CN103716338B - A kind of information-pushing method and device - Google Patents
A kind of information-pushing method and device Download PDFInfo
- Publication number
- CN103716338B CN103716338B CN201210370622.4A CN201210370622A CN103716338B CN 103716338 B CN103716338 B CN 103716338B CN 201210370622 A CN201210370622 A CN 201210370622A CN 103716338 B CN103716338 B CN 103716338B
- Authority
- CN
- China
- Prior art keywords
- user
- information
- pushed
- historical
- degree
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention discloses a kind of information-pushing method and device, this method includes：Obtain the user information of the user information and historical user of active user；According to the user information of the active user, the user information of historical user, the similar users circle of the active user is obtained, the similar users circle includes one or more historical users related with the active user；Preference degree of the historical user in the similar users circle to each information to be pushed is obtained, and information to be pushed is chosen according to the preference degree and is pushed to the active user.Using the present invention, in process of information push the characteristics of consideration user, targetedly user is pushed into row information.
Description
Technical field
The present invention relates to internet arena more particularly to a kind of information-pushing methods and device.
Background technology
With the continuous development of Internet technology, there are more and more information interchanges on the internet.Such as, service provider can
To issue various advertisements on the internet, and these advertisements then can be directly pushed to use by the distributor for possessing user resources
Family.
Push (Push) technology is a kind of to send information to client by server active based on client server mechanism
Technology, transmission information be typically user institute in advance it is scheduled.Compared with traditional pull technology (PULL), the former is by taking
Business device actively sends information, and the latter is then by client computer active request information.The advantage of push technology is the active of information
Property and promptness, before user plane can be pushed information at any time.
Existing internet information push may purposefully, on time by user be felt mainly using push (Push) technology
The information of interest is actively sent in the computer of user.It is like broadcasting station broadcast, " push " technology actively will be newest
News and data are pushed to client, and user need not internet searching.The major advantage of Push technologies be user is required it is low, generally
Suitable for the public, special technology is not required；Second is that promptness is good, information source is constantly updated to user " push " in time
Multidate information.
But in existing internet information push technology, the fresh demand for considering user less, only being pushed away to user simply
It send.On the one hand, meaningless Internet resources are occupied；On the other hand, user to these information and may lose interest in, and also result in
The waste of Internet resources.
Invention content
Technical problem to be solved of the embodiment of the present invention is, provides a kind of information-pushing method and device.In information
The characteristics of user is considered during push targetedly pushes user into row information.
In order to solve the above-mentioned technical problem, an embodiment of the present invention provides a kind of information-pushing methods, including：
Obtain the user information of the user information and historical user of active user；
According to the user information of the active user, the user information of historical user, the similar of the active user is obtained
User encloses, and the similar users circle includes one or more historical users related with the active user；
Preference degree of the historical user in the similar users circle to each information to be pushed is obtained, and according to the preference degree
It chooses information to be pushed and is pushed to the active user.
In another aspect, the embodiment of the present invention additionally provides a kind of information push-delivery apparatus, including：
Data obtaining module, the user information of user information and historical user for obtaining active user；
User encloses acquisition module, for user information, the user information of historical user according to the active user, obtains
The similar users circle of the active user, the similar users circle include one or more history related with the active user
User；
Preference degree acquisition module, the hobby for obtaining the historical user in the similar users circle to each information to be pushed
Degree；
Pushing module is pushed to the active user for choosing information to be pushed according to the preference degree.
Implement the embodiment of the present invention, has the advantages that：When being pushed to a certain user into row information, the use is investigated
The information of the relevant historical user at family forms similar users circle related with the user；According to the historical behavior of similar users circle
After these users are investigated to the preference degree of pushed information, pushed into row information further according to preference degree.So, it is pushed to user
Information just have the demand that larger probability meets user, improve the accuracy of push, also improve the utilization of Internet resources
Rate.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Obtain other attached drawings according to these attached drawings.
Fig. 1 is an idiographic flow schematic diagram of the information-pushing method in the embodiment of the present invention；
Fig. 2 is a concrete composition schematic diagram of the information push-delivery apparatus in the embodiment of the present invention；
Fig. 3 is a concrete composition schematic diagram of user's circle acquisition module in the embodiment of the present invention；
Fig. 4 is a concrete composition schematic diagram of the preference degree acquisition module in the embodiment of the present invention.
