CN110163665A - A kind of advertisement inventory inquiry amount method, apparatus, equipment and storage medium - Google Patents
A kind of advertisement inventory inquiry amount method, apparatus, equipment and storage medium Download PDFInfo
- Publication number
- CN110163665A CN110163665A CN201910364693.5A CN201910364693A CN110163665A CN 110163665 A CN110163665 A CN 110163665A CN 201910364693 A CN201910364693 A CN 201910364693A CN 110163665 A CN110163665 A CN 110163665A
- Authority
- CN
- China
- Prior art keywords
- advertisement
- inventory
- targeted advertisements
- training
- coefficient
- 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 89
- 238000003860 storage Methods 0.000 title claims abstract description 21
- 238000012549 training Methods 0.000 claims description 116
- 238000012417 linear regression Methods 0.000 claims description 112
- 241001269238 Data Species 0.000 claims description 9
- 238000012512 characterization method Methods 0.000 claims description 9
- 238000012937 correction Methods 0.000 claims description 8
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 4
- 238000004590 computer program Methods 0.000 claims description 3
- 238000013480 data collection Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 20
- 230000008569 process Effects 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 238000012545 processing Methods 0.000 description 7
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 2
- 238000000151 deposition Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- 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
- G06Q30/0241—Advertisements
- G06Q30/0272—Period of advertisement exposure
Abstract
The embodiment of the present application discloses a kind of advertisement inventory inquiry amount method, apparatus, equipment and storage medium, wherein this method comprises: determining that targeted advertisements launch each corresponding inventory of advertisement position on the page；Determine the corresponding theoretical available stock of targeted advertisements；It is modified by losing coefficient theoretical available stock corresponding to the targeted advertisements, obtain the corresponding actually available inventory of the targeted advertisements, wherein, losing coefficient can characterize when advertisement delivery system carries out advertisement dispensing using duplicate removal rule to the influence degree of advertisement exposure amount.This method considers the influence of duplicate removal rule during inquiry amount, theoretical available stock is modified using coefficient is lost, guarantee that finally obtained actually available inventory is more accurate, reduce advertisement and lack the generation for broadcasting phenomenon, guarantees the actual demand for meeting contract advertiser.
Description
Technical field
This application involves Internet technical field more particularly to a kind of advertisement inventory inquiry amount method, apparatus, equipment and storage
Medium.
Background technique
Currently, advertisement launching platform is provides contract ad placement services to advertiser, so-called contract advertisement, which refers to, supports value
The dispensing requirement that the advertisement of guarantor's amount, i.e. guarantee advertisement reach advertiser in light exposure.Therefore, advertisement launching platform is sold in flow
When, it needs to know current advertisement inventory, just needs to determine advertisement inventory by advertisement inventory's inquiry amount method at this time, if asking
It is higher to measure result, will cause the scarce of advertisement and broadcast, be unable to satisfy the actual demand of contract advertiser.
Existing advertisement inventory inquiry amount method think each advertisement position be it is independent, mutually it is incoherent, cannot be compatible with
Advertisement duplicate removal rule, and advertisement dispensing is all based on the progress of advertisement duplicate removal rule in practical applications, this results in inquiry amount knot
Fruit is higher.
Summary of the invention
The embodiment of the present application provides a kind of advertisement inventory inquiry amount method, apparatus, equipment and storage medium, in inquiry amount advertisement
Advertisement duplicate removal rule has been comprehensively considered during inventory, guarantees that finally determining inquiry amount result is more accurate.
In view of this, the application first aspect provides a kind of advertisement inventory inquiry amount method, which comprises
Determine that targeted advertisements launch each corresponding inventory of advertisement position on the page；
Determine the corresponding theoretical available stock of targeted advertisements to be put；
It is modified by losing coefficient theoretical available stock corresponding to the targeted advertisements, obtains the targeted advertisements
Corresponding actually available inventory, the coefficient characterization advertisement delivery system of losing use duplicate removal rule when carrying out advertisement dispensing to wide
Accuse the influence degree of light exposure.
The application second aspect provides a kind of advertisement inventory inquiry amount device, and described device includes:
First determining module, for determining that targeted advertisements launch each corresponding inventory of advertisement position on the page；
Second determining module, for determining the corresponding theoretical available stock of targeted advertisements to be put；
Correction module is obtained for being modified by losing coefficient theoretical available stock corresponding to the targeted advertisements
The corresponding actually available inventory of the targeted advertisements is obtained, the coefficient characterization advertisement delivery system of losing is carried out using duplicate removal rule
To the influence degree of advertisement exposure amount when advertisement is launched.
The application third aspect provides a kind of equipment, and the equipment includes processor and memory:
Said program code is transferred to the processor for storing program code by the memory；
The processor is used to execute the advertisement base as described in above-mentioned first aspect according to the instruction in said program code
The step of depositing inquiry amount method.
The application fourth aspect provides a kind of computer readable storage medium, and the computer readable storage medium is for depositing
Program code is stored up, said program code is for executing advertisement inventory inquiry amount method described in above-mentioned first aspect.
The 5th aspect of the application provides a kind of computer program product including instruction, when run on a computer,
So that the computer executes advertisement inventory inquiry amount method described in above-mentioned first aspect.
As can be seen from the above technical solutions, the embodiment of the present application has the advantage that
The embodiment of the present application provides a kind of advertisement inventory inquiry amount method, and this method considers during inquiry amount advertisement inventory
Influence of the duplicate removal rule for advertisement inventory's inquiry amount result.Specifically, first determining target when carrying out inquiry amount for targeted advertisements
Each corresponding inventory of advertisement position on the page is launched in advertisement, then according to the corresponding inventory of each advertisement position and targeted advertisements
Stereotactic conditions, determine the corresponding theoretical available stock of the targeted advertisements；In turn, corresponding to the targeted advertisements by losing coefficient
Theoretical available stock be modified, obtain the corresponding actually available inventory of the targeted advertisements, wherein losing coefficient can characterize
Advertisement delivery system uses duplicate removal rule when carrying out advertisement dispensing to the influence degree of advertisement exposure amount.In this way, being using losing
Several pairs of theoretical available stocks are modified to obtain actually available inventory, guarantee that the actually available inventory finally determined is to consider to go
It is determined in the case where the influence of weight-normality then, that is, guarantees that the inquiry amount result finally determined is more accurate, reduction advertisement, which lacks, broadcasts phenomenon
Generation, guarantee to meet the actual demand of contract advertiser.
Detailed description of the invention
Fig. 1 is the schematic diagram that page duplicate removal principle is launched in advertisement；
Fig. 2 is the application scenarios schematic diagram of advertisement inventory inquiry amount method provided by the embodiments of the present application；
Fig. 3 is the flow diagram of advertisement inventory inquiry amount method provided by the embodiments of the present application；
Fig. 4 is the derivation schematic diagram of the corresponding actually available inventory of targeted advertisements provided by the embodiments of the present application；
Fig. 5 is the structural schematic diagram of the first advertisement inventory inquiry amount device provided by the embodiments of the present application；
Fig. 6 is the structural schematic diagram of second of advertisement inventory inquiry amount device provided by the embodiments of the present application；
Fig. 7 is the structural schematic diagram of the third advertisement inventory inquiry amount device provided by the embodiments of the present application；
Fig. 8 is the structural schematic diagram of the 4th kind of advertisement inventory's inquiry amount device provided by the embodiments of the present application；
Fig. 9 is the structural schematic diagram of the 5th kind of advertisement inventory's inquiry amount device provided by the embodiments of the present application；
Figure 10 is the structural schematic diagram of the 6th kind of advertisement inventory's inquiry amount device provided by the embodiments of the present application；
Figure 11 is a kind of structural schematic diagram of server provided by the embodiments of the present application；
Figure 12 is a kind of structural schematic diagram of terminal device provided by the embodiments of the present application.
Specific embodiment
In order to make those skilled in the art more fully understand application scheme, below in conjunction in the embodiment of the present application
Attached drawing, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described embodiment is only this
Apply for a part of the embodiment, instead of all the embodiments.Based on the embodiment in the application, 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.
The description and claims of this application and term " first ", " second ", " third ", " in above-mentioned attached drawing
The (if present)s such as four " are to be used to distinguish similar objects, without being used to describe a particular order or precedence order.It should manage
The data that solution uses in this way are interchangeable under appropriate circumstances, so that embodiments herein described herein for example can be to remove
Sequence other than those of illustrating or describe herein is implemented.In addition, term " includes " and " having " and theirs is any
Deformation, it is intended that cover it is non-exclusive include, for example, containing the process, method of a series of steps or units, system, production
Product or equipment those of are not necessarily limited to be clearly listed step or unit, but may include be not clearly listed or for this
A little process, methods, the other step or units of product or equipment inherently.
Advertisement launching platform is when providing contract ad placement services for advertiser, it usually needs according to determining for contract advertisement
Advertisement inventory inquiry amount is carried out to condition, obtains the remaining advertisement inventory for meeting the stereotactic conditions, and then according to the residue advertisement base
The predetermined degree of exposure of setting contract advertisement is deposited, i.e. the corresponding available stock of setting contract advertisement.If the available stock that inquiry amount determines
It is higher, the scarce generation for broadcasting phenomenon of advertisement will be caused, the demand of advertiser is unable to satisfy；If the available stock that inquiry amount determines is relatively low,
Will lead to flow can not sufficiently be sold, and the income of advertisement launching platform is influenced.As it can be seen that accurately determining the corresponding available library of advertisement
It deposits and all has very important meaning for advertisement launching platform and advertiser.
Nowadays, major advertisement launching platform is in order to advanced optimize user experience, during launching advertisement correspondingly
It is additionally arranged duplicate removal rule, that is, requires in certain window ranges, identical advertisement cannot be launched, it is right in the window ranges to need
Advertisement carries out duplicate removal processing.Referring to Fig. 1, Fig. 1 is that schematic diagram is launched in the corresponding advertisement of duplicate removal principle；As shown in Figure 1, it is assumed that certain is wide
Accusing and launching the duplicate removal window ranges of the page is 3, then cannot launch in a certain duplicate removal window ranges that the page is launched in the advertisement
Identical advertisement, i.e., the advertisement launched in advertisement position 1, advertisement position 2 and the advertisement position 3 in a certain duplicate removal window ranges cannot
It is identical.
However, existing advertisement inventory inquiry amount method think each advertisement position be it is independent, mutually incoherent, therefore,
The prior art does not consider influence of the duplicate removal rule for advertisement inventory's inquiry amount result usually in inquiry amount advertisement inventory, and then leads
It causes inquiry amount result higher, advertisement often occurs and lack the phenomenon that broadcasting.
In order to solve the above-mentioned problems of the prior art, the embodiment of the present application provides a kind of advertisement inventory inquiry amount side
Method, this method are carried out when determining advertisement inventory's inquiry amount result by means of losing coefficient theoretical available stock corresponding to advertisement
Amendment, this is lost coefficient and can characterize when advertisement delivery system carries out advertisement dispensing using duplicate removal rule to the shadow of advertisement exposure amount
The degree of sound guarantees in this way, considering influence of the duplicate removal rule for inquiry amount result during determining advertisement inventory's inquiry amount result
Finally determining inquiry amount result is more accurate, reduces advertisement and lacks the generation for broadcasting phenomenon, guarantees the actual demand for meeting advertiser.
It should be understood that advertisement inventory inquiry amount method provided by the embodiments of the present application can be applied to have data processing function
Equipment, such as terminal device, server.Wherein, terminal device is specifically as follows computer, personal digital assistant (Personal
Digital Assitant, PDA), smart phone, tablet computer etc.；Server is specifically as follows application server, or
Web server, in practical application deployment, which can be separate server, or cluster server.
Technical solution provided by the embodiments of the present application in order to facilitate understanding, carrying out advertisement inventory inquiry amount below with server is
Example, is introduced advertisement inventory inquiry amount method provided by the embodiments of the present application.
Referring to fig. 2, Fig. 2 is the application scenarios schematic diagram of advertisement inventory inquiry amount method provided by the embodiments of the present application.This is answered
With including: server 201 in scene, it is used to execute advertisement inventory inquiry amount method provided by the embodiments of the present application, for target
Advertisement determines its corresponding actually available inventory.
When advertiser needs to launch targeted advertisements by certain advertisement launching platform, advertisement launching platform needs to utilize service
Device 201 carries out advertisement inventory inquiry amount for the targeted advertisements, to determine the corresponding actually available inventory of the targeted advertisements, that is, determines
The corresponding light exposure of the targeted advertisements.
When specific progress advertisement inventory inquiry amount, server 201 first determines that each advertisement position is respectively on the targeted advertisements dispensing page
Corresponding inventory；Then, according to the corresponding inventory of above-mentioned each advertisement position and the stereotactic conditions of targeted advertisements, the mesh is determined
Mark the corresponding theoretical available stock of advertisement；In turn, server 201 is corresponding to the targeted advertisements theoretical available using coefficient is lost
Inventory is modified, and obtains the corresponding actually available inventory of the targeted advertisements, wherein advertisement dispensing system can be characterized by losing coefficient
System uses duplicate removal rule when carrying out advertisement dispensing to the influence degree of advertisement exposure amount.
Server 201 is during carrying out advertisement inventory inquiry amount, using losing coefficient to the corresponding theory of targeted advertisements
Available stock is modified, and obtains the corresponding actually available inventory of the targeted advertisements, in this way, comprehensively considered duplicate removal rule for
The influence of advertisement inventory's inquiry amount result guarantees that finally determining inquiry amount result is more accurate, reduces advertisement and lacks the hair for broadcasting phenomenon
It is raw, guarantee the actual demand for meeting the advertiser of targeted advertisements.
