CN106227834A - The recommendation method and device of multimedia resource - Google Patents

The recommendation method and device of multimedia resource Download PDF

Info

Publication number
CN106227834A
CN106227834A CN201610594958.7A CN201610594958A CN106227834A CN 106227834 A CN106227834 A CN 106227834A CN 201610594958 A CN201610594958 A CN 201610594958A CN 106227834 A CN106227834 A CN 106227834A
Authority
CN
China
Prior art keywords
multimedia resource
recommended
kind multimedia
resource
click volume
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201610594958.7A
Other languages
Chinese (zh)
Other versions
CN106227834B (en
Inventor
滕飞
赵磊
单明辉
尹玉宗
姚键
潘柏宇
王冀
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Alibaba China Co Ltd
Youku Network Technology Beijing Co Ltd
Original Assignee
1Verge Internet Technology Beijing Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 1Verge Internet Technology Beijing Co Ltd filed Critical 1Verge Internet Technology Beijing Co Ltd
Priority to CN201610594958.7A priority Critical patent/CN106227834B/en
Publication of CN106227834A publication Critical patent/CN106227834A/en
Application granted granted Critical
Publication of CN106227834B publication Critical patent/CN106227834B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
    • G06F16/43—Querying
    • G06F16/435—Filtering based on additional data, e.g. user or group profiles

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The present invention relates to the recommendation method and device of multimedia resource.The method includes: multimedia resource to be recommended is divided into first kind multimedia resource and Equations of The Second Kind multimedia resource；History click data according to first kind multimedia resource determines the click volume discreet value of first kind multimedia resource；Search the first kind multimedia resource mated with Equations of The Second Kind multimedia resource, calculate the matching degree of Equations of The Second Kind multimedia resource and the first kind multimedia resource mated, and calculate the click volume discreet value of Equations of The Second Kind multimedia resource according to matching degree with the click volume discreet value of the first kind multimedia resource mated；Multimedia resource to be recommended is recommended by the click volume discreet value according to multimedia resource to be recommended.It is recommended that the recommendation method and device of the multimedia resource according to the present invention can make the multimedia resource newly uploaded also have an opportunity, and which thereby enhances the recommendation accuracy of multimedia resource, thus improves the recommendation effect of multimedia resource.

Description

The recommendation method and device of multimedia resource
Technical field
The present invention relates to areas of information technology, particularly relate to the recommendation method and device of a kind of multimedia resource.
Background technology
Existing video recommendations technology mainly includes video recommendations technology based on click volume, video based on uplink time Recommended technology and video recommendations technology based on text message.Wherein, text message refers to title and/or the uploader of video The description information added for video.In existing video recommendations technology based on click volume, that some are newly uploaded, click volume is relatively Few video is difficult to recommended.Existing video recommendations technology based on uplink time does not accounts for the temperature of video, it is recommended that effect The most poor.In existing video recommendations technology based on text message, due to text message and the video content of some videos Dependency relatively low, the accuracy causing video recommendations is relatively low.
Summary of the invention
Technical problem
In view of this, the technical problem to be solved in the present invention is, the recommendation effect of the recommended technology of existing multimedia resource The most poor problem.
Solution
In order to solve above-mentioned technical problem, according to one embodiment of the invention, it is provided that the recommendation of a kind of multimedia resource Method, including:
Multimedia resource to be recommended is divided into first kind multimedia resource and Equations of The Second Kind multimedia resource；
For each described first kind multimedia resource, the history click data according to described first kind multimedia resource is true The click volume discreet value of fixed described first kind multimedia resource；
For each described Equations of The Second Kind multimedia resource, search many with the first kind that described Equations of The Second Kind multimedia resource mates Media resource, calculates described Equations of The Second Kind multimedia resource and the matching degree of the first kind multimedia resource mated, and according to institute The click volume discreet value of the first kind multimedia resource stating matching degree and mated calculates the point of described Equations of The Second Kind multimedia resource The amount of hitting discreet value；
Described multimedia resource to be recommended is carried out by the click volume discreet value according to described multimedia resource to be recommended Recommend.
For said method, in a kind of possible implementation, multimedia resource to be recommended is divided into the first kind Multimedia resource and Equations of The Second Kind multimedia resource, including:
Obtain the history click data of each multimedia resource to be recommended respectively；
For each described multimedia resource to be recommended, according to the history hits of described multimedia resource to be recommended It is judged that whether the current click volume of described multimedia resource to be recommended is in the decline phase, the most then by described to be recommended Multimedia resource is defined as first kind multimedia resource, otherwise described multimedia resource to be recommended is defined as the many matchmakers of Equations of The Second Kind Body resource.
For said method, in a kind of possible implementation, multimedia resource to be recommended is divided into the first kind Multimedia resource and Equations of The Second Kind multimedia resource, including:
Obtain the uplink time of each multimedia resource to be recommended respectively；
For each described multimedia resource to be recommended, it is judged that the uplink time of described multimedia resource to be recommended away from Whether the time span from present system time is more than the first preset value, the most then by true for described multimedia resource to be recommended It is set to first kind multimedia resource, otherwise described multimedia resource to be recommended is defined as Equations of The Second Kind multimedia resource.
For said method, in a kind of possible implementation, for each described first kind multimedia resource, according to The history click data of described first kind multimedia resource determines the click volume discreet value of described first kind multimedia resource, bag Include:
For each described first kind multimedia resource, obtain the history click data of described first kind multimedia resource；
History click data according to described first kind multimedia resource trains the time of described first kind multimedia resource Attenuation quotient；
History click data and described time attenuation quotient according to described first kind multimedia resource calculate described first The click volume discreet value of class multimedia resource.
For said method, in a kind of possible implementation, according to the history point of described first kind multimedia resource Hit data and train the time attenuation quotient of described first kind multimedia resource, including:
According to the history click data of first kind multimedia resource i, employing formula 1 trains described first kind multimedia resource i Time attenuation quotient θ, wherein, the history of described first kind multimedia resource i is clicked on packet and is included described first kind multimedia Resource i click volume before appointed day；
F (t)=Vi0×eθtFormula 1；
Wherein, Vi0Represent the described first kind multimedia resource i click volume at described appointed day；T=0 represents described finger Fix the date, f (0)=Vi0；As t ＞ 0, t represents the t days before described appointed day, and f (t) represents the many matchmakers of the described first kind The click volume of before described appointed day the t days of body resource i；As t ＜ 0, t represents the-t after described appointed day My god, f (t) represents the click volume discreet value in the described first kind multimedia resource i-t sky after described appointed day；
History click data and described time attenuation quotient according to described first kind multimedia resource calculate described first The click volume discreet value of class multimedia resource, including:
The formula 1 after training is used to calculate the point in the described first kind multimedia resource i-t sky after described appointed day The amount of hitting discreet value.
For said method, in a kind of possible implementation, according to the history point of described first kind multimedia resource Hit data and train the time attenuation quotient of described first kind multimedia resource, including:
According to the history click data of first kind multimedia resource i, employing formula 2 trains described first kind multimedia resource i Time attenuation quotient θ, wherein, the history of described first kind multimedia resource i is clicked on packet and is included described first kind multimedia Resource i click volume before appointed day；
G (t)=lg (Vi0×eθt) formula 2；
Wherein, Vi0Represent the described first kind multimedia resource i click volume at described appointed day；T=0 represents described finger Fix the date, g (0)=lgVi0；As t ＞ 0, t represents the t days before described appointed day, and g (t) represents that the described first kind is many The logarithm value of the click volume of before described appointed day the t days of media resource i；As t ＜ 0, t represents at described appointed day After-t sky, g (t) represents that the click volume in the described first kind multimedia resource i-t sky after described appointed day is estimated The logarithm value of value；
History click data and described time attenuation quotient according to described first kind multimedia resource calculate described first The click volume discreet value of class multimedia resource, including:
The formula 2 after training is used to calculate the point in the described first kind multimedia resource i-t sky after described appointed day The logarithm value of the amount of hitting discreet value.
For said method, in a kind of possible implementation, for each described Equations of The Second Kind multimedia resource, search The first kind multimedia resource mated with described Equations of The Second Kind multimedia resource, calculates described Equations of The Second Kind multimedia resource and is mated The matching degree of first kind multimedia resource, including:
For each described Equations of The Second Kind multimedia resource, according to the appointment information searching of described Equations of The Second Kind multimedia resource with The first kind multimedia resource of described Equations of The Second Kind multimedia resource coupling；Described appointment information include following at least one: upload Time, time span and uploader information；
The appointment letter of the appointment information according to described Equations of The Second Kind multimedia resource and the first kind multimedia resource that mated Breath calculates the matching degree of described Equations of The Second Kind multimedia resource and the first kind multimedia resource mated.
For said method, in a kind of possible implementation, many with the first kind mated according to described matching degree The click volume discreet value of media resource calculates the click volume discreet value of described Equations of The Second Kind multimedia resource, particularly as follows:
Using described matching degree with the product of the click volume discreet value of the first kind multimedia resource mated as described The click volume discreet value of two class multimedia resources.
For said method, in a kind of possible implementation, according to the click of described multimedia resource to be recommended Described multimedia resource to be recommended is recommended by amount discreet value, including:
Click volume discreet value according to described multimedia resource to be recommended order from high to low is waited to push away described in all The multimedia resource recommended is ranked up；
The preceding M that sorts is taken out individual described to be recommended many from all described multimedia resource to be recommended after sequence Media resource is recommended, and wherein, the size of M pre-sets.
In order to solve above-mentioned technical problem, according to another embodiment of the present invention, it is provided that pushing away of a kind of multimedia resource Recommend device, including:
Divide module, for multimedia resource to be recommended being divided into first kind multimedia resource and Equations of The Second Kind multimedia Resource；
First click volume discreet value determines module, for for each described first kind multimedia resource, according to described the The history click data of one class multimedia resource determines the click volume discreet value of described first kind multimedia resource；
Second click volume discreet value determines module, for for each described Equations of The Second Kind multimedia resource, searches with described The first kind multimedia resource of Equations of The Second Kind multimedia resource coupling, calculates described Equations of The Second Kind multimedia resource and first mated The matching degree of class multimedia resource, and according to the click volume discreet value of described matching degree with the first kind multimedia resource mated Calculate the click volume discreet value of described Equations of The Second Kind multimedia resource；
Recommending module, is used for the click volume discreet value according to described multimedia resource to be recommended to described to be recommended many Media resource is recommended.
For said apparatus, in a kind of possible implementation, described division module includes:
History clicks on data acquisition submodule, for obtaining the history hits of each multimedia resource to be recommended respectively According to；
First divides submodule, for for each described multimedia resource to be recommended, according to described to be recommended many The history click data of media resource judges whether the current click volume of described multimedia resource to be recommended is in the decline phase, if It is then described multimedia resource to be recommended to be defined as first kind multimedia resource, otherwise by described multimedia to be recommended Resource is defined as Equations of The Second Kind multimedia resource.
For said apparatus, in a kind of possible implementation, described division module includes:
Uplink time obtains submodule, for obtaining the uplink time of each multimedia resource to be recommended respectively；
Second divides submodule, for for each described multimedia resource to be recommended, it is judged that described to be recommended many Whether the time span of the uplink time distance present system time of media resource is more than the first preset value, the most then by described Multimedia resource to be recommended is defined as first kind multimedia resource, and described multimedia resource to be recommended is otherwise defined as Two class multimedia resources.
For said apparatus, in a kind of possible implementation, described first click volume discreet value determines that module includes:
History clicks on data acquisition submodule, for for each described first kind multimedia resource, obtains described first The history click data of class multimedia resource；
Time attenuation quotient training submodule, trains for the history click data according to described first kind multimedia resource The time attenuation quotient of described first kind multimedia resource；
First click volume discreet value calculating sub module, for the history click data according to described first kind multimedia resource With the click volume discreet value that described time attenuation quotient calculates described first kind multimedia resource.
For said apparatus, in a kind of possible implementation, described time attenuation quotient training submodule is specifically used In:
According to the history click data of first kind multimedia resource i, employing formula 1 trains described first kind multimedia resource i Time attenuation quotient θ, wherein, the history of described first kind multimedia resource i is clicked on packet and is included described first kind multimedia Resource i click volume before appointed day；
F (t)=Vi0×eθtFormula 1；
Wherein, Vi0Represent the described first kind multimedia resource i click volume at described appointed day；T=0 represents described finger Fix the date, f (0)=Vi0；As t ＞ 0, t represents the t days before described appointed day, and f (t) represents the many matchmakers of the described first kind The click volume of before described appointed day the t days of body resource i；As t ＜ 0, t represents the-t after described appointed day My god, f (t) represents the click volume discreet value in the described first kind multimedia resource i-t sky after described appointed day；
Described first click volume discreet value calculating sub module specifically for:
The formula 1 after training is used to calculate the point in the described first kind multimedia resource i-t sky after described appointed day The amount of hitting discreet value.
For said apparatus, in a kind of possible implementation, described time attenuation quotient training submodule is specifically used In:
According to the history click data of first kind multimedia resource i, employing formula 2 trains described first kind multimedia resource i Time attenuation quotient θ, wherein, the history of described first kind multimedia resource i is clicked on packet and is included described first kind multimedia Resource i click volume before appointed day；
G (t)=lg (Vi0×eθt) formula 2；
Wherein, Vi0Represent the described first kind multimedia resource i click volume at described appointed day；T=0 represents described finger Fix the date, g (0)=lgVi0；As t ＞ 0, t represents the t days before described appointed day, and g (t) represents that the described first kind is many The logarithm value of the click volume of before described appointed day the t days of media resource i；As t ＜ 0, t represents at described appointed day After-t sky, g (t) represents that the click volume in the described first kind multimedia resource i-t sky after described appointed day is estimated The logarithm value of value；
Described first click volume discreet value calculating sub module specifically for:
The formula 2 after training is used to calculate the point in the described first kind multimedia resource i-t sky after described appointed day The logarithm value of the amount of hitting discreet value.
For said apparatus, in a kind of possible implementation, described second click volume discreet value determines that module includes:
Matched sub-block, for for each described Equations of The Second Kind multimedia resource, according to described Equations of The Second Kind multimedia resource Specify the first kind multimedia resource that mates with described Equations of The Second Kind multimedia resource of information searching；Described appointment information include with Descend at least one: uplink time, time span and uploader information；
Matching degree calculating sub module, for the appointment information according to described Equations of The Second Kind multimedia resource and first mated The appointment information of class multimedia resource calculate described Equations of The Second Kind multimedia resource and the first kind multimedia resource mated Degree of joining.
For said apparatus, in a kind of possible implementation, described second click volume discreet value determines that module includes:
Second click volume discreet value calculating sub module, for by described matching degree and the first kind multimedia resource mated The product of click volume discreet value as the click volume discreet value of described Equations of The Second Kind multimedia resource.
For said apparatus, in a kind of possible implementation, described recommending module includes:
Sorting sub-module, is used for the order from high to low of the click volume discreet value according to described multimedia resource to be recommended All described multimedia resources to be recommended are ranked up；
Recommend submodule, for taking out the preceding M that sorts from all described multimedia resource to be recommended after sequence Individual described multimedia resource to be recommended is recommended, and wherein, the size of M pre-sets.
Beneficial effect
By multimedia resource to be recommended being divided into first kind multimedia resource and Equations of The Second Kind multimedia resource, according to The history click data of first kind multimedia resource determines the click volume discreet value of first kind multimedia resource, according to Equations of The Second Kind The click volume discreet value of first kind multimedia resource and the matching degree between the two of multimedia resource coupling calculate Equations of The Second Kind The click volume discreet value of multimedia resource, further according to the click volume discreet value of multimedia resource to be recommended to many matchmakers to be recommended Body resource is recommended, and the recommendation method and device of multimedia resource according to embodiments of the present invention can make the many matchmakers newly uploaded Body resource is also had an opportunity recommended, which thereby enhances the recommendation accuracy of multimedia resource, thus improves multimedia resource Recommendation effect.
According to below with reference to the accompanying drawings detailed description of illustrative embodