted coupon simultaneously
It is pushed to the target user；
Optionally, described device further include: training module 210 determines several users according to the user preferential certificate of each user
Type determines training sample according to each discount coupon and each user of the user type to each user type determined,
And training is for predicting that the user of the user type executes the probability of buying behavior under the different service conditions to discount coupon
Model, wherein be directed to each training sample, which includes a user and a discount coupon, if in the training sample
The user for including holds the discount coupon for including in the training sample, then the training sample is positive example, if including in the training sample
User do not hold the discount coupon for including in the training sample, then the training sample be counter-example.
Optionally, training module 210, according to the counter-example in the training sample, training is obtained for predicting that user does not hold
The first model that the probability of buying behavior is executed when having a discount coupon, according in the training sample positive example and each positive example pass through
The first probability that first model is calculated executes buying behavior when training is obtained for predicting that user holds discount coupon
Second model of probability, according to positive example in the training sample, training is obtained for predicting that user executes purchase using discount coupon
The third model of the probability of behavior.
Optionally, training module 210, according to the counter-example and grad enhancement decision tree GBDT in the training sample, instruction
Practice the GBDT decision-tree model for handling user characteristics, according to the GBDT decision-tree model that training is completed, handles each in counter-example
The user characteristics of user, and using processing result as the input of linear regression LR model, with training for predicting that user does not hold
The LR model that the probability of buying behavior is executed when discount coupon, as first model.
Optionally, training module 210, for each positive example in the training sample, completed according to training described the
One model calculates the probability that buying behavior is executed when the user for including in the positive example does not hold discount coupon, and first as positive example is general
Rate is held when training is for predicting that user holds discount coupon according to the first probability of positive example and positive example in the training sample
The LR model of the probability of row buying behavior, as second model.
Optionally, training module 210, according to the positive example and grad enhancement decision tree GBDT in the training sample, instruction
Practice the GBDT decision-tree model for handling user characteristics and discount coupon feature, the GBDT decision tree mould completed according to training
Type handles the discount coupon feature of the user characteristics and discount coupon of user in each positive example, and using processing result as linear regression
The input of LR model, with training for predicting user using the LR model of the probability of discount coupon execution buying behavior, as described
Third model.
Optionally, probability evaluation entity 204 determine the target user couple according to the user information of the target user
The user type answered, according to the user characteristics of the target user, by being directed to the corresponding user class of the target user in advance
The first probability of target user described in first model prediction of type training, according to the user characteristics of the target user and should be to
The discount coupon feature for pushing discount coupon, it is pre- by the second model for being directed to the corresponding user type training of the target user in advance
The second probability for surveying the target user, according to the user characteristics of the target user and the discount coupon of the discount coupon to be pushed
Feature, by be directed to target user described in the third model prediction of the target user corresponding user type training in advance the
Three probability.
Optionally, computing module 206 is returned, according to the difference of second probability and first probability and described
The number of users of the corresponding user type of target user determines the receipts that the discount coupon to be pushed is pushed to the target user
Benefit, according to the third probability, first probability, the corresponding user type of the target user number of users and should be to
The preferential cost of discount coupon is pushed, the push cost that the discount coupon to be pushed is pushed to the target user is determined, according to institute
Income and the push cost are stated, determines the return value that the discount coupon to be pushed is pushed to the target user.
Optionally, pushing module 208 are pushed to the return of the target user according to the discount coupon respectively to be pushed determined
The sequence of value from big to small, determines the discount coupon to be pushed of specified quantity, is pushed to the target user as targeted coupon.
This specification embodiment additionally provides a kind of computer readable storage medium, which is stored with computer journey
Sequence, computer program can be used for executing the discount coupon method for pushing that above-mentioned Fig. 1 is provided.
Based on the method for discount coupon shown in FIG. 1 push, this specification embodiment also proposed electronic equipment shown in Fig. 4
Schematic configuration diagram.Such as Fig. 4, in hardware view, the electronic equipment include processor, internal bus, network interface, memory and
Nonvolatile memory is also possible that hardware required for other business certainly.Processor is read from nonvolatile memory
Corresponding computer program is taken then to run into memory, to realize discount coupon method for pushing described in above-mentioned Fig. 1.
Certainly, other than software realization mode, other implementations, such as logical device suppression is not precluded in this specification
Or mode of software and hardware combining etc., that is to say, that the executing subject of following process flow is not limited to each logic unit,
It is also possible to hardware or logical device.
In the 1990s, the improvement of a technology can be distinguished clearly be on hardware improvement (for example,
Improvement to circuit structures such as diode, transistor, switches) or software on improvement (improvement for method flow).So
And with the development of technology, the improvement of current many method flows can be considered as directly improving for hardware circuit.
Designer nearly all obtains corresponding hardware circuit by the way that improved method flow to be programmed into hardware circuit.Cause
This, it cannot be said that the improvement of a method flow cannot be realized with hardware entities module.For example, programmable logic device
(Programmable Logic Device, PLD) (such as field programmable gate array (Field Programmable Gate
Array, FPGA)) it is exactly such a integrated circuit, logic function determines device programming by user.By designer
Voluntarily programming comes a digital display circuit " integrated " on a piece of PLD, designs and makes without asking chip maker
Dedicated IC chip.Moreover, nowadays, substitution manually makes IC chip, this programming is also used instead mostly " is patrolled
Volume compiler (logic compiler) " software realizes that software compiler used is similar when it writes with program development,
And the source code before compiling also write by handy specific programming language, this is referred to as hardware description language
(Hardware Description Language, HDL), and HDL is also not only a kind of, but there are many kind, such as ABEL
(Advanced Boolean Expression Language)、AHDL(Altera Hardware Description
Language)、Confluence、CUPL(Cornell University Programming Language)、HDCal、JHDL
(Java Hardware Description Language)、Lava、Lola、MyHDL、PALASM、RHDL(Ruby
Hardware Description Language) etc., VHDL (Very-High-Speed is most generally used at present
Integrated Circuit Hardware Description Language) and Verilog.Those skilled in the art also answer
This understands, it is only necessary to method flow slightly programming in logic and is programmed into integrated circuit with above-mentioned several hardware description languages,
The hardware circuit for realizing the logical method process can be readily available.
Controller can be implemented in any suitable manner, for example, controller can take such as microprocessor or processing
The computer for the computer readable program code (such as software or firmware) that device and storage can be executed by (micro-) processor can
Read medium, logic gate, switch, specific integrated circuit (Application Specific Integrated Circuit,
ASIC), the form of programmable logic controller (PLC) and insertion microcontroller, the example of controller includes but is not limited to following microcontroller
Device: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20 and Silicone Labs C8051F320 are deposited
Memory controller is also implemented as a part of the control logic of memory.It is also known in the art that in addition to
Pure computer readable program code mode is realized other than controller, can be made completely by the way that method and step is carried out programming in logic
Controller is obtained to come in fact in the form of logic gate, switch, specific integrated circuit, programmable logic controller (PLC) and insertion microcontroller etc.
Existing identical function.Therefore this controller is considered a kind of hardware component, and to including for realizing various in it
The device of function can also be considered as the structure in hardware component.Or even, it can will be regarded for realizing the device of various functions
For either the software module of implementation method can be the structure in hardware component again.
System, device, module or the unit that above-described embodiment illustrates can specifically realize by computer chip or entity,
Or it is realized by the product with certain function.It is a kind of typically to realize that equipment is computer.Specifically, computer for example may be used
Think personal computer, laptop computer, cellular phone, camera phone, smart phone, personal digital assistant, media play
It is any in device, navigation equipment, electronic mail equipment, game console, tablet computer, wearable device or these equipment
The combination of equipment.
For convenience of description, it is divided into various units when description apparatus above with function to describe respectively.Certainly, implementing this
The function of each unit can be realized in the same or multiple software and or hardware when specification.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
In a typical configuration, calculating equipment includes one or more processors (CPU), input/output interface, net
Network interface and memory.
Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/or
The forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flash RAM).Memory is computer-readable medium
Example.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable
Programmable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM),
Digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape magnetic disk storage or other magnetic storage devices
Or any other non-transmission medium, can be used for storage can be accessed by a computing device information.As defined in this article, it calculates
Machine readable medium does not include temporary computer readable media (transitory media), such as the data-signal and carrier wave of modulation.
It should also be noted that, the terms "include", "comprise" or its any other variant are intended to nonexcludability
It include so that the process, method, commodity or the equipment that include a series of elements not only include those elements, but also to wrap
Include other elements that are not explicitly listed, or further include for this process, method, commodity or equipment intrinsic want
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including described want
There is also other identical elements in the process, method of element, commodity or equipment.