It should be understood that above-mentioned application scenarios shown in Fig. 2 are only a kind of example, in practical applications, can also be set by terminal
It is standby to independently execute advertisement inventory inquiry amount method provided by the embodiments of the present application, herein not to advertisement base provided by the embodiments of the present application
The application scenarios for depositing inquiry amount method do any restriction.
Advertisement inventory inquiry amount method provided by the present application is introduced below by embodiment.
Referring to Fig. 3, Fig. 3 is the flow diagram of advertisement inventory inquiry amount method provided by the embodiments of the present application；For the ease of
Description, following embodiments are described using server as executing subject, it should be appreciated that the execution master of advertisement inventory's inquiry amount method
Body is not limited in server, other can also have the equipment of data processing function for terminal device etc..As shown in figure 3, should
Advertisement inventory inquiry amount method the following steps are included:
Step 301: determining that targeted advertisements launch each corresponding inventory of advertisement position on the page.
Targeted advertisements launch the page and refer to the page for launching targeted advertisements, which launches the page usually can be with
It is determined according to the stereotactic conditions that advertiser proposes.The corresponding inventory of advertisement position refers to that a certain advertisement position can be allocated for throwing
Put the remaining inventory of targeted advertisements；Under normal conditions, inventory allocation first can be carried out to the advertisement launched, guarantees to meet and has thrown
The dispensing demand for the advertisement put；In turn, based on the remaining inventory after advertisement bit allocation, determine that the advertisement position can be allocated for
The inventory for launching targeted advertisements, as the corresponding inventory of the advertisement position.
When server needs to carry out advertisement inventory inquiry amount for targeted advertisements, server can correspondingly determine that target is wide
It accuses and launches each corresponding inventory of advertisement position on interface；It is respectively corresponded to it should be understood that targeted advertisements launch each advertisement position on interface
Inventory may be the same or different.
Step 302: determining the corresponding theoretical available stock of targeted advertisements to be put.
Determine that targeted advertisements are launched on the page after each corresponding inventory of advertisement position, server can be based further on
The stereotactic conditions of each corresponding inventory of advertisement position and the targeted advertisements determine that the corresponding theory of the targeted advertisements can use library
It deposits；Wherein, the stereotactic conditions of targeted advertisements are usually set by advertiser, can characterize the fixed condition of targeted advertisements, for example,
Release time launches object, dispensing place etc..
It should be noted that above-mentioned theory available stock be actually do not consider duplicate removal rule under conditions of, for
The available stock that targeted advertisements determine；The theory available stock is usually bigger than normal, directly that the theory available stock is wide as target
Corresponding advertisement inventory inquiry amount is accused as a result, normally resulting in targeted advertisements lacks the phenomenon that broadcasting.
In one possible implementation, server can be calculated by high-water line (High Water Mark, HWM)
Method determines the corresponding theoretical available stock of the targeted advertisements according to the fixed condition of targeted advertisements to be put.
When server determines targeted advertisements corresponding theoretical available stock by HWM algorithm, targeted advertisements can be first determined
Priority.Specifically, server can first calculate meet each flow node of each contract advertisement fixed condition flow and,
In turn, according to each corresponding flow of contract advertisement and the priority of contract advertisement is determined, flow and smaller shows that its is right
The priority answered is higher, so determines the priority of targeted advertisements and each contract advertisement.
Then, theoretical available stock is determined using HWM algorithm, at this time, it usually needs choose different value of inventory and try one by one
It tests, that is, takes different available stocks to substitute into HWM algorithm, finally find out so that HWM allocation algorithm has the maximum value of solution, as mesh
Mark the corresponding theoretical available stock of advertisement.Specifically, inventory allocation is done in the contract advertisement for being first higher than targeted advertisements to priority, it will
Remaining inventory distributes to targeted advertisements, obtains maximum available stock Max；Then, then by the priority of targeted advertisements it is set as
It is minimum, inventory allocation is carried out to the contract advertisement of other high priorities, remaining inventory is distributed into targeted advertisements, obtains minimum
Available stock Min.In turn, the binary chop between maximum available stock Max and minimum available stock Min, substitutes into allocation algorithm
Iterative solution is found out so that HWM allocation algorithm has the maximum value of solution, the as corresponding theoretical available stock of the targeted advertisements.
Specific calculating process is as follows:
Assuming that theoretical on-hand initial value I (0)=(Max+Min)/2；
If HWM algorithm has solution, I (t+1)=(Max+I (t))/2；
If HWM algorithm is without solution, I (t+1)=(Min+I (t))/2；
So iterative solution finally makes HWM allocation algorithm have the maximum value of solution, the as target until I (t) restrains
The corresponding theoretical available stock of advertisement.
It should be noted that during determining theoretical available stock using HWM algorithm, when executing HWM algorithm every time
The corresponding advertisement of the available stock can be obtained and play probability.
Advertisement plays probability P=I (t)/Sj；
Correspondingly, the corresponding advertisement of targeted advertisements plays probability, as the corresponding theoretical available stock of the targeted advertisements with
The ratio of its corresponding flow sum.
It should be understood that can also pass through other than HWM algorithm can be used to determine the corresponding theoretical available stock of targeted advertisements
Other modes determine the corresponding theoretical available stock of targeted advertisements, do not do any limit to theoretical on-hand method of determination herein
It is fixed.
Step 303: being modified by losing coefficient theoretical available stock corresponding to the targeted advertisements, described in acquisition
The corresponding actually available inventory of targeted advertisements, the coefficient characterization advertisement delivery system of losing is using the progress advertisement throwing of duplicate removal rule
To the influence degree of advertisement exposure amount when putting.
It is after server determines the corresponding theoretical available stock of targeted advertisements, i.e., available to lose coefficient to the targeted advertisements
Corresponding theory available stock is modified, to obtain the corresponding actually available inventory of the targeted advertisements；This loses coefficient energy
Enough characterization advertisement delivery system uses duplicate removal rule when carrying out advertisement dispensing to the influence degree of advertisement exposure amount.
Specifically, assuming that losing coefficient is Q, the corresponding theoretical available stock of targeted advertisements is RO, then targeted advertisements are corresponding
Actually available inventory RdIt can use theoretical available stock ROIt is multiplied by and loses coefficient Q and obtain, i.e. Rd=Ro×Q。
In one possible implementation, server can be wide according to target by the linear regression model (LRM) of pre-training
It accuses corresponding advertisement and plays probability, obtain that the targeted advertisements are corresponding to lose coefficient；The linear regression model (LRM) of the pre-training is logical
It crosses and generation is trained to training dataset, it includes wide in training data that it includes multiple training datas which, which concentrates,
Corresponding advertisement is accused to play probability and lose coefficient.
