Summary of the invention
The embodiment of the invention discloses a kind of information-pushing method and device, with improve successfully to client push all toThe probability of pushed information.
In order to achieve the above objectives, the embodiment of the invention discloses a kind of information-pushing methods, which comprises
It receives the information that client is sent and pushes request；
It is pushed and is requested according to the information, determine at least two information to be pushed；
The jump mistake rate of information to be pushed determined by obtaining；
The recommendation determined according to the jump mistake rate of recommendation coefficient and the pushed information by identified each information to be pushedDegree obtains the push sequence of identified pushed information；
According to push sequence obtained, successively to pushed information determined by the client push.
In a specific embodiment of the present invention, the jump mistake rate of information to be pushed determined by the acquisition, comprising:
From preset jump mistake rate list, the jump mistake rate of identified information to be pushed is obtained；
Wherein, the preset jump mistake rate list obtains in the following manner:
According to preset classifying rules, according to the corresponding push parameter of pushed information each in preset period of time, determining shouldThe generic of pushed information；
Count the jump mistake rate that each pushed information in the preset period of time is directed to its generic；
The jump mistake rate obtained according to above-mentioned statistics generates and jumps the list of mistake rate.
In a specific embodiment of the present invention, in the jump mistake rate obtained according to above-mentioned statistics, generates to jump and loseAfter rate list, further includes:
Detect the element in jump mistake rate list generated with the presence or absence of value for sky；
If it exists, it is not the value of empty element according to value in the generic that each value is empty element, updatesThe value is the value of empty element.
In a specific embodiment of the present invention, the preset classifying rules, including it is following rule at leastIt is a kind of:
The classifying rules of video length is playing for the client；
For the classifying rules of current pushed information duration；
For the classifying rules of the push serial number of current pushed information；
For the classifying rules of the push serial number pushed information adjacent with the push serial number of current pushed information.
It include for current push letter in the preset classifying rules in a specific embodiment of the present inventionIn the case where the classifying rules of the push serial number of breath,
The basis is determined by the recommendation coefficient of identified each information to be pushed and the jump mistake rate of the pushed informationRecommendation obtains the push sequence of identified pushed information, comprising:
According to the sequence that push serial number n is ascending, following steps are passed sequentially through from identified N number of information to be pushedMiddle selection pushes the corresponding information to be pushed of serial number n:
Not selected information to be pushed in N number of information to be pushed determined by obtaining；
It is above-mentioned each not selected wait push away in the case where selecting to push serial number from preset jump mistake rate list as nThe corresponding jump mistake rate of breath of delivering letters；
According to expression formula P (1-BR), the corresponding recommendation of above-mentioned each not selected information to be pushed is estimated,In, the P indicates the recommendation coefficient of information to be pushed, and the BR indicates described and preset jumps record in the list of mistake rate, pushIn the case that serial number is n, the jump mistake rate of information to be pushed；
According to above-mentioned estimation result, the highest information to be pushed of recommendation for selecting estimation to obtain is as push serial number nCorresponding information to be pushed.
In a specific embodiment of the present invention, the information-pushing method further include:
To after the selected each information to be pushed of the client push, according to should be to the client pushThe push serial number of pushed information and user are directed to the feedback of the information to be pushed, update the preset jump mistake rate list.
In a specific embodiment of the present invention, the recommendation coefficient of identified each information to be pushed, are as follows:
The recommendation coefficient obtained according to the price of identified each information to be pushed；Or
According to the recommendation system of each information to be pushed determined by the corresponding historical viewings record acquisition of the clientNumber.
In order to achieve the above objectives, the embodiment of the invention discloses a kind of information push-delivery apparatus, described device includes:
Request receiving module is pushed, for receiving the information push request of client transmission；
Pushed information determining module requests for being pushed according to the information, determines at least two information to be pushed；
It jumps mistake rate and obtains module, for obtaining the jump mistake rate of identified information to be pushed；
Push sequence obtains module, for according to the recommendation coefficient and push letter by identified each information to be pushedThe recommendation of breath jumping mistake rate and determining obtains the push sequence of identified pushed information；
Info push module, for successively being pushed away to determined by the client push according to push sequence obtainedIt delivers letters breath.
In a specific embodiment of the present invention, described obtain jumps mistake rate acquisition module, is specifically used for from preset jumpIn mistake rate list, the jump mistake rate of identified information to be pushed is obtained；
The information push-delivery apparatus further include: jump the list of mistake rate and obtain module；
Wherein, the jump mistake rate list obtains module, comprising:
Classification determines submodule, is used for according to preset classifying rules, according to pushed information pair each in preset period of timeThe push parameter answered determines the generic of the pushed information；
Mistake rate statistic submodule is jumped, is directed to its generic for counting each pushed information in the preset period of timeJump mistake rate；
It jumps the list of mistake rate and generates submodule, the jump mistake rate for being obtained according to above-mentioned statistics generates and jumps the list of mistake rate.
In a specific embodiment of the present invention, the jump mistake rate list obtains module, further includes:
Element detection submodule, for detecting after the jump mistake rate list generates submodule generation jump mistake rate listThe element generated jumped in the list of mistake rate with the presence or absence of value for sky；
Element update submodule, for the testing result of the Element detection submodule be in the case where, according to everyA value is the value that value is not empty element in the generic of empty element, and updating the value is taking for empty elementValue.
In a specific embodiment of the present invention, the preset classifying rules, including it is following rule at leastIt is a kind of:
The classifying rules of video length is playing for the client；
For the classifying rules of current pushed information duration；
For the classifying rules of the push serial number of current pushed information；
For the classifying rules of the push serial number pushed information adjacent with the push serial number of current pushed information.
In a specific embodiment of the present invention, the push sequence obtains module, is specifically used for described defaultClassifying rules in include for current pushed information push serial number classifying rules in the case where, according to push serial numberN ascending sequence, successively selection pushes the corresponding information to be pushed of serial number n from identified N number of information to be pushed；
The push sequence obtains module, comprising:
Information to be pushed obtains submodule, not selected wait push in identified N number of information to be pushed for obtainingInformation；
It jumps mistake rate and selects submodule, in the case where selecting to push serial number from preset jump mistake rate list as n, onState the corresponding jump mistake rate of each not selected information to be pushed；
Recommendation estimates submodule, for estimating above-mentioned each not selected wait push according to expression formula P (1-BR)The corresponding recommendation of information, wherein the P indicates that the recommendation coefficient of information to be pushed, the BR indicate that the preset jump is lostIn the case that record in rate list, push serial number is n, the jump mistake rate of information to be pushed；
Information to be pushed selects submodule, for according to above-mentioned estimation result, the recommendation that selection estimation obtains to be highestInformation to be pushed is as the corresponding information to be pushed of push serial number n.
In a specific embodiment of the present invention, the information push-delivery apparatus further include:
Mistake rate list update module is jumped, for after the selected each information to be pushed of the client push,According to the feedback for being directed to the information to be pushed to the push serial number of the client push information to be pushed and user, moreThe new preset jump mistake rate list.
In a specific embodiment of the present invention, the recommendation coefficient of identified each information to be pushed, are as follows:
The recommendation coefficient obtained according to the price of identified each information to be pushed；Or
According to the recommendation system of each information to be pushed determined by the corresponding historical viewings record acquisition of the clientNumber.
As seen from the above, in scheme provided in an embodiment of the present invention, server is according to the received information push request of institute, reallyAfter making information to be pushed, obtain determined by information to be pushed jump mistake rate, then further according to by it is identified each toPushed information recommends coefficient and the recommendation of the pushed information jumping mistake rate and determining, the push of pushed information determined by acquisitionSequentially, and according to push sequence obtained, successively to pushed information determined by client push.By being determined in acquisitionInformation to be pushed push sequence when, it is contemplated that the jump mistake rate of each information to be pushed, and jump mistake rate and be able to reflect serverWhen to client push information, successfully push the probability of the information, therefore, can according to the jump mistake rate of each information to be pushed,Preferentially push and successfully push the higher information to be pushed of probability, and then can be improved successfully to the whole letter to be pushed of client pushThe probability of breath.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, completeSite preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based onEmbodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every otherEmbodiment shall fall within the protection scope of the present invention.
Fig. 1 is a kind of flow diagram of information-pushing method provided in an embodiment of the present invention, this method comprises:
S101: it receives the information that client is sent and pushes request.
Wherein, in above-mentioned information push request, may include:
Client is playing the type of video, for example, movement class film, the TV play of describing love affairs class, costume piece, modern play,Hong Kong and Taiwan films, American-European film etc.；
Client is playing the duration of video, for example, 50 minutes, 150 minutes etc.；
Client is playing the playback progress of video, such as: 0 point 0 second, 30 points 15 seconds etc..
It should be noted that information included in above-mentioned information push request is not limited to that, in practical applicationIt can be determines according to actual conditions.
S102: it is pushed and is requested according to information, determine at least two information to be pushed.
Server receive client transmission above- mentioned information push request after, according to the information push request, determine toPushed information, information to be pushed can be advertising information, news information, weather forecast information etc..
Specifically, the type that the client for including in request is playing video can be pushed according to information, user is estimatedThen the information such as hobby, age determine information to be pushed according to the information that estimation obtains.
For example, if the type that client is playing video is Hong Kong and Taiwan films, describing love affairs class TV play, in this way, user mayIt is young woman, then can determines information to be pushed are as follows: cosmetics, clothing advertising information etc..
Furthermore it is also possible to according to pushing, the client for including in request is playing the duration of video and client is being broadcastThe playback progress for putting video, determines information to be pushed.
It is understood that user does not like usually and is disturbed when watching video, if not understanding in video also in userHold, in the case where not attracted by client video being played on, to the pushed information of the longer duration of client push, Yi YinqiUser's dislike, so, it is long when can push shorter to push away when the playback progress that client is playing video is less than preset thresholdIt delivers letters breath, as the playback progress that client is playing video increases, in view of factors such as developments of action, user usually will not be byIn do not like viewing pushed information and close the video being currently viewed, cause pushed information to fail, so, such caseUnder, it can suitably push the pushed information of longer duration.
For example, client video being played on when it is 60 minutes a length of in the case where,
If information push request in include client be playing video playback progress be 0 point 0 second, can only to15 seconds advertising informations of client push；
If client be playing video playback progress be 25 points 0 second, can be to client push advertisement in 30 secondsInformation；
If client be playing video playback progress be 50 points 20 seconds, can be to client push advertisement in 60 secondsInformation.
S103: the jump mistake rate of information to be pushed determined by obtaining.
Under normal conditions, mistake rate is jumped, refers to that display user enters the page by corresponding entrance, but only has accessed a pageThe access times that face is just left account for the ratio of the total access times of the page.
Defined with above-mentioned jump mistake rate it is similar, in the application, can by jump mistake rate understand are as follows: in preset period of time, serviceDevice accounts for the ratio to the client push information to be pushed total degree to the number of any information to be pushed failure of client push.
Wherein, server is to client push information to be pushed, and information to be pushed success is in client terminal playingSituation belongs to server to the success of client push information；
Server does not succeed due to selecting " skipping " etc. user to client push information to be pushedThe case where client terminal playing, belongs to server and fails to client push information.
It, can be with specifically, the jump mistake rate of identified information to be pushed can be obtained from preset jump mistake rate listAccording to the push times relevant to identified information to be pushed recorded in server, e.g., push should be wait push away in preset period of timeThe number etc. that information to be pushed failure is pushed in the successful number of breath of delivering letters, preset period of time is calculated, certainly, identifiedThe jump mistake rate of information to be pushed can also obtain by other means, and the application is defined not to this.
In a kind of optional implementation of the invention, when obtaining jump mistake rate from preset jump mistake rate list, Ke YitongThe method for crossing embodiment illustrated in fig. 2 offer, which obtains, jumps the list of mistake rate.
S104: it is determined according to the jump mistake rate of recommendation coefficient and the pushed information by identified each information to be pushedRecommendation obtains the push sequence of identified pushed information.
Specifically, the recommendation coefficient of identified each information to be pushed can be according to identified letter each to be pushedThe recommendation coefficient that the price of breath obtains, from the angle for maximizing media platform income, the price of information to be pushed is higher, clothesBusiness device more tends to the client push information, so that media platform can obtain higher income.
The recommendation coefficient of identified each information to be pushed can also be according to the corresponding historical viewings record of clientThe recommendation coefficient of the identified each information to be pushed obtained.Specifically, can be remembered according to the corresponding historical viewings of clientThe user informations such as hobby, age, the gender for determining user are recorded, then according between information to be pushed and user informationWith degree, the recommendation coefficient of identified each information to be pushed is obtained.In this way, same information to be pushed is pushed to by serverWhen different client, determine that obtained recommendation coefficient may be different.
For example, user likes seeing describing love affairs class TV play, can determine that the use from the point of view of the corresponding historical viewings record of clientFamily may be young woman, and identified information to be pushed is respectively as follows: automotive-type advertising information, shopping series advertisements information, health careCategory advertising information, and under normal conditions, young woman is interested in the shopping series advertisements in above-mentioned three series advertisements, instituteWith, the recommendation coefficient of these three types of information to be pushed, can according to shopping series advertisements, health care products advertisement, automobile series advertisements it is suitableSequence is set as the numerical value being sequentially reduced.
S105: according to push sequence obtained, successively to pushed information determined by client push.
In practical application, when in order to guarantee to obtain the jump mistake rate of identified information to be pushed, it can obtain relatively accurateData, in a preferred embodiment of the present invention, after each information to be pushed selected to client push,Can also according to the feedback for being directed to the information to be pushed to the push serial number of the client push information to be pushed and user,Update preset jump mistake rate list.
As seen from the above, in scheme provided in this embodiment, server is determined according to the received information push request of instituteAfter information to be pushed, obtain determined by information to be pushed jump mistake rate, then further according to by identified each wait pushInformation recommends coefficient and the recommendation of the pushed information jumping mistake rate and determining, the push of pushed information determined by acquisition is suitableSequence, and according to push sequence obtained, successively to pushed information determined by client push.As determined by the acquisitionWhen the push sequence of information to be pushed, it is contemplated that the jump mistake rate of each information to be pushed, and jump mistake rate be able to reflect server toWhen client push information, the probability of the information is successfully pushed, it therefore, can be excellent according to the jump mistake rate of each information to be pushedThe successfully push higher information to be pushed of probability is first pushed, and then can be improved successfully to client push whole information to be pushedProbability.
It should be noted that under normal conditions, media platform by server to client push information when, usually pressIt charges according to success to the number of user's pushed information, is root in the recommendation coefficient of identified each information to be pushed thereforeIn the case where the recommendation coefficient obtained according to the price of identified each information to be pushed, implement the side provided according to the present inventionCase, in addition to can be improved successfully to the probability of client push whole information to be pushed other than, it can also be ensured that media platformMaximum revenue.
In one particular embodiment of the present invention, referring to fig. 2, a kind of stream for obtaining and jumping the method for mistake rate list is providedJourney schematic diagram, this method comprises:
S201: according to preset classifying rules, according to the corresponding push parameter of pushed information each in preset period of time, reallyThe generic of the fixed pushed information.
Above-mentioned preset classifying rules may include at least one of following rule:
The classifying rules of video length is playing for client；
For the classifying rules of current pushed information duration；
For the classifying rules of the push serial number of current pushed information；
For the classifying rules of the push serial number pushed information adjacent with the push serial number of current pushed information.
For example, video length T can be playing according to clientVIt is divided into following a few classes:
TV≤ 5 minutes classifications, 5 minutes < TV≤ 10 minutes classifications, 10 minutes < TV≤ 30 minutes classifications, 30 pointsClock < TV≤ 60 minutes classifications, TV> 60 minutes classifications etc..
It can be according to current pushed information duration TPIt is divided into following a few classes:
TP≤ 15 seconds classifications, 15 seconds < TP≤ 30 seconds classifications, 30 seconds < TP≤ 45 seconds classifications, 45 seconds < TP≤60The classification of second, TP> 60 seconds classifications etc..
When each may include: to the client push pushed information in the corresponding push parameter of pushed information, clientHold the duration for being playing video, the duration of the pushed information, the push sequence of the pushed information, the pushed information in clientPlay position etc..
S202: each pushed information is directed to the jump mistake rate of its generic in statistics preset period of time.
S203: the jump mistake rate obtained according to above-mentioned statistics generates and jumps the list of mistake rate.
It is empty element according to that may have value in the jump mistake rate list of above method generation in practical application, thisSample, when obtaining jump mistake rate from preset jump mistake rate list, not can guarantee the jump mistake rate obtained every time is that effective jump is lostRate, in consideration of it, in the jump mistake rate obtained according to above-mentioned statistics, generating in a kind of relatively good implementation of the invention and jumping mistake rateAfter list, the element in jump mistake rate list generated with the presence or absence of value for sky can also be detected, and if it exists, according to eachValue is the value that value is not empty element in the generic of empty element, updates the value that the value is empty element.
It should be noted that above-mentioned jump mistake rate list generated can be the storage in the form of a bivariate table, it canIt is stored in the form of multiple bivariate tables by be, in addition, the data in the jump mistake rate list are in addition to can be with table dataForm present outer, can also be presented in the form of multi-dimensional matrix.The application is not compared and is defined.
In a specific embodiment of the invention, referring to Fig. 3, the process signal of another information-pushing method is providedFigure, compared with previous embodiment, in the present embodiment, in preset classifying rules include for current pushed information push it is suitableIn the case where the classifying rules of serial number,
The recommendation determined according to the jump mistake rate of recommendation coefficient and the pushed information by identified each information to be pushedDegree obtains the push sequence of identified pushed information, comprising:
S104A: the initial value of setting push serial number n is 1.
S104B: judging whether n is greater than the quantity N of identified information to be pushed, if it has not, executing S104C, S105.
It should be noted that the initial value of n may be set to be other values in practical application, judged in this stepWhen, need to judge whether n is greater than (initial value of n+identified information to be pushed quantity -1).
S104C: not selected information to be pushed in N number of information to be pushed determined by obtaining.
S104D: selected from preset jump mistake rate list push serial number for n in the case where, it is above-mentioned each not to be selectedThe corresponding jump mistake rate of information to be pushed.
S104E: according to expression formula P (1-BR), the corresponding recommendation of above-mentioned each not selected information to be pushed is estimatedDegree.
Wherein, P indicate information to be pushed recommendation coefficient, BR indicate it is preset jump the list of mistake rate in record, push it is suitableIn the case where serial number n, the jump mistake rate of information to be pushed.
S104F: according to above-mentioned estimation result, the highest information to be pushed of recommendation for selecting estimation to obtain is suitable as pushingThe corresponding information to be pushed of serial number n.
S104G: the value for updating n is n+1, and returns to S104B.
As seen from the above, in scheme provided in this embodiment, in identified push sequence, suitable number of each push is correspondingInformation to be pushed relative to the corresponding pushed information recommendation of each push serial number be thereafter it is highest, therefore, can be into oneStep improves successfully to the probability of client push whole information to be pushed.
Above-mentioned S104A to S104G is illustrated below by a specific example.
It is assumed that identified information to be pushed is respectively as follows: InfP1、InfP2、InfP3, then determined by information to be pushedQuantity N is 3, and each information to be pushed recommends coefficient and its jump mistake rate, referring to such as the following table 1.
Table 1
According to S104A: the initial value of setting push serial number n is 1.
According to S104B: due to n=1 at this time, it is seen that n < 3 is not more than 3, executes S104C.
According to S104C: not selected information to be pushed includes: InfP1、InfP2、InfP3。
According to S104D:InfP1、InfP2、InfP3, in the case where pushing serial number is 1, jump mistake rate be respectively as follows: 0.6,0.7、0.3。
According to S104E:InfP1Recommendation=5x (1-0.6)=2.0, InfP2Recommendation=8x (1-0.7)=2.4,InfP3Recommendation=3x (1-0.3)=2.1.
According to S104F: according to above-mentioned estimation result, selecting InfP2As the push corresponding information to be pushed of serial number 1.
According to S104G: the value for updating n is 2, returns to S104B.
According to S104B: due to n=2 at this time, it is seen that n < 3 is not more than 3, executes S104C.
According to S104C: not selected information to be pushed includes: InfP1、InfP3。
According to S104D:InfP1、InfP3, in the case where pushing serial number is 2, jumps mistake rate and be respectively as follows: 0.5,0.4.
According to S104E:InfP1Recommendation=5x (1-0.5)=2.5, InfP3Recommendation=3x (1-0.4)=1.8.
According to S104F: according to above-mentioned estimation result, selecting InfP1As the push corresponding information to be pushed of serial number 2.
According to S104G: the value for updating n is 3, returns to S104B.
According to S104B: due to n=3 at this time, it is seen that n=3 is not more than 3, executes S104C.
According to S104C: not selected information to be pushed includes: InfP3。
According to S104D:InfP3, in the case where pushing serial number is 3, jumps mistake rate and be respectively as follows: 0.6.
According to S104E:InfP3Recommendation=3x (1-0.6)=1.2.
According to S104F: according to above-mentioned estimation result, selecting InfP3As the push corresponding information to be pushed of serial number 3.
According to S104G: the value for updating n is 4, returns to S104B.
According to S104B: due to n=4 at this time, it is seen that n > 4 executes S105.
So far, the push sequence of pushed information determined by can get are as follows: InfP2、InfP1、InfP3。
Corresponding with above-mentioned information-pushing method, the embodiment of the invention also provides a kind of information push-delivery apparatus.
Fig. 4 is a kind of structural schematic diagram of information push-delivery apparatus provided in an embodiment of the present invention, which includes:
Request receiving module 401 is pushed, for receiving the information push request of client transmission；
Pushed information determining module 402 requests for being pushed according to the information, determines at least two information to be pushed；
It jumps mistake rate and obtains module 403, for obtaining the jump mistake rate of identified information to be pushed；
Push sequence obtains module 404, for being pushed away according to the recommendation coefficient by identified each information to be pushed with thisThe recommendation that the jump mistake rate of breath of delivering letters determines, obtains the push sequence of identified pushed information；
Info push module 405 is used for according to push sequence obtained, successively to determined by the client pushPushed information.
Specifically, the jump mistake rate acquisition module 403 that obtains can be specifically used for from preset jump mistake rate list, it is true to obtain instituteThe jump mistake rate of fixed information to be pushed；
In this case, above-mentioned information push-delivery apparatus can also include: and jump the list of mistake rate to obtain module；
Wherein, the jump mistake rate list obtains module, comprising:
Classification determines submodule, is used for according to preset classifying rules, according to pushed information pair each in preset period of timeThe push parameter answered determines the generic of the pushed information；
Mistake rate statistic submodule is jumped, is directed to its generic for counting each pushed information in the preset period of timeJump mistake rate；
It jumps the list of mistake rate and generates submodule, the jump mistake rate for being obtained according to above-mentioned statistics generates and jumps the list of mistake rate.
Optionally, the jump mistake rate list, which obtains module, to include:
Element detection submodule, for detecting after the jump mistake rate list generates submodule generation jump mistake rate listThe element generated jumped in the list of mistake rate with the presence or absence of value for sky；
Element update submodule, for the testing result of the Element detection submodule be in the case where, according to everyA value is the value that value is not empty element in the generic of empty element, and updating the value is taking for empty elementValue.
Optionally, the preset classifying rules may include at least one of following rule:
The classifying rules of video length is playing for the client；
For the classifying rules of current pushed information duration；
For the classifying rules of the push serial number of current pushed information；
For the classifying rules of the push serial number pushed information adjacent with the push serial number of current pushed information.
In an alternate embodiment of the present invention where, the information push-delivery apparatus further include:
Mistake rate list update module is jumped, for after the selected each information to be pushed of the client push,According to the feedback for being directed to the information to be pushed to the push serial number of the client push information to be pushed and user, moreThe new preset jump mistake rate list.
Specifically, the recommendation coefficient of each information to be pushed determined by described can be with are as follows:
The recommendation coefficient obtained according to the price of identified each information to be pushed；Or
According to the recommendation system of each information to be pushed determined by the corresponding historical viewings record acquisition of the clientNumber.
As seen from the above, in scheme provided in this embodiment, server is determined according to the received information push request of instituteAfter information to be pushed, obtain determined by information to be pushed jump mistake rate, then further according to by identified each wait pushInformation recommends coefficient and the recommendation of the pushed information jumping mistake rate and determining, the push of pushed information determined by acquisition is suitableSequence, and according to push sequence obtained, successively to pushed information determined by client push.As determined by the acquisitionWhen the push sequence of information to be pushed, it is contemplated that the jump mistake rate of each information to be pushed, and jump mistake rate be able to reflect server toWhen client push information, the probability of the information is successfully pushed, it therefore, can be excellent according to the jump mistake rate of each information to be pushedThe successfully push higher information to be pushed of probability is first pushed, and then can be improved successfully to client push whole information to be pushedProbability.
In one particular embodiment of the present invention, referring to Fig. 5, the structural representation of another information push-delivery apparatus is providedFigure, compared with foregoing individual embodiments, in the present embodiment,
The push sequence obtains module 404, specifically for including being directed to work as to be pushed forward in the preset classifying rulesDeliver letters breath push serial number classifying rules in the case where, according to the ascending sequence of push serial number n, successively from reallyThe corresponding information to be pushed of selection push serial number n in fixed N number of information to be pushed；
The push sequence obtains module 404, comprising:
Information to be pushed obtains submodule 4041, in N number of information to be pushed determined by obtaining it is not selected toPushed information；
It jumps mistake rate and selects submodule 4042, push the case where serial number is n for selecting from preset jumps mistake rate listUnder, the corresponding jump mistake rate of above-mentioned each not selected information to be pushed；
Recommendation estimates submodule 4043, for according to expression formula P (1-BR), estimate it is above-mentioned it is each it is not selected toThe corresponding recommendation of pushed information, wherein the P indicates that the recommendation coefficient of information to be pushed, the BR indicate described presetIn the case where record in the list of mistake rate, push serial number is jumped as n, the jump mistake rate of information to be pushed；
Information to be pushed selects submodule 4044, for according to above-mentioned estimation result, the recommendation that selection estimation obtains to be mostHigh information to be pushed is as the corresponding information to be pushed of push serial number n.
As seen from the above, in scheme provided in this embodiment, in identified push sequence, suitable number of each push is correspondingInformation to be pushed relative to the corresponding pushed information recommendation of each push serial number be thereafter it is highest, therefore, can be into oneStep improves successfully to the probability of client push whole information to be pushed.
For device embodiment, since it is substantially similar to the method embodiment, related so being described relatively simplePlace illustrates referring to the part of embodiment of the method.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a realityBody or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operationIn any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended toNon-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including thoseElement, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipmentIntrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded thatThere is also other identical elements in process, method, article or equipment including the element.
Those of ordinary skill in the art will appreciate that all or part of the steps in realization above method embodiment is canIt is completed with instructing relevant hardware by program, the program can store in computer-readable storage medium,The storage medium designated herein obtained, such as: ROM/RAM, magnetic disk, CD.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.It is allAny modification, equivalent replacement, improvement and so within the spirit and principles in the present invention, are all contained in protection scope of the present inventionIt is interior.