It will be understood by those skilled in the art that the embodiment of this specification can provide as the production of method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or implementation combining software and hardware aspects can be used in this specification
The form of example.Moreover, it wherein includes the computer of computer usable program code that this specification, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
This specification can describe in the general context of computer-executable instructions executed by a computer, such as journey
Sequence module.Generally, program module include routines performing specific tasks or implementing specific abstract data types, programs, objects,
Component, data structure etc..This specification can also be practiced in a distributed computing environment, in these distributed computing environment
In, by executing task by the connected remote processing devices of communication network.In a distributed computing environment, program module
It can be located in the local and remote computer storage media including storage equipment.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for system reality
For applying example, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring to embodiment of the method
Part explanation.
The foregoing is merely the embodiments of this specification, are not limited to this specification.For art technology
For personnel, this specification can have various modifications and variations.It is all made any within the spirit and principle of this specification
Modification, equivalent replacement, improvement etc., should be included within the scope of the claims of this specification.
Claims (12)
1. a kind of discount coupon method for pushing characterized by comprising
According to the historical behavior data of target user, the user characteristics of the target user are determined；
It is special according to the user characteristics of the target user and the discount coupon for being somebody's turn to do discount coupon to be pushed for respectively discount coupon to be pushed
Sign determines that the target user executes under the different service conditions of discount coupon to be pushed to this by model trained in advance
The probability of buying behavior；
According to the cost of the probability and the discount coupon to be pushed, determine that the discount coupon to be pushed, which is pushed to the target, to be used
The return value at family；
According to for return value that respectively discount coupon to be pushed is determined respectively, from respectively wait push in discount coupon, determining targeted advantages
Certificate is simultaneously pushed to the target user.
2. the method as described in right wants 1, which is characterized in that preparatory training pattern specifically includes:
According to the user information of each user, several user types are determined；
Training sample is determined according to each discount coupon and each user of the user type for each user type determined,
And training is for predicting that the user of the user type executes the probability of buying behavior under the different service conditions to discount coupon
Model；
Wherein, for each training sample, which includes a user and a discount coupon, if wrapping in the training sample
The user contained holds the discount coupon for including in the training sample, then the training sample is positive example, if including in the training sample
User does not hold the discount coupon for including in the training sample, then the training sample is counter-example.
3. method according to claim 2, which is characterized in that preparatory training pattern specifically includes:
According to the counter-example in the training sample, buying behavior is executed when training is obtained for predicting that user does not hold discount coupon
First model of probability；
According to the positive example and the first probability for being calculated by first model of each positive example in the training sample, training
The second model of the probability of buying behavior is executed when obtaining for predicting that user holds discount coupon；
According to positive example in the training sample, training is obtained for predicting that user uses the probability of discount coupon execution buying behavior
Third model.
4. method as claimed in claim 3, which is characterized in that according to the counter-example in the training sample, training is used for
Prediction user executes the first model of the probability of buying behavior when not holding discount coupon, specifically include:
According to the counter-example and grad enhancement decision tree GBDT in the training sample, the GBDT for handling user characteristics is trained
Decision-tree model；
According to training complete GBDT decision-tree model, handle counter-example in each user user characteristics, and using processing result as
The input of linear regression LR model executes the LR mould of the probability of buying behavior when with training for predicting that user does not hold discount coupon
Type, as first model.
5. method as claimed in claim 3, which is characterized in that according in the training sample positive example and each positive example pass through
The first probability that first model is calculated executes buying behavior when training is obtained for predicting that user holds discount coupon
Second model of probability, specifically includes:
For each positive example in the training sample, include according to training first model completed to calculate in the positive example
User executes the probability of buying behavior, the first probability as positive example when not holding discount coupon；
According to the first probability of positive example and positive example in the training sample, held when training is for predicting that user holds discount coupon
The LR model of the probability of row buying behavior, as second model.
6. method as claimed in claim 3, which is characterized in that according to positive example in the training sample, training is obtained for pre-
The third model that user executes the probability of buying behavior using discount coupon is surveyed, is specifically included:
According to the positive example and grad enhancement decision tree GBDT in the training sample, training is for handling user characteristics and excellent
The GBDT decision-tree model of favour certificate feature；
According to the GBDT decision-tree model that training is completed, the discount coupon of the user characteristics and discount coupon of user in each positive example is handled
Feature, and using processing result as the input of linear regression LR model, with training for predicting that user executes purchase using discount coupon
The LR model for buying the probability of behavior, as the third model.
7. method as claimed in claim 3, which is characterized in that according to the user characteristics of the target user and be somebody's turn to do wait push
The discount coupon feature of discount coupon determines the target user to the discount coupon to be pushed by model trained in advance
The probability that buying behavior is executed under different service conditions, specifically includes:
According to the user information of the target user, the corresponding user type of the target user is determined；
According to the user characteristics of the target user, by be directed to the corresponding user type training of the target user in advance the
The first probability of target user described in one model prediction；
According to the discount coupon feature of the user characteristics of the target user and the discount coupon to be pushed, by advance for described
The second probability of target user described in second model prediction of the corresponding user type training of target user；
According to the discount coupon feature of the user characteristics of the target user and the discount coupon to be pushed, by advance for described
The third probability of target user described in the third model prediction of the corresponding user type training of target user.
8. the method for claim 7, which is characterized in that according to the probability and should discount coupon be pushed cost,
It determines the return value that the discount coupon to be pushed is pushed to the target user, specifically includes:
According to second probability and the difference of first probability and the user of the corresponding user type of the target user
Quantity determines the income that the discount coupon to be pushed is pushed to the target user；
According to the third probability, first probability, the corresponding user type of the target user number of users and should
The preferential cost of discount coupon to be pushed determines the push cost that the discount coupon to be pushed is pushed to the target user；
According to the income and the push cost, the return that the discount coupon to be pushed is pushed to the target user is determined
Value.
9. the method as described in claim 1, which is characterized in that according to for the return that respectively discount coupon to be pushed is determined respectively
Value, from respectively wait push in discount coupon, determining targeted coupon and being pushed to the target user, specifically includes:
It is pushed to the sequence of the return value of the target user from big to small according to the discount coupon respectively to be pushed determined, determination refers to
The discount coupon to be pushed of fixed number amount is pushed to the target user as targeted coupon.
10. a kind of discount coupon driving means characterized by comprising
User characteristics determining module determines the user characteristics of the target user according to the historical behavior data of target user；
Probability evaluation entity according to the user characteristics of the target user and is somebody's turn to do excellent wait push for respectively discount coupon to be pushed
The discount coupon feature of favour certificate determines the target user in the difference to the discount coupon to be pushed by model trained in advance
The probability of buying behavior is executed under service condition；
Computing module is returned, according to the cost of the probability and the discount coupon to be pushed, determination pushes away the discount coupon to be pushed
Give the return value of the target user；
Pushing module, according to each return value determined, from respectively wait push in discount coupon, determining targeted coupon and be pushed to institute
State target user.
11. a kind of computer readable storage medium, which is characterized in that the storage medium is stored with computer program, the meter
The claims 1-9 any method is realized when calculation machine program is executed by processor.
12. a kind of electronic equipment including memory, processor and stores the calculating that can be run on a memory and on a processor
Machine program, which is characterized in that the processor realizes the claims 1-9 any method when executing described program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910335996.4A CN110070399B (en) | 2019-04-24 | 2019-04-24 | Coupon pushing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910335996.4A CN110070399B (en) | 2019-04-24 | 2019-04-24 | Coupon pushing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110070399A true CN110070399A (en) | 2019-07-30 |
| CN110070399B CN110070399B (en) | 2024-11-01 |
Family
ID=67368715
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910335996.4A Active CN110070399B (en) | 2019-04-24 | 2019-04-24 | Coupon pushing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110070399B (en) |
Cited By (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110555168A (en) * | 2019-08-22 | 2019-12-10 | 阿里巴巴集团控股有限公司 | Information pushing method and device |
| CN110728323A (en) * | 2019-10-12 | 2020-01-24 | 中诚信征信有限公司 | Target type user identification method and device, electronic equipment and storage medium |
| CN111160951A (en) * | 2019-12-10 | 2020-05-15 | 北京三快在线科技有限公司 | Method and device for predicting excitation result, electronic equipment and readable storage medium |
| CN111178972A (en) * | 2019-12-30 | 2020-05-19 | 北京每日优鲜电子商务有限公司 | Message pushing method and device, storage medium and equipment |
| CN111507773A (en) * | 2020-04-22 | 2020-08-07 | 深圳市元征科技股份有限公司 | Coupon management method, device and storage medium |
| CN111553736A (en) * | 2020-04-30 | 2020-08-18 | 中国银行股份有限公司 | Method and device for processing preference information, readable medium and equipment |
| CN111563778A (en) * | 2020-05-12 | 2020-08-21 | 北京口袋财富信息科技有限公司 | Information pushing method and device |
| CN111586162A (en) * | 2020-05-06 | 2020-08-25 | 支付宝(杭州)信息技术有限公司 | Information pushing method and system |
| CN111881399A (en) * | 2020-07-20 | 2020-11-03 | 汉海信息技术(上海)有限公司 | Message pushing method and device |
| CN111931106A (en) * | 2020-07-29 | 2020-11-13 | 上海博泰悦臻电子设备制造有限公司 | Data processing method and related device |
| CN112070532A (en) * | 2020-08-27 | 2020-12-11 | 中国平安财产保险股份有限公司 | Information pushing method, device, equipment and storage medium |
| CN112270492A (en) * | 2020-11-12 | 2021-01-26 | 北京百家科技集团有限公司 | Resource allocation method, device, computer equipment and storage medium |
| CN112365283A (en) * | 2020-11-05 | 2021-02-12 | 广州视琨电子科技有限公司 | Coupon issuing method, device, terminal equipment and storage medium |
| CN112422696A (en) * | 2020-12-09 | 2021-02-26 | 上海益世界信息技术集团有限公司 | Content resource pushing determination method and related device |
| CN112465594A (en) * | 2020-11-27 | 2021-03-09 | 康键信息技术（深圳）有限公司 | Live broadcast interaction method, device, equipment and storage medium |
| CN112600756A (en) * | 2020-09-04 | 2021-04-02 | 京东数字科技控股股份有限公司 | Service data processing method and device |
| WO2021088600A1 (en) * | 2019-11-05 | 2021-05-14 | 北京三快在线科技有限公司 | Model training |
| CN112819526A (en) * | 2021-01-29 | 2021-05-18 | 北京骑胜科技有限公司 | To-be-pushed resource determination method, device, electronic equipment, medium and program product |
| CN112990954A (en) * | 2019-12-17 | 2021-06-18 | 北京沃东天骏信息技术有限公司 | Method and device for distributing coupons |
| CN113095723A (en) * | 2021-04-30 | 2021-07-09 | 中国银行股份有限公司 | Coupon recommendation method and device |
| CN113760521A (en) * | 2020-09-22 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Virtual resource allocation method and device |
| CN113824974A (en) * | 2021-08-27 | 2021-12-21 | 北京达佳互联信息技术有限公司 | Virtual asset sending method and device, electronic equipment and storage medium |
| WO2022048161A1 (en) * | 2020-09-01 | 2022-03-10 | 北京沃东天骏信息技术有限公司 | Electronic certificate assignment method and apparatus, device and storage medium |
| CN114169906A (en) * | 2020-09-11 | 2022-03-11 | 腾讯科技（深圳）有限公司 | Electronic ticket pushing method and device |
| CN114862472A (en) * | 2022-05-19 | 2022-08-05 | 上海钧正网络科技有限公司 | File generation and delivery method, file generation and delivery device, electronic equipment and medium |
| CN115860820A (en) * | 2023-01-19 | 2023-03-28 | 武汉利楚商务服务有限公司 | Intelligent encryption putting and destroying method and device for electronic coupons |
| CN117273821A (en) * | 2023-11-20 | 2023-12-22 | 阿里健康科技(杭州)有限公司 | Issuing method, training method and related device of electronic equity certificates |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107578281A (en) * | 2017-08-31 | 2018-01-12 | 湖南大学 | User preferential certificate behavior prediction method and model building method under e-commerce environment |
| CN107818478A (en) * | 2017-10-31 | 2018-03-20 | 携程计算机技术（上海）有限公司 | Overseas reward voucher distribution method and system based on provisional profit |
| CN107909392A (en) * | 2017-11-02 | 2018-04-13 | 北京三快在线科技有限公司 | Stimulation behavior susceptibility Forecasting Methodology, device, equipment and readable storage medium storing program for executing |
| CN108985809A (en) * | 2017-06-02 | 2018-12-11 | 北京京东尚科信息技术有限公司 | Motivate method, apparatus, electronic equipment and the storage medium of push |
-
2019
- 2019-04-24 CN CN201910335996.4A patent/CN110070399B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108985809A (en) * | 2017-06-02 | 2018-12-11 | 北京京东尚科信息技术有限公司 | Motivate method, apparatus, electronic equipment and the storage medium of push |
| CN107578281A (en) * | 2017-08-31 | 2018-01-12 | 湖南大学 | User preferential certificate behavior prediction method and model building method under e-commerce environment |
| CN107818478A (en) * | 2017-10-31 | 2018-03-20 | 携程计算机技术（上海）有限公司 | Overseas reward voucher distribution method and system based on provisional profit |
| CN107909392A (en) * | 2017-11-02 | 2018-04-13 | 北京三快在线科技有限公司 | Stimulation behavior susceptibility Forecasting Methodology, device, equipment and readable storage medium storing program for executing |
Cited By (38)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110555168A (en) * | 2019-08-22 | 2019-12-10 | 阿里巴巴集团控股有限公司 | Information pushing method and device |
| CN110555168B (en) * | 2019-08-22 | 2023-08-08 | 创新先进技术有限公司 | Information pushing method and device |
| CN110728323A (en) * | 2019-10-12 | 2020-01-24 | 中诚信征信有限公司 | Target type user identification method and device, electronic equipment and storage medium |
| CN110728323B (en) * | 2019-10-12 | 2022-06-21 | 中诚信征信有限公司 | Target type user identification method and device, electronic equipment and storage medium |
| WO2021088600A1 (en) * | 2019-11-05 | 2021-05-14 | 北京三快在线科技有限公司 | Model training |
| CN111160951A (en) * | 2019-12-10 | 2020-05-15 | 北京三快在线科技有限公司 | Method and device for predicting excitation result, electronic equipment and readable storage medium |
| CN112990954A (en) * | 2019-12-17 | 2021-06-18 | 北京沃东天骏信息技术有限公司 | Method and device for distributing coupons |
| CN111178972A (en) * | 2019-12-30 | 2020-05-19 | 北京每日优鲜电子商务有限公司 | Message pushing method and device, storage medium and equipment |
| CN111507773A (en) * | 2020-04-22 | 2020-08-07 | 深圳市元征科技股份有限公司 | Coupon management method, device and storage medium |
| CN111553736B (en) * | 2020-04-30 | 2023-06-20 | 中国银行股份有限公司 | Method and device for processing preferential information, readable medium and equipment |
| CN111553736A (en) * | 2020-04-30 | 2020-08-18 | 中国银行股份有限公司 | Method and device for processing preference information, readable medium and equipment |
| CN111586162A (en) * | 2020-05-06 | 2020-08-25 | 支付宝(杭州)信息技术有限公司 | Information pushing method and system |
| CN111563778A (en) * | 2020-05-12 | 2020-08-21 | 北京口袋财富信息科技有限公司 | Information pushing method and device |
| CN111563778B (en) * | 2020-05-12 | 2021-08-03 | 北京口袋财富信息科技有限公司 | Information pushing method and device |
| CN111881399B (en) * | 2020-07-20 | 2023-10-27 | 汉海信息技术(上海)有限公司 | Message pushing method and device |
| CN111881399A (en) * | 2020-07-20 | 2020-11-03 | 汉海信息技术(上海)有限公司 | Message pushing method and device |
| CN111931106A (en) * | 2020-07-29 | 2020-11-13 | 上海博泰悦臻电子设备制造有限公司 | Data processing method and related device |
| CN112070532A (en) * | 2020-08-27 | 2020-12-11 | 中国平安财产保险股份有限公司 | Information pushing method, device, equipment and storage medium |
| WO2022048161A1 (en) * | 2020-09-01 | 2022-03-10 | 北京沃东天骏信息技术有限公司 | Electronic certificate assignment method and apparatus, device and storage medium |
| CN112600756A (en) * | 2020-09-04 | 2021-04-02 | 京东数字科技控股股份有限公司 | Service data processing method and device |
| CN112600756B (en) * | 2020-09-04 | 2023-08-04 | 京东科技控股股份有限公司 | Service data processing method and device |
| CN114169906A (en) * | 2020-09-11 | 2022-03-11 | 腾讯科技（深圳）有限公司 | Electronic ticket pushing method and device |
| CN114169906B (en) * | 2020-09-11 | 2024-03-22 | 腾讯科技（深圳）有限公司 | Electronic coupon pushing method and device |
| CN113760521A (en) * | 2020-09-22 | 2021-12-07 | 北京沃东天骏信息技术有限公司 | Virtual resource allocation method and device |
| CN112365283B (en) * | 2020-11-05 | 2024-05-17 | 广州视琨电子科技有限公司 | Coupon issuing method and device, terminal equipment and storage medium |
| CN112365283A (en) * | 2020-11-05 | 2021-02-12 | 广州视琨电子科技有限公司 | Coupon issuing method, device, terminal equipment and storage medium |
| CN112270492A (en) * | 2020-11-12 | 2021-01-26 | 北京百家科技集团有限公司 | Resource allocation method, device, computer equipment and storage medium |
| CN112465594A (en) * | 2020-11-27 | 2021-03-09 | 康键信息技术（深圳）有限公司 | Live broadcast interaction method, device, equipment and storage medium |
| CN112465594B (en) * | 2020-11-27 | 2024-06-18 | 康键信息技术（深圳）有限公司 | Live interaction method, device, equipment and storage medium |
| CN112422696A (en) * | 2020-12-09 | 2021-02-26 | 上海益世界信息技术集团有限公司 | Content resource pushing determination method and related device |
| CN112819526A (en) * | 2021-01-29 | 2021-05-18 | 北京骑胜科技有限公司 | To-be-pushed resource determination method, device, electronic equipment, medium and program product |
| CN113095723A (en) * | 2021-04-30 | 2021-07-09 | 中国银行股份有限公司 | Coupon recommendation method and device |
| CN113824974A (en) * | 2021-08-27 | 2021-12-21 | 北京达佳互联信息技术有限公司 | Virtual asset sending method and device, electronic equipment and storage medium |
| CN113824974B (en) * | 2021-08-27 | 2022-10-04 | 北京达佳互联信息技术有限公司 | Virtual asset sending method and device, electronic equipment and storage medium |
| CN114862472A (en) * | 2022-05-19 | 2022-08-05 | 上海钧正网络科技有限公司 | File generation and delivery method, file generation and delivery device, electronic equipment and medium |
| CN115860820A (en) * | 2023-01-19 | 2023-03-28 | 武汉利楚商务服务有限公司 | Intelligent encryption putting and destroying method and device for electronic coupons |
| CN117273821A (en) * | 2023-11-20 | 2023-12-22 | 阿里健康科技(杭州)有限公司 | Issuing method, training method and related device of electronic equity certificates |
| CN117273821B (en) * | 2023-11-20 | 2024-03-01 | 阿里健康科技(杭州)有限公司 | Issuing method, training method and related device of electronic equity certificates |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110070399B (en) | 2024-11-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110070399A (en) | A kind of discount coupon method for pushing and device | |
| CN110033314B (en) | Advertisement data processing method and device | |
| CN103823908B (en) | Content recommendation method and server based on user preference | |
| CN108305102A (en) | Electronics red packet distribution method, device and client | |
| WO2019161731A1 (en) | Method and apparatus for channel attribution | |
| CN106295832A (en) | Product information method for pushing and device | |
| CN108537568A (en) | A kind of information recommendation method and device | |
| CN109284864B (en) | Behavior sequence obtaining method and device and user conversion rate prediction method and device | |
| CN108399477A (en) | A kind of method and device that risk threshold value determines | |
| CN106202088A (en) | A kind of method and system mating business scenario | |
| CN110647687B (en) | Service recommendation method and device | |
| CN109741177A (en) | Appraisal procedure, device and the intelligent terminal of user credit | |
| WO2020173275A1 (en) | Information processing method, apparatus and device | |
| CN106952114A (en) | A kind of advertisement editing customized method, device and advertisement editing machine | |
| CN110046932A (en) | A kind of data processing method, device and equipment | |
| CN110245973A (en) | Data processing method and device | |
| CN107993088A (en) | A kind of Buying Cycle Forecasting Methodology and device, electronic equipment | |
| CN110262998B (en) | Account checking data processing method and device | |
| CN108830705A (en) | A kind of method of summary of transaction data, device and equipment | |
| CN109003090A (en) | risk control method and device | |
| WO2023134644A1 (en) | Method, apparatus, and device for determining pre-delivery quantity, and storage medium | |
| CN110517095A (en) | A kind of method and device for launching advertisement | |
| CN108632358B (en) | Token management method and device based on WIFI sharing | |
| CN110134860A (en) | User's portrait generation method, device and equipment | |
| CN106909567A (en) | Data processing method and device |
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 | ||
| GR01 | Patent grant |