Through inventor the study found that losing between coefficient and advertisement broadcasting probability there are specific relationship, that is, lose coefficient
The multinomial of probability is played for advertisement.The relationship lost between coefficient and advertisement broadcasting probability is derived below with reference to Fig. 4,
Assuming that it includes 3 advertisement positions on the page that targeted advertisements, which are launched, the corresponding inventory of each advertisement position is S, which launches
The corresponding duplicate removal window ranges of the page are 3, and it is P that the corresponding advertisement of targeted advertisements A to be put, which plays probability,.
The corresponding theoretical available stock R of targeted advertisements Ao=3 × S × P；
The corresponding actually available inventory R of targeted advertisementsdTheory inventory R corresponding equal to targeted advertisementsOIt subtracts because of duplicate removal rule
Bring loses inventory RS；That is, Rd=Ro-Rs。
Wherein, because duplicate removal rule bring loses inventory RSFor on advertisement position 2 because duplicate removal rule bring inventory loss with
Because of the sum of duplicate removal rule bring inventory loss on advertisement position 3；
That is, RSThe inventory of 3 duplicate removal of the inventory+advertisement position loss of 2 duplicate removal of=advertisement position loss
=S × (advertisement A appears in the probability on advertisement position 1,2)+S × (advertisement A appears in the probability on advertisement position 1,3)
+ S × (advertisement A appears in the probability on advertisement position 2,3)
=S × P × P+S × P × P+S × (1-P) × P × P
In turn,
That is,
Further genralrlization is greater than 3, and the situation that the corresponding inventory of each advertisement position is different to advertisement bit quantity, loses coefficient
Q can be indicated are as follows: Q=w0+w1P+w2P2+…+wnPn
Based on the above-mentioned relationship lost between coefficient and advertisement broadcasting probability, machine learning algorithm can be used, packet is utilized
It includes advertisement and plays probability and its corresponding training data for losing coefficient, training obtains that determine the probability can be played according to advertisement
The linear regression model (LRM) of coefficient is lost, specifically trains the mode of the linear regression model (LRM) that will be situated between in subsequent processes embodiment
It continues, referring particularly to subsequent embodiment.
In turn, when server needs utilization is lost coefficient theoretical available stock corresponding to targeted advertisements and is modified,
The corresponding advertisement of targeted advertisements can be played probability and input above-mentioned linear regression model (LRM) by server, utilize the linear regression model (LRM)
Probability is played to the corresponding advertisement of targeted advertisements correspondingly to be handled, and is obtained the advertisement and is played the corresponding correction factor of probability,
And then it is modified using the correction factor theoretical available stock corresponding to targeted advertisements.
In alternatively possible implementation, the history that server can also acquire advertisement delivery system launches data,
It includes the actual exposure amount of advertisement and the predetermined degree of exposure of advertisement in history release time section that the history, which launches data,；In turn, root
The ratio that the predetermined degree of exposure of the actual exposure amount and advertisement of advertisement in data is launched according to history, determines and above-mentioned loses coefficient.
Specifically, server can according to actual needs, using past certain time period as history release time section, example
Such as, using the time of the previous moon as history release time section；Then, the history release time is acquired in advertisement delivery system
History in section launches data, which launches in data and generally include the corresponding actual exposure amount of multiple advertisements, and
The corresponding predetermined degree of exposure of these advertisements；In turn, right divided by its using its corresponding actual exposure amount for a certain advertisement
The predetermined degree of exposure answered obtains losing coefficient corresponding to the advertisement, in this way, in this manner it is achieved that determining each advertisement respectively
It is corresponding to lose coefficient, finally, the corresponding folding of the advertisement delivery system is determined based on the corresponding coefficient of losing of each advertisement
Damage coefficient.
It should be noted that in practical applications, it can also only include the corresponding reality of an advertisement that history, which is launched in data,
Border light exposure and predetermined degree of exposure do not launch the data volume for including in data to history herein and do any restriction.
It should be understood that other than the mode of coefficient is lost in above two determination, it in practical applications, can also be according to practical need
It asks and coefficient is lost using other modes determination, any restriction is not done to the method for determination for losing coefficient herein.
Above-mentioned advertisement inventory inquiry amount method is corresponding to advertisement by means of losing coefficient when determining advertisement inventory's inquiry amount result
Theoretical available stock be modified, this lose coefficient can characterize advertisement delivery system using duplicate removal rule carry out advertisement dispensing
When to the influence degree of advertisement exposure amount, in this way, consider during determining advertisement inventory's inquiry amount result duplicate removal rule for
The influence of inquiry amount result guarantees that finally determining inquiry amount result is more accurate, reduces advertisement and lacks the generation for broadcasting phenomenon, guarantees to meet
The actual demand of advertiser.
Being mentioned to server in the embodiment shown in fig. 3 can to the corresponding theory of targeted advertisements by losing coefficient
Be modified with inventory, in a kind of wherein possible implementation, this lose coefficient can be it is true using linear regression model (LRM)
Fixed；This kind loses the method for determination of coefficient in order to facilitate understanding, below with reference to Fig. 5 to the training method of linear regression model (LRM) into
Row is introduced.
Referring to Fig. 5, Fig. 5 is the flow diagram of the training method of linear regression model (LRM) provided by the embodiments of the present application；For
Convenient for description, following embodiments are described using server as executing subject, it should be appreciated that the training of the linear regression model (LRM)
The executing subject of method is not limited in server, other can also have the equipment of model training function for terminal device etc.；
It should be noted that for train the server of linear regression model (LRM) be used for inquiry amount advertisement inventory server can be same
Platform server, or different servers.As shown in figure 5, the training method of the linear regression model (LRM) the following steps are included:
Step 501: determining training dataset.
When server training linear regression model (LRM), it usually needs instructed using multiple training datas to linear regression model (LRM)
Practice；For this purpose, server needs first to obtain multiple training data composition training datasets before training linear regression model (LRM).It answers
Understand, since linear regression model (LRM) is for playing the model that determine the probability loses coefficient, correspondingly, server institute according to advertisement
It include that advertisement plays probability and its corresponding loses coefficient in the training data of acquisition.
In one possible implementation, server can collect training data in the following manner, form training number
According to collection: first obtaining advertisement when advertisement is launched and play when probability, the corresponding actual exposure amount of advertisement and advertisement are launched because of duplicate removal
Reason and caused by lose light exposure；Then, according to the corresponding actual exposure amount of advertisement and loss light exposure, the advertisement pair is determined
The correction factor answered；In turn, the corresponding advertisement of the advertisement is played into probability and identified correction factor as one group of trained number
According to passing through the different training data of multiple groups and form training dataset.
Specifically, the corresponding actual exposure amount of advertisement refers to the practical light exposure generated during launching the advertisement；
The corresponding loss light exposure of advertisement refers to the light exposure for causing advertisement not to be exposed for some reason and losing, and generates loss exposure
When amount, advertisement launching platform would generally correspondingly recording loss light exposure Producing reason, when acquiring training data, server can
To filter out the loss light exposure generated by duplicate removal rule from all loss light exposures, the loss light exposure and advertisement pair are utilized
Coefficient is lost in the actual exposure amount calculating answered.
It is specific to be calculated when losing coefficient using loss light exposure and the corresponding actual exposure amount of advertisement, it can first calculate because going
Weight-normality then caused by loss the sum of light exposure and actual exposure amount, in turn, using actual exposure amount divided by caused by because of duplicate removal rule
Loss light exposure and the sum of actual exposure amount, obtain losing coefficient.
By the corresponding advertisement of the advertisement play probability with through it is above-mentioned be calculated lose coefficient, as one group of trained number
According to；In this way, obtaining the different training data of multiple groups in the manner described above, training dataset is formed using these training datas.
It should be understood that in practical applications, server can also obtain training dataset by other means, herein not to instruction
The acquisition modes for practicing data set do any restriction.
Step 502: linear regression model (LRM) is obtained by the training of least square method combined training data set.
After server determines training dataset, it can be based on the training dataset, using least square method to initial
Linear regression model (LRM) is trained, and obtains the linear regression mould that determine the probability correction factor can be accurately played according to advertisement
Type.
It should be noted that least square method is a kind of mathematical optimization techniques, by minimize the quadratic sum of error come
Find the optimal function matching between data；Unknown data can be easily acquired using least square method, and make these
The quadratic sum of the error between data and real data acquired is minimum.
When especially by least square in training linear regression model (LRM), server can play advertisement in training data general
Each exponential term of rate as characteristic item, for example, using advertisement play exponential term P, P^2 of probability P, P^3 ..., P^n is as special
Levy item；In turn, using least square in training linear regression model (LRM), loss function specific manifestation shape used in training process
Shown in formula such as formula (1):
Formula (1) is a convex function, and there are minimum value, the training process to linear regression model (LRM) is actually to find
So that formula (1) reaches the coefficient vector W of minimum value；When specific training, it can be solved using gradient descent method, choose one group
Initial value of the random number as coefficient vector W is modified coefficient vector W according to formula (2) and formula (3) using training data:
Wherein, a is learning rate, can choose a lesser numerical value according to the actual situation；Iteration executes above-mentioned formula (2)
With formula (3) until coefficient W restrains, coefficient vector W is so obtained.
Initial linear regression model (LRM) is trained using different parameters, and then more different linear regression model (LRM)s
Root-mean-square error；Under normal conditions, as polynomial order n=2, regularization coefficient λ=0.0001, calculated model is equal
Square error is minimum, at this point, i.e. it is believed that linear regression model (LRM) training is completed.
It should be understood that can also judge by other means whether linear regression model (LRM) trains completion other than aforesaid way,
Any restriction is not done to the mode for judging that linear regression model (LRM) training is completed herein.
It should be noted that the linear regression model (LRM) that training obtains before will be fitted no longer when duplicate removal window changes
The page is launched in advertisement after changing for duplicate removal window, for this purpose, needing re -training suitable after duplicate removal window changes
Linear regression model (LRM) for new duplicate removal window；At this point, server needs to collect another training dataset, received again based on this
The training dataset of collection carries out re -training to linear regression model (LRM), and the linear regression model (LRM) that the re -training obtains is used for
In advertisement delivery system, coefficient is lost with determination advertisement to be put.
The training method used when re -training linear regression model (LRM), the instruction with above-mentioned linear regression model (LRM) shown in fig. 5
It is similar to practice method, needs to collect after duplicate removal window changes the corresponding actual exposure amount of advertisement and advertisement because of duplicate removal rule
Caused by lose light exposure, in turn, based on the actual exposure amount and lose light exposure in the manner described above to calculate the advertisement corresponding
Lose coefficient, this is lost into corresponding with the advertisement advertisement of coefficient and plays probability as one group of training data, is so obtained more
The training data that group duplicate removal window generates after changing, forms new training dataset；In turn, based on the new training data
Collection, using the new linear regression model (LRM) of least square in training.
The training method of above-mentioned linear regression model (LRM), based on it is multiple include that advertisement plays probability and its corresponding loses coefficient
Training data, linear regression model (LRM) is obtained using least square in training, which can be according to the wide of input
It accuses and plays probability, accurately determine its and corresponding lose coefficient.Training data for training linear regression model (LRM) is basis
What history advertisement played data determined, it is more objective and accurate, correspondingly, the linear regression obtained based on the training of these training datas
Model, which also can be determined relatively accurately, loses coefficient, in turn, loses coefficient to theory using what the linear regression model (LRM) determined
When available stock is modified, it can guarantee to obtain accurate reasonable actually available inventory.
For the ease of further understanding advertisement inventory inquiry amount method provided by the embodiments of the present application, below with server by utilizing
For coefficient is lost in linear regression model (LRM) determination, whole introduction is done to advertisement inventory inquiry amount method provided by the embodiments of the present application.
Coefficient is lost to determine using linear regression model (LRM), server needs preparatory training that can be suitable for current duplicate removal
The linear regression model (LRM) of window ranges；Specifically, server can first acquire multiple groups training data composition training dataset, adopted
In the training data of collection include advertisement play probability and its it is corresponding lose coefficient, this loses coefficient can be according to the reality of advertisement
Border light exposure and because duplicate removal rule caused by lose light exposure be calculated；In turn, server can be based on above-mentioned trained number
According to collection, linear regression model (LRM) is obtained using least square in training, which can play according to the advertisement of input
Probability correspondingly determines that its corresponding loses coefficient.
It should be noted that server can gone when advertisement launches the duplicate removal window ranges on the page and changes
Weight window ranges resurvey multiple training datas after changing and form new training dataset, in turn, using least square method knot
Close the training dataset re -training linear regression model (LRM).
When advertiser needs to launch targeted advertisements by advertisement launching platform, server can first determine the targeted advertisements
Launch each corresponding inventory of advertisement position and the corresponding theoretical available stock of the targeted advertisements on the page；It is specific to determine mesh
When marking the corresponding theoretical available stock of advertisement, server can be i.e. fixed according to the fixed condition of the targeted advertisements by HWM algorithm
To condition, determine that it is corresponding targeted advertisements can also to be calculated at the same time in the corresponding theoretical available stock of the targeted advertisements
Advertisement plays probability.
It should be understood that above-mentioned theory available stock is the available stock being calculated under conditions of not considering duplicate removal rule,
It is usually higher, directly using the theory available stock as inquiry amount as a result, advertisement is easily caused to lack the generation for broadcasting phenomenon.
In turn, the linear regression model (LRM) that server by utilizing is obtained through the training of above-mentioned training process, it is corresponding according to targeted advertisements
Advertisement play probability, correspondingly determine that its corresponding loses coefficient；When due to training linear regression model (LRM), used training
It include that advertisement plays probability, and determines based on advertisement actual exposure amount and because losing light exposure caused by duplicate removal rule in data
Lose coefficient, therefore, what which determined lose coefficient can characterize correspondingly carried out using duplicate removal rule it is wide
Accuse influence degree when launching to advertisement exposure amount.
Finally, the corresponding theoretical available stock of server by utilizing targeted advertisements, is multiplied by losing for linear regression model (LRM) determination
Coefficient obtains the corresponding actually available inventory of the targeted advertisements.
Experimental studies have found that, advertisement inventory inquiry amount method provided by the embodiments of the present application is applied into Mr. Yu's news through inventor
APP, the scarce of information flow advertisement broadcast rate and obviously drop to 7% from 27%, it is seen then that advertisement inventory's inquiry amount method can guarantee institute
Determining inquiry amount result is more accurate, is effectively reduced the scarce of advertisement and broadcasts rate, promotes the experience of advertiser.
For above-described advertisement inventory inquiry amount method, present invention also provides corresponding advertisement inventory inquiry amount device,
So that above-mentioned advertisement inventory inquiry amount method is able to apply and realize in practice.
It is a kind of advertisement inventory inquiry amount dress corresponding with advertisement inventory inquiry amount method shown in figure 3 above referring to Fig. 6, Fig. 6
600 structural schematic diagram is set, advertisement inventory's inquiry amount device 600 includes:
First determining module 601, for determining that targeted advertisements launch each corresponding inventory of advertisement position on the page；
Second determining module 602, for determining the corresponding theoretical available stock of targeted advertisements to be put；
Correction module 603, for being modified by losing coefficient theoretical available stock corresponding to the targeted advertisements,
Obtain the corresponding actually available inventory of the targeted advertisements, it is described lose coefficient characterization advertisement delivery system using duplicate removal rule into
To the influence degree of advertisement exposure amount when row advertisement is launched.
It optionally, is the embodiment of the present application referring to Fig. 7, Fig. 7 on the basis of advertisement inventory's inquiry amount device shown in Fig. 6
Another advertisement inventory inquiry amount device of offer；Advertisement inventory's inquiry amount device 700 further include:
Module 701 is obtained to be broadcast for the linear regression model (LRM) by pre-training according to the corresponding advertisement of the targeted advertisements
It puts probability and obtains that the targeted advertisements are corresponding to lose coefficient, wherein the linear regression model (LRM) is assembled for training by training data
Practice generation, the training dataset includes multiple training datas, and the training data includes that the corresponding advertisement of advertisement plays generally
Rate and lose coefficient.
It optionally, is the embodiment of the present application referring to Fig. 8, Fig. 8 on the basis of advertisement inventory's inquiry amount device shown in Fig. 7
Another advertisement inventory inquiry amount device of offer；Advertisement inventory's inquiry amount device 800 further include:
Training data determining module 801, for determining the training dataset；
Training module 802, for obtaining the linear regression in conjunction with training dataset training by least square method
Model.
Optionally, on the basis of advertisement inventory's inquiry amount device shown in Fig. 8, the training data determining module 801 has
Body is used for:
It obtains advertisement of the advertisement when launching and plays probability and the corresponding actual exposure amount of advertisement；
Obtain advertisement launch when due to duplicate removal reason caused by lose light exposure；
According to the corresponding actual exposure amount of advertisement and loss light exposure, determines that advertisement is corresponding and lose coefficient；
The corresponding advertisement of advertisement is played into probability and the corresponding coefficient of losing of advertisement as one group of training data, passes through multiple groups
Different training datas forms the training dataset.
It optionally, is the embodiment of the present application referring to Fig. 9, Fig. 9 on the basis of advertisement inventory's inquiry amount device shown in Fig. 8
Another advertisement inventory inquiry amount device of offer；Advertisement inventory's inquiry amount device 900 further include:
Training module 901 is updated, when changing for the duplicate removal window used by advertisement delivery system, then based on more
Duplicate removal window after new, collects another training dataset, based on linear regression described in another training dataset re -training
Model, in the advertisement delivery system.
It optionally, is the application implementation referring to Figure 10, Figure 10 on the basis of advertisement inventory's inquiry amount device shown in Fig. 6
Another advertisement inventory inquiry amount device that example provides；Advertisement inventory's inquiry amount device 1000 further include:
Acquisition module 1001, the history for acquiring advertisement delivery system launch data, and the history launches data and includes
The actual exposure amount of advertisement and the predetermined degree of exposure of advertisement in history release time section；
Third determining module 1002, for launching the actual exposure amount and advertisement of advertisement described in data according to the history
Predetermined degree of exposure ratio, determine described in lose coefficient.
Optionally, on the basis of advertisement inventory's inquiry amount device shown in Fig. 6, second determining module is specifically used for:
By high-water line algorithm, according to the fixed condition of targeted advertisements to be put, determine that the targeted advertisements are corresponding
Theoretical available stock.
Above-mentioned advertisement inventory inquiry amount device is corresponding to advertisement by means of losing coefficient when determining advertisement inventory's inquiry amount result
Theoretical available stock be modified, this lose coefficient can characterize advertisement delivery system using duplicate removal rule carry out advertisement dispensing
When to the influence degree of advertisement exposure amount, in this way, consider during determining advertisement inventory's inquiry amount result duplicate removal rule for
The influence of inquiry amount result guarantees that finally determining inquiry amount result is more accurate, reduces advertisement and lacks the generation for broadcasting phenomenon, guarantees to meet
The actual demand of advertiser.
Present invention also provides a kind of equipment for inquiry amount advertisement inventory, which is specifically as follows server, referring to
Figure 11, Figure 11 are a kind of structural schematic diagram of the server for inquiry amount advertisement inventory provided by the embodiments of the present application.The service
Device 1100 can generate bigger difference because configuration or performance are different, may include one or more central processing units
(central processing units, CPU) 1122 (for example, one or more processors) and memory 1132, one
The storage medium 1130 of a or more than one storage application program 1142 or data 1144 (such as deposit by one or more magnanimity
Store up equipment).Wherein, memory 1132 and storage medium 1130 can be of short duration storage or persistent storage.It is stored in storage medium
1130 program may include one or more modules (diagram does not mark), and each module may include in server
Series of instructions operation.Further, central processing unit 1122 can be set to communicate with storage medium 1130, in server
The series of instructions operation in storage medium 1130 is executed on 1100.
Server 1100 can also include one or more power supplys 1126, one or more wired or wireless nets
Network interface 1150, one or more input/output interfaces 1158, and/or, one or more operating systems 1141, example
Such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM etc..
The step as performed by server can be based on server architecture shown in the Figure 11 in above-described embodiment.
Wherein, CPU 1122 is for executing following steps:
Determine that targeted advertisements launch each corresponding inventory of advertisement position on the page；
Determine the corresponding theoretical available stock of targeted advertisements to be put；
It is modified by losing coefficient theoretical available stock corresponding to the targeted advertisements, obtains the targeted advertisements
Corresponding actually available inventory, the coefficient characterization advertisement delivery system of losing use duplicate removal rule when carrying out advertisement dispensing to wide
Accuse the influence degree of light exposure.
Optionally, any specific implementation side of advertisement inventory inquiry amount method in the embodiment of the present application can also be performed in CPU1122
The method and step of formula.
The embodiment of the present application also provides another equipment for being used for inquiry amount advertisement inventory, which is specifically as follows terminal
Equipment.As shown in figure 12, for ease of description, part relevant to the embodiment of the present application is illustrated only, particular technique details is not
It discloses, please refers to the embodiment of the present application method part.The terminal can be include mobile phone, tablet computer, personal digital assistant
(Personal Digital Assistant, PDA), point-of-sale terminal (Point of Sales, POS), vehicle-mounted computer etc. are any
Terminal device, taking the terminal as an example:
Figure 12 shows the block diagram of the part-structure of mobile phone relevant to terminal provided by the embodiments of the present application.With reference to figure
12, mobile phone includes: radio frequency (Radio Frequency, RF) circuit 1210, memory 1220, input unit 1230, display unit
1240, sensor 1250, voicefrequency circuit 1260, Wireless Fidelity (wireless fidelity, WiFi) module 1070, processor
The components such as 1280 and power supply 1290；It wherein, include input panel 1231 and other input equipments in input unit 1230
1232, include display panel 1241 in display unit 1240, includes loudspeaker 1261 and microphone 1262 in voicefrequency circuit 1260.
It, can be with it will be understood by those skilled in the art that handset structure shown in Figure 12 does not constitute the restriction to mobile phone
Including perhaps combining certain components or different component layouts than illustrating more or fewer components.
In the embodiment of the present application, processor 1280 included by the terminal is with the following functions:
Determine that targeted advertisements launch each corresponding inventory of advertisement position on the page；
Determine the corresponding theoretical available stock of targeted advertisements to be put；
It is modified by losing coefficient theoretical available stock corresponding to the targeted advertisements, obtains the targeted advertisements
Corresponding actually available inventory, the coefficient characterization advertisement delivery system of losing use duplicate removal rule when carrying out advertisement dispensing to wide
Accuse the influence degree of light exposure.
Optionally, any specific implementation of advertisement inventory inquiry amount method in the embodiment of the present application can also be performed in processor 1280
The method and step of mode.
The embodiment of the present application also provides a kind of computer readable storage medium, for storing program code, the program code
For executing any one embodiment in a kind of advertisement inventory inquiry amount method described in foregoing individual embodiments.
The embodiment of the present application also provides a kind of computer program product including instruction, when run on a computer,
So that computer executes any one embodiment in a kind of advertisement inventory inquiry amount method described in foregoing individual embodiments.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
In several embodiments provided herein, it should be understood that disclosed system, device and method can be with
It realizes by another way.For example, the apparatus embodiments described above are merely exemplary, for example, the unit
It divides, only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components
It can be combined or can be integrated into another system, or some features can be ignored or not executed.Another point, it is shown or
The mutual coupling, direct-coupling or communication connection discussed can be through some interfaces, the indirect coupling of device or unit
It closes or communicates to connect, can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, each functional unit in each embodiment of the application can integrate in one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, the technical solution of the application is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) executes the complete of each embodiment the method for the application
Portion or part steps.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (full name in English: Read-Only
Memory, english abbreviation: ROM), random access memory (full name in English: Random Access Memory, english abbreviation:
RAM), the various media that can store program code such as magnetic or disk.
The above, above embodiments are only to illustrate the technical solution of the application, rather than its limitations；Although referring to before
Embodiment is stated the application is described in detail, those skilled in the art should understand that: it still can be to preceding
Technical solution documented by each embodiment is stated to modify or equivalent replacement of some of the technical features；And these
It modifies or replaces, the spirit and scope of each embodiment technical solution of the application that it does not separate the essence of the corresponding technical solution.
Claims (13)
1. a kind of advertisement inventory inquiry amount method characterized by comprising
Determine that targeted advertisements launch each corresponding inventory of advertisement position on the page；
Determine the corresponding theoretical available stock of targeted advertisements to be put；
It is modified by losing coefficient theoretical available stock corresponding to the targeted advertisements, it is corresponding to obtain the targeted advertisements
Actually available inventory, it is described lose coefficient characterization advertisement delivery system using duplicate removal rule carry out advertisement dispensing when to advertisement expose
The influence degree of light quantity.
2. the method according to claim 1, wherein losing coefficient described in determining in the following manner:
By the linear regression model (LRM) of pre-training, it is wide that the probability acquisition target is played according to the corresponding advertisement of the targeted advertisements
It accuses and corresponding loses coefficient, wherein the linear regression model (LRM) is generated by training dataset training, the training data
Collection includes multiple training datas, and the training data includes that the corresponding advertisement of advertisement plays probability and loses coefficient.
3. according to the method described in claim 2, it is characterized in that, training the linear regression model (LRM) by the following method:
Determine the training dataset；
The linear regression model (LRM) is obtained in conjunction with training dataset training by least square method.
4. according to the method in claim 2 or 3, which is characterized in that determine the training dataset in the following manner, wrap
It includes:
It obtains advertisement of the advertisement when launching and plays probability and the corresponding actual exposure amount of advertisement；
Obtain advertisement launch when due to duplicate removal reason caused by lose light exposure；
According to the corresponding actual exposure amount of advertisement and loss light exposure, determines that advertisement is corresponding and lose coefficient；
The corresponding advertisement of advertisement is played into probability and the corresponding coefficient of losing of advertisement as one group of training data, passes through multiple groups difference
Training data form the training dataset.
5. according to the method described in claim 2, it is characterized in that, the duplicate removal window used by advertisement delivery system becomes
When change, then it is based on updated duplicate removal window, collects another training dataset, is based on another training dataset re -training
The linear regression model (LRM), in the advertisement delivery system.
6. the method according to claim 1, wherein losing coefficient described in determining in the following manner:
The history for acquiring advertisement delivery system launches data, and it includes advertisement in history release time section that the history, which launches data,
The predetermined degree of exposure of actual exposure amount and advertisement；
The ratio of the actual exposure amount of the advertisement in data and the predetermined degree of exposure of the advertisement is launched according to the history,
Coefficient is lost described in determination.
7. method according to any one of claim 1 to 6, which is characterized in that determination targeted advertisements to be put
Corresponding theory available stock, comprising:
The corresponding reason of the targeted advertisements is determined according to the fixed condition of targeted advertisements to be put by high-water line algorithm
By available stock.
8. a kind of advertisement inventory inquiry amount device characterized by comprising
First determining module, for determining that targeted advertisements launch each corresponding inventory of advertisement position on the page；
Second determining module, for determining the corresponding theoretical available stock of targeted advertisements to be put；
Correction module obtains institute for being modified by losing coefficient theoretical available stock corresponding to the targeted advertisements
The corresponding actually available inventory of targeted advertisements is stated, the coefficient characterization advertisement delivery system of losing is using the progress advertisement of duplicate removal rule
To the influence degree of advertisement exposure amount when dispensing.
9. device according to claim 8, which is characterized in that described device further include:
Module is obtained, for the linear regression model (LRM) by pre-training, probability is played according to the corresponding advertisement of the targeted advertisements
Obtain that the targeted advertisements are corresponding to lose coefficient, wherein the linear regression model (LRM) is generated by training dataset training
, the training dataset includes multiple training datas, and the training data includes that the corresponding advertisement of advertisement plays probability and folding
Damage coefficient.
10. device according to claim 8, which is characterized in that described device further include:
Acquisition module, the history for acquiring advertisement delivery system launch data, and it includes that history is launched that the history, which launches data,
The actual exposure amount of advertisement in period and the predetermined degree of exposure of advertisement；
4th determining module, for launching the actual exposure amount of advertisement and the predetermined exposure of advertisement described in data according to the history
The ratio of light quantity, determine described in lose coefficient.
11. a kind of equipment, which is characterized in that the equipment includes processor and memory:
Said program code is transferred to the processor for storing program code by the memory；
The processor is used for according to the instruction execution advertisement base of any of claims 1-7 in said program code
Deposit inquiry amount method.
12. a kind of computer readable storage medium, which is characterized in that the computer readable storage medium is for storing program generation
Code, said program code require advertisement inventory's inquiry amount described in any one of 1-7 for perform claim.
13. a kind of computer program product including instruction, which is characterized in that when run on a computer, so that described
Computer perform claim requires advertisement inventory's inquiry amount method described in any one of 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910364693.5A CN110163665A (en) | 2019-04-30 | 2019-04-30 | A kind of advertisement inventory inquiry amount method, apparatus, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910364693.5A CN110163665A (en) | 2019-04-30 | 2019-04-30 | A kind of advertisement inventory inquiry amount method, apparatus, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110163665A true CN110163665A (en) | 2019-08-23 |
Family
ID=67633226
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910364693.5A Pending CN110163665A (en) | 2019-04-30 | 2019-04-30 | A kind of advertisement inventory inquiry amount method, apparatus, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110163665A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111260398A (en) * | 2020-01-13 | 2020-06-09 | 腾讯科技（深圳）有限公司 | Advertisement putting control method and device, electronic equipment and storage medium |
| CN113723983A (en) * | 2020-05-26 | 2021-11-30 | 北京达佳互联信息技术有限公司 | Information processing method and device, server and storage medium |
| CN115456529A (en) * | 2022-09-13 | 2022-12-09 | 上海数禾信息科技有限公司 | Material inventory early warning method and system, computer equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102265299A (en) * | 2008-11-24 | 2011-11-30 | 埃德玛威尔公司 | Mobile ad optimization architecture |
| CN104899769A (en) * | 2015-06-30 | 2015-09-09 | 北京奇艺世纪科技有限公司 | Advertising position allocation method and device |
| CN107371048A (en) * | 2017-07-13 | 2017-11-21 | 北京奇艺世纪科技有限公司 | A kind of Forecasting Methodology and device of the stock of publicity orders |
| CN108256879A (en) * | 2016-12-28 | 2018-07-06 | 北京奇虎科技有限公司 | Rate adjusting method, device and equipment are launched in a kind of advertisement |
| CN108269109A (en) * | 2016-12-30 | 2018-07-10 | 深圳市优朋普乐传媒发展有限公司 | A kind of Advertisement arrangement injected volume equalization methods and device |
-
2019
- 2019-04-30 CN CN201910364693.5A patent/CN110163665A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102265299A (en) * | 2008-11-24 | 2011-11-30 | 埃德玛威尔公司 | Mobile ad optimization architecture |
| CN104899769A (en) * | 2015-06-30 | 2015-09-09 | 北京奇艺世纪科技有限公司 | Advertising position allocation method and device |
| CN108256879A (en) * | 2016-12-28 | 2018-07-06 | 北京奇虎科技有限公司 | Rate adjusting method, device and equipment are launched in a kind of advertisement |
| CN108269109A (en) * | 2016-12-30 | 2018-07-10 | 深圳市优朋普乐传媒发展有限公司 | A kind of Advertisement arrangement injected volume equalization methods and device |
| CN107371048A (en) * | 2017-07-13 | 2017-11-21 | 北京奇艺世纪科技有限公司 | A kind of Forecasting Methodology and device of the stock of publicity orders |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111260398A (en) * | 2020-01-13 | 2020-06-09 | 腾讯科技（深圳）有限公司 | Advertisement putting control method and device, electronic equipment and storage medium |
| CN111260398B (en) * | 2020-01-13 | 2023-11-07 | 腾讯科技（深圳）有限公司 | Advertisement putting control method and device, electronic equipment and storage medium |
| CN113723983A (en) * | 2020-05-26 | 2021-11-30 | 北京达佳互联信息技术有限公司 | Information processing method and device, server and storage medium |
| CN115456529A (en) * | 2022-09-13 | 2022-12-09 | 上海数禾信息科技有限公司 | Material inventory early warning method and system, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jiao et al. | Toward an automated auction framework for wireless federated learning services market | |
| CN109902849B (en) | User behavior prediction method and device, and behavior prediction model training method and device | |
| US10929879B2 (en) | Method and apparatus for identification of fraudulent click activity | |
| CN110163665A (en) | A kind of advertisement inventory inquiry amount method, apparatus, equipment and storage medium | |
| CN104572278B (en) | The method, device and equipment of light application calling local side ability | |
| CN106878415B (en) | Load balancing method and device for data consumption | |
| CN109889397B (en) | Lottery method, block generation method, equipment and storage medium | |
| CN115018019B (en) | Model training method and system based on federal learning and storage medium | |
| CN106815254A (en) | A kind of data processing method and device | |
| CN110427494B (en) | Knowledge graph display method and device, storage medium and electronic device | |
| US20240064348A1 (en) | Image analysis system | |
| CN110874626B (en) | Quantization method and quantization device | |
| CN107798332B (en) | User behavior prediction method and device | |
| CN110460650A (en) | The decision-making technique and device of unloading are calculated under multiple edge server scene | |
| US20230281656A1 (en) | Data processing method and related apparatus | |
| CN110533459A (en) | A kind of terminal replacement probability prediction technique and device based on gradient boosting algorithm | |
| CN108174227A (en) | Display methods, device and the storage medium of virtual objects | |
| CN110381378A (en) | Projectional technique, device, server and the storage medium of network multimedia playback volume | |
| CN108810056A (en) | Information-pushing method and device | |
| CN108734366A (en) | User identification method and its system | |
| CN109947850A (en) | Data distributing method, device and equipment | |
| CN106998341A (en) | A kind of information processing method and server | |
| US20230153222A1 (en) | Scaled-down load test models for testing real-world loads | |
| CN112738815B (en) | Method and device for evaluating number of accessible users | |
| CN113038242B (en) | Method, device and equipment for determining display position of live broadcast card and storage 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 |