Specific implementation mode
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, complete
Site preparation describes, 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 on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other without creative efforts
Embodiment shall fall within the protection scope of the present invention.
In embodiments of the present invention be based on a basic ideas, i.e., when some user carry out network behavior (e.g., searching resource,
Watch online multimedia data) when, the behavior of user's circle similar with the user is investigated, what which included is to go through
History user, these historical users once carried out various network behaviors, then according to the network behavior of these historical users come
Predict that the information that some user may like just has higher accuracy, e.g., click preference according to the search of historical user and calculate
Hobby index predicts advertisement that user's maximum possible is liked, may like the film seen, the webpage etc. that may like browsing.
Wherein, can be by the various dimensions to user when obtaining similar users circle, such as educational background, region, age, receipts
Enter, education degree etc. carries out the division of similar users circle.Certainly, according to the opportunity of pushed information, pushed information etc.
Difference can also have more dimensions to select.Then each specific embodiment of the present invention is described below.
As shown in Figure 1, for an idiographic flow schematic diagram of the information-pushing method in the embodiment of the present invention.The flow packet
Include following steps.
101, the user information of the user information and historical user of active user is obtained.As previously mentioned, the user information can
With the various aspects information including user, e.g., educational background, age, region, income, occupation, hobby etc..
The user can refer to the user, the user for logging in some Web Community, some search engine for logging in some application
Login user, even some terminal device network address etc., carry out the user of a certain network behavior.Certainly, if
It is the network address of some terminal device, then when collecting user information, then can investigates once logged by the network address
Some user relevant information.
For the ease of subsequent calculating, which can be the array of numeralization, and the array is referred to as user's letter
Array is ceased, the element in the user information array corresponds to the numeralization value of one or more attributes of user.
102, according to the user information of the active user, the user information of historical user, obtain the active user's
Similar users circle, the similar users circle include one or more historical users related with the active user.
Historical user described herein refers to carrying out the user of certain network behavior in history, the specific network behavior
Can be consistent with active user above-mentioned, such as all it is the user for carrying out web search, can not also be consistent, such as some Web Community
In historical user, what which carried out is to deliver new post, and active user may be only browsing video etc..Statistics
Historical user source, it is related with the data source of server, be not limited herein.
When user information above-mentioned is numeralization array, this step is concretely：Calculate the user of the active user
The total amount of the difference of corresponding element between information array and the user information array of each historical user；Determine the total amount of the difference
Less than the historical user of predetermined value as the compartment history user in the similar users circle of the active user.
Certainly, calculate the total amount of above-mentioned difference specific algorithm can there are many, as calculate two arrays between it is each right
The absolute value for answering the difference of element, then sums, or calculate each corresponding element difference square after sum again, or meter
Calculate the difference of each corresponding element square after multiplied by sum after some coefficient, or by space projection, by two arrays
A bit in a certain space is projected to, then the distance etc. for calculating point-to-point transmission, passes through various existing mathematical algorithms in a word and obtain
Difference under the various forms of two arrays.
Such as, when by nearest neighbor algorithm calculating difference, the active user can be calculated according to formula 1 or formula 2 and described is gone through
The total amount of the difference of history user：
Again in the similar users circle from the user of the D (k) obtained in D (k), k=1~m≤υ as the active user
Historical user, wherein υ is predetermined threshold.Such as, it is predefined constant.
Wherein, K (n) is the value of n-th of attribute of k-th of historical user, and X (n) is n-th of category of the active user
Property value, H (n) is the average value of described k-th of historical user, n-th of attribute, and D (k) is the active user and the kth
The total amount of the difference of a historical user, n are the sum of the element in the array, and m is the sum of the historical user, and n and m are equal
For the integer more than or equal to 1.
It is that formula 1 is taken to calculate in a particular embodiment, formula 2 is still taken to calculate, needs to consider concrete condition.That is, for
The value of each attribute of user, some can directly be weighed with numerical value, such as emolument, age, some then cannot directly use number
Value is weighed, such as hobby class, then at this point, for X (n)-K's (n) as a result, can be to take X by hobby is identical
(n)-K (n) values are 0, and different X (n)-K (n) values that take are 1.Differing greatly for different dimensions attribute is so considered at this time, such as
Age and emolument, calculating difference can be very big, general to recommend to be calculated using formula 2, to balance the numerical value model between different attribute
The difference enclosed.Certainly, if in the case that the difference of different dimensions attribute is little, formula 1 can be used to simplify the calculation and calculate distance.
103, preference degree of the historical user in the similar users circle to each information to be pushed is obtained.Specifically calculating
When, it can be complete to the number of processes, processing density, processing of each information to be pushed according to the historical user in the similar users circle
The preference degree of one or more calculating each information to be pushed in degree." processing " described herein and specific pushed information
Related, e.g., pushed information is video ads, then processing is specially to play, and is not enumerated herein.
Illustrate that the specific calculating process of preference degree includes so that pushed information is Playable content as an example below：According to the phase
Historical user in being enclosed like user plays the number of each information to be pushed and time weighting calculates each information to be pushed
Density；The number and time weighting that each information to be pushed is clicked according to the historical user in the similar users circle, with
And the density of each information to be pushed calculates the click degree of each information to be pushed；According to going through in the similar users circle
History user completely plays the number of each information to be pushed and the density of time weighting and each information to be pushed calculates
The integrity degree of each information to be pushed；The preference degree of each information to be pushed is calculated according to the click degree and integrity degree.
Wherein, by taking the advertisement that pushed information is playable as an example, illustrate to calculate above-mentioned density, click degree, integrity degree and happiness
The specific algorithm spent well is as follows：
(1) density is calculated as follows：
Wherein, d is the broadcasting number of days of pushed information, and t is initial calculation time, qviFor the phase when time is i
Historical user in being enclosed like user plays the overall number of the message, and Intensity (l) is the history in the similar users circle
The density of the 1st information to be pushed of user couple.Wherein, τ is constant, is defined in this example, and τ can be 0.01 or 0.05.Certainly, τ
Also other values are may be defined as, can specifically be determined according to actual needs.
It should be noted that in this example, the density of advertisement is the advertising display number of history for a period of time according to 1/eτ*i
(i be number of days) apart from current time decayed after accumulated value.Remoter apart from current time, decaying is more.
(2) the click degree is calculated according to the following equation：
Click (l)=Click_num (l)/Intensity (l),
Wherein, clickiFor time i when hits, Click (l) is the historical user couple the in the similar users circle
The click degree of 1 information to be pushed.
It is similar, in this example, the hits not instead of ad click number of advertisement it is simple cumulative, when one section of history
Between advertisement be clicked number according to 1/eτ*i(i be number of days) apart from current time decayed after accumulated value.Distance is current
Time is remoter, and decaying is more.
(3) integrity degree is calculated according to the following equation：
Integrity (l)=Integrity_num (l)/Intensity (l),
Wherein, IntegrityiFor time i when complete broadcasting time, Integrity (l) be the similar users circle in
The 1st information to be pushed of historical user couple integrity degree.
In this example, advertisement completely plays number nor the simple of advertisement broadcasting time is added up, but history is for a period of time
The number that is completely played of advertisement according to 1/eτ*i(i be number of days) apart from current time decayed after accumulated value.Distance
Current time is remoter, and decaying is more.
(4) preference degree is calculated according to the following equation：
Adc (l)=Click (l)α*Integrity(l)β
Wherein, α, β are constant parameter, and 0≤α≤1,0≤β≤1, alpha+beta=1, Adc (l) are the similar users
The preference degree of the 1st information to be pushed of historical user couple in circle.
As can be seen from the above equation, the higher advertisement of Adc exponential quantities, in user's clicking rate, complete of history nearest a period of time
Whole broadcasting rate is higher, is calculated using single " click/exposure " or " complete broadcasting/exposure " compared to traditional,
The algorithm can synthetically weigh advertisement, while can embody the timeliness of historical data.As can be seen from the above description,
The calculating of preference degree is with reference to the impression of advertisement, hits, complete broadcasting number etc..Certainly, other kinds of push is believed
Breath, then can accordingly refer to corresponding content.
Meanwhile using e in above-mentioned algorithmτ*iThis variable weight mode is conducive to capture the variation feelings occurred recently
Condition.
104, information to be pushed is chosen according to the preference degree and is pushed to the active user.Such as, most by the preference degree
Big pushed information is pushed to the active user.It is of course also possible to a threshold constant be arranged, when preference degree is more than the constant
Information is then pushed to user.
As it can be seen that in embodiments of the present invention, after being calculated by the preference degree to the similar circle of user, higher happiness can be pushed
Good angle value it is corresponding some, certain class or multiple pushed informations.To realize personalized recommendation, the profit of Internet resources is improved
With rate.
Correspondingly, the embodiment of the present invention additionally provides an information push-delivery apparatus, as shown in Fig. 2, the device 1 includes：Letter
Cease acquisition module 10, the user information of user information and historical user for obtaining active user；User encloses acquisition module 12,
For user information, the user information of historical user according to the active user, the similar users of the active user are obtained
Circle, the similar users circle include one or more historical users related with the active user；Preference degree acquisition module 14,
Preference degree for obtaining the historical user in the similar users circle to each pushed information；Pushing module 16, for according to institute
It states preference degree selection information to be pushed and is pushed to the active user.
Wherein, the user information is numeralization array, and the array is referred to as user information array, the user information
Element in array corresponds to the numeralization value of one or more attributes of user, as shown in figure 3, the user encloses acquisition module
12 may include：Difference computing unit 120, the user of user information array and each historical user for calculating the active user
The total amount of the difference of corresponding element between information array；Similar circle determination unit 122, for determining that the total amount of the difference is less than
The historical user of predetermined value is as the historical user in the similar users circle of the active user.
Specifically, the difference computing unit 120, it can be used for calculating the active user according to following formula and described go through
The total amount of the difference of history user：
Wherein K (n) is the value of n-th of attribute of k-th of historical user, and X (n) is n-th of category of the active user
Property value, H (n) is the average value of described k-th of historical user, n-th of attribute, and D (k) is the active user and the kth
The total amount of the difference of a historical user, n are the sum of the element in the array, and m is the sum of the historical user, and n and m are equal
For the integer more than or equal to 1；
The similar circle determination unit 122 is used for, described in user's conduct of the D (k) obtained from D (k), k=1~m≤υ
Historical user in the similar users circle of active user, wherein υ is predetermined threshold.
Wherein, the preference degree acquisition module 14 is additionally operable to, according to the historical user in the similar users circle to push
The preference degree of the processing density of information, number of processes, one or more calculating pushed information in processing integrity degree.
Specifically, preference degree acquisition module 14 is additionally operable to according to the historical user in the similar users circle to respectively waiting pushing
The preference degree of the number of processes of information, processing density, one or more calculating each information to be pushed in processing integrity degree.
As shown in figure 4, preference degree acquisition module 14 may include：Density computational submodule 140, for according in the similar users circle
Historical user plays the number of each information to be pushed and time weighting calculates the density of each information to be pushed；Click degree
Computational submodule 142, for clicked according to the historical user in the similar users circle each information to be pushed number and
The density of time weighting and each information to be pushed calculates the click degree of each information to be pushed；Integrity degree calculates son
Module 144, number for completely playing each information to be pushed according to the historical user in the similar users circle and when
Between weight and the density of each information to be pushed calculate the integrity degree of each information to be pushed；Preference degree calculates submodule
Block 146, the preference degree for calculating each information to be pushed according to the click degree and integrity degree.
Such as, specifically, density computational submodule 140, can be calculated as follows the density：
Wherein, d is the broadcasting number of days of pushed information, and t is initial calculation time, qviFor the phase when time is i
Historical user in being enclosed like user plays the overall number of the message, and Intensity (l) is the history in the similar users circle
The density of the 1st information to be pushed of user couple；
Click degree computational submodule 142 can calculate the click degree according to the following equation：
Click (l)=Click_num (l)/Intensity (l),
Wherein, clickiFor time i when hits, Click (l) is the historical user couple the in the similar users circle
The click degree of 1 information to be pushed；
Integrity degree computational submodule 144 can calculate the integrity degree according to the following equation：
Integrity (l)=Integrity_num (l)/Intensity (l),
Wherein, IntegrityiFor time i when complete broadcasting time, Integrity (l) be the similar users circle in
The 1st information to be pushed of historical user couple integrity degree；
Preference degree computational submodule 146 can calculate the preference degree according to the following equation：
Adc (l)=Click (l)α*Integrity(l)β
Wherein, α, β are constant parameter, and 0≤α≤1,0≤β≤1, alpha+beta=1, Adc (l) are the similar users
The preference degree of the 1st information to be pushed of historical user couple in circle.
Specifically, the pushing module 16 is additionally operable to, the maximum pushed information of the preference degree is pushed to described current
User.
It is understood that the technical detail in above-mentioned apparatus embodiment is consistent in preceding method embodiment, herein
It does not do and repeats one by one.
In conclusion in embodiments of the present invention, when being pushed to a certain user into row information, being examined using nearest neighbor algorithm
The information for examining the relevant historical user of the user is formed in the related similar users circle of the user, according to similar users circle
Historical behavior investigates preference degree of these users to pushed information, is pushed into row information further according to preference degree.So, Xiang Yong
The information of family push just has the demand that larger probability meets user, improves the utilization rate of Internet resources.
One of ordinary skill in the art will appreciate that realizing all or part of flow in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a computer read/write memory medium
In, the program is when being executed, it may include such as the flow of the embodiment of above-mentioned each method.Wherein, the storage medium can be magnetic
Dish, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access
Memory, RAM) etc..
It is above disclosed to be only a preferred embodiment of the present invention, the power of the present invention cannot be limited with this certainly
Sharp range, therefore equivalent changes made in accordance with the claims of the present invention, are still within the scope of the present invention.
Claims (11)
1. a kind of information-pushing method, which is characterized in that the method includes：
Obtain the user information of the user information and historical user of active user；
According to the user information of the active user, the user information of historical user, the similar users of the active user are obtained
Circle, the similar users circle include one or more historical users related with the active user；
Play the number of each information to be pushed according to the historical user in the similar users circle and time weighting calculate it is described each
The density of information to be pushed, the information to be pushed are Playable content, and the time weighting is 1/eτ*i, i be apart from it is current when
Between number of days, τ is preset constant；
The number and time weighting of each information to be pushed, Yi Jisuo are clicked according to the historical user in the similar users circle
The density for stating each information to be pushed calculates the click degree of each information to be pushed；
The number and time weighting that each information to be pushed is completely played according to the historical user in the similar users circle, with
And the density of each information to be pushed calculates the integrity degree of each information to be pushed；
The preference degree of each information to be pushed is calculated according to the click degree and integrity degree, and is waited for according to preference degree selection
Pushed information is pushed to the active user.
2. the method as described in claim 1, which is characterized in that the user information is numeralization array, and claims the array
For user information array, the element in the user information array corresponds to the numeralization value of one or more attributes of user,
User information, the user information of historical user according to the active user obtains the similar users of the active user
Circle includes：
Calculate the difference of corresponding element between the user information array of the active user and the user information array of each historical user
Other total amount；
Determine that the total amount of the difference is less than the historical user of predetermined value as going through in the similar users circle of the active user
History user.
3. method as claimed in claim 2, which is characterized in that the user information array for calculating the active user and institute
The total amount for stating the difference of corresponding element between the user information array of each historical user in similar users circle includes：
The total amount of the difference is calculated according to following formula：
Wherein, K (n) is the value of n-th of attribute of k-th of historical user, and X (n) is n-th of attribute of the active user
Value, D (k) are the total amount of the active user and the difference of k-th of historical user, and n is the element in the array
Sum, m are the sum of the historical user, and n and m are the integer more than or equal to 1；
The total amount of the determination difference is less than the historical user of predetermined value as in the similar users circle of the active user
Historical user include：
The history in similar users circle from the user of the D (k) obtained in D (k), k=1~m≤υ as the active user is used
Family, wherein υ is predetermined threshold.
4. method as claimed in claim 2, which is characterized in that the user information array for calculating the active user and institute
The total amount for stating the difference of corresponding element between the user information array of each historical user in similar users circle includes：
The total amount of the difference of the active user and the historical user is calculated according to following formula：
Wherein, K (n) is the value of n-th of attribute of k-th of historical user, and X (n) is n-th of attribute of the active user
Value, H (n) are the average value of described k-th of historical user, n-th of attribute, and D (k) is that the active user goes through with described k-th
The total amount of the difference of history user, n are the sum of the element in the array, and m is the sum of the historical user, and n and m are big
In the integer equal to 1；
The total amount of the determination difference is less than the historical user of predetermined value as in the similar users circle of the active user
Historical user include：
The history in similar users circle from the user of the D (k) obtained in D (k), k=1~m≤υ as the active user is used
Family, wherein υ is predetermined threshold.
5. the method as described in claim 1, which is characterized in that the density is calculated as follows：
Wherein, d is the broadcasting number of days of pushed information, and t is initial calculation time, qviFor the similar use when time is i
Historical user in the circle of family plays the overall number of the message, and Intensity (l) is the historical user in the similar users circle
To the density of first of information to be pushed；
The click degree is calculated according to the following equation：
Click (l)=Click_num (l)/Intensity (l),
Wherein, clickiFor time i when hits, Click (l) is that the historical user couple in the similar users circle waits for for first
The click degree of pushed information；
The integrity degree is calculated according to the following equation：
Integrity (l)=Integrity_num (l)/Intensity (l),
Wherein, IntegrityiFor time i when complete broadcasting time, Integrity (l) be the similar users circle in going through
The integrity degree of first of information to be pushed of history user couple；
The preference degree is calculated according to the following equation：
Adc (l)=Click (l) α * Integrity (l)β
Wherein, α, β are constant parameter, and 0≤α≤1,0≤β≤1, alpha+beta=1, Adc (l) is in the similar users circles
The preference degree of first of information to be pushed of historical user couple.
6. the method as described in claim 1, which is characterized in that described to be pushed to according to preference degree selection information to be pushed
The active user includes：
The maximum information to be pushed of the preference degree is pushed to the active user.
7. a kind of information push-delivery apparatus, which is characterized in that described device includes：
Data obtaining module, the user information of user information and historical user for obtaining active user；
User encloses acquisition module, for user information, the user information of historical user according to the active user, described in acquisition
The similar users circle of active user, the similar users circle include that one or more history related with the active user are used
Family；
Preference degree acquisition module, the preference degree for obtaining the historical user in the similar users circle to each information to be pushed；
Pushing module is pushed to the active user for choosing information to be pushed according to the preference degree；
The wherein described preference degree acquisition module includes density computational submodule, click degree computational submodule, integrity degree calculating submodule
Block and preference degree computational submodule, wherein：
Density computational submodule, time for playing each information to be pushed according to the historical user in the similar users circle
Number and time weighting calculate the density of each information to be pushed, and the information to be pushed is Playable content, the time power
Weight is 1/eτ*i, i is the number of days apart from current time, and τ is preset constant；
Click degree computational submodule, for clicking each information to be pushed according to the historical user in the similar users circle
The density of number and time weighting and each information to be pushed calculates the click degree of each information to be pushed；
Integrity degree computational submodule, for according to the historical user in the similar users circle completely play it is described respectively wait for push letter
The density of the number and time weighting of breath and each information to be pushed calculates the integrity degree of each information to be pushed；
Preference degree computational submodule, the preference degree for calculating each information to be pushed according to the click degree and integrity degree.
8. device as claimed in claim 7, which is characterized in that the user information is numeralization array, and claims the array
For user information array, the element in the user information array corresponds to the numeralization value of one or more attributes of user,
The user encloses acquisition module：
Difference computing unit, for calculating the user information array of the active user and the user information array of each historical user
Between corresponding element difference total amount；
Similar circle determination unit, for determining that the total amount of the difference is less than the historical user of predetermined value as the active user
Similar users circle in historical user.
9. device as claimed in claim 8, which is characterized in that the difference computing unit is specifically used for according to following formula
Calculate the total amount of the difference of the active user and the historical user：
Or,
Wherein K (n) is the value of n-th of attribute of k-th of historical user, and X (n) is n-th of attribute of the active user
Value, H (n) are the average value of described k-th of historical user, n-th of attribute, and D (k) is that the active user goes through with described k-th
The total amount of the difference of history user, n are the sum of the element in the array, and m is the sum of the historical user, and n and m are big
In the integer equal to 1；
The similar circle determination unit is specifically used for, and works as described in user's conduct of the D (k) obtained from D (k), k=1~m≤υ
Historical user in the similar users circle of preceding user, wherein υ is predetermined threshold.
10. device as claimed in claim 7, which is characterized in that the density computational submodule, specifically for being calculated as follows
The density：
Wherein, d is the broadcasting number of days of pushed information, and t is initial calculation time, qviFor the similar use when time is i
Historical user in the circle of family plays the overall number of the message, and Intensity (l) is the historical user in the similar users circle
To the density of first of information to be pushed；
The click degree computational submodule, specifically for calculating the click degree according to the following equation：
Click (l)=Click_num (l)/Intensity (l),
Wherein, clickiFor time i when hits, Click (l) is that the historical user couple in the similar users circle waits for for first
The click degree of pushed information；
The integrity degree computational submodule, specifically for calculating the integrity degree according to the following equation：
Integrity (l)=Integrity_num (l)/Intensity (l),
Wherein, IntegrityiFor time i when complete broadcasting time, Integrity (l) be the similar users circle in going through
The integrity degree of first of information to be pushed of history user couple；
The preference degree computational submodule, specifically for calculating the preference degree according to the following equation：
Adc (l)=Click (l)α*Integrity(l)β
Wherein, α, β are constant parameter, and 0≤α≤1,0≤β≤1, alpha+beta=1, Adc (l) is in the similar users circles
The preference degree of first of information to be pushed of historical user couple.
11. device as claimed in claim 7, which is characterized in that the pushing module is additionally operable to, and the preference degree is maximum
Information to be pushed is pushed to the active user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210370622.4A CN103716338B (en) | 2012-09-28 | 2012-09-28 | A kind of information-pushing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210370622.4A CN103716338B (en) | 2012-09-28 | 2012-09-28 | A kind of information-pushing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103716338A CN103716338A (en) | 2014-04-09 |
| CN103716338B true CN103716338B (en) | 2018-08-17 |
Family
ID=50408918
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210370622.4A Active CN103716338B (en) | 2012-09-28 | 2012-09-28 | A kind of information-pushing method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103716338B (en) |
Families Citing this family (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106575418B (en) * | 2014-05-16 | 2021-06-08 | 微软技术许可有限责任公司 | Suggested keywords |
| CN104636470B (en) * | 2015-02-11 | 2018-04-24 | 广州华多网络科技有限公司 | A kind of method and device of recommendation business information |
| CN106294457B (en) * | 2015-05-29 | 2021-05-14 | 腾讯科技（深圳）有限公司 | Network information pushing method and device |
| CN104935964A (en) * | 2015-06-02 | 2015-09-23 | 四川九天揽月文化传媒有限公司 | Program grouping screening and push method for intelligent television |
| CN104918070A (en) * | 2015-06-02 | 2015-09-16 | 四川九天揽月文化传媒有限公司 | Smart television-based video program push system and push method |
| CN104967690B (en) * | 2015-06-30 | 2019-04-12 | 北京奇艺世纪科技有限公司 | A kind of information-pushing method and device |
| CN105159910A (en) * | 2015-07-03 | 2015-12-16 | 安一恒通（北京）科技有限公司 | Information recommendation method and device |
| CN110460645A (en) * | 2015-08-04 | 2019-11-15 | 阿里巴巴集团控股有限公司 | A kind of information-pushing method and device |
| CN105354724A (en) * | 2015-11-04 | 2016-02-24 | 北京金山安全软件有限公司 | Advertisement information pushing method and server |
| CN106919580B (en) * | 2015-12-25 | 2021-04-13 | 腾讯科技（深圳）有限公司 | Information pushing method and device |
| CN106407361A (en) * | 2016-09-07 | 2017-02-15 | 北京百度网讯科技有限公司 | Method and device for pushing information based on artificial intelligence |
| CN107066582B (en) * | 2017-04-14 | 2020-06-26 | 聚好看科技股份有限公司 | Method and device for realizing virtual resource recommendation |
| CN107846463A (en) * | 2017-11-08 | 2018-03-27 | 北京摩拜科技有限公司 | Information-pushing method, server, client and Vehicular system |
| CN109816414A (en) * | 2017-11-22 | 2019-05-28 | 北京国双科技有限公司 | Advertisement effectiveness data prediction technique and device |
| CN108629045A (en) * | 2018-05-14 | 2018-10-09 | 平安科技（深圳）有限公司 | A kind of information push method, storage medium and server |
| CN108875043B (en) * | 2018-06-27 | 2022-02-25 | 腾讯科技（北京）有限公司 | User data processing method and device, computer equipment and storage medium |
| CN109657140A (en) * | 2018-12-13 | 2019-04-19 | 平安普惠企业管理有限公司 | Information-pushing method, device, computer equipment and storage medium |
| CN110246002A (en) * | 2019-04-25 | 2019-09-17 | 深圳壹账通智能科技有限公司 | Method for pushing, device and the computer equipment of financing recommendation information |
| CN113032662B (en) * | 2021-03-31 | 2021-11-26 | 艾普深瞳(北京)智能科技有限公司 | Block chain big data recommendation method and system based on artificial intelligence and cloud platform |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1975866A1 (en) * | 2007-03-31 | 2008-10-01 | Sony Deutschland Gmbh | Method and system for recommending content items |
| KR101552147B1 (en) * | 2008-04-24 | 2015-09-11 | 삼성전자주식회사 | Method for recommending broadcasting contents and apparatus thereof |
| CN101685458B (en) * | 2008-09-27 | 2012-09-19 | 华为技术有限公司 | Recommendation method and system based on collaborative filtering |
| CN101937547A (en) * | 2010-09-15 | 2011-01-05 | 宇龙计算机通信科技(深圳)有限公司 | Software and/or software information pushing method, system, acquisition device, software shop service system and mobile terminal |
| CN102592223A (en) * | 2011-01-18 | 2012-07-18 | 卓望数码技术（深圳）有限公司 | Commodity recommending method and commodity recommending system |
| CN102332006B (en) * | 2011-08-03 | 2016-08-03 | 百度在线网络技术（北京）有限公司 | A kind of information push control method and device |
-
2012
- 2012-09-28 CN CN201210370622.4A patent/CN103716338B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN103716338A (en) | 2014-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103716338B (en) | A kind of information-pushing method and device | |
| US11763200B2 (en) | Method and system for creating a predictive model for targeting web-page to a surfer | |
| US9800910B2 (en) | Recommending media items based on take rate signals | |
| KR102159274B1 (en) | Methods and systems for creating a data-driven attribution model for assigning attribution credit to a plurality of events | |
| US8893012B1 (en) | Visual indicator based on relative rating of content item | |
| KR101650117B1 (en) | Conversion path performance measures and reports | |
| KR101923065B1 (en) | User-initiated boosting of social networking objects | |
| US10032176B2 (en) | Real time statistics extraction from arbitrary advertising audiences | |
| US8744989B1 (en) | Ranking and vote scheduling using statistical confidence intervals | |
| CN108140041B (en) | For the viewing time cluster of video search | |
| Cleger-Tamayo et al. | Top-N news recommendations in digital newspapers | |
| US20100023475A1 (en) | Method and system for creating a predictive model for targeting webpage to a surfer | |
| CN106294830A (en) | The recommendation method and device of multimedia resource | |
| CN109889891B (en) | Method, device and storage medium for acquiring target media file | |
| CN107832437A (en) | Audio/video method for pushing, device, equipment and storage medium | |
| CN113535991B (en) | Multimedia resource recommendation method and device, electronic equipment and storage medium | |
| CN104169910A (en) | Method to evaluate the geographic popularity of geographically located user-generated content items | |
| US20100030717A1 (en) | Framework to evaluate content display policies | |
| CN111143738A (en) | Resource display method and device, electronic equipment and storage medium | |
| CN104967690B (en) | A kind of information-pushing method and device | |
| CN103700004A (en) | Method and device for pushing microblog advertising service information | |
| CN109635192A (en) | Magnanimity information temperature seniority among brothers and sisters update method and platform towards micro services | |
| Medo et al. | Identification and impact of discoverers in online social systems | |
| JP5386660B1 (en) | Information processing apparatus, information processing method, and information processing program | |
| CN113015010A (en) | Push parameter determination method, device, equipment and computer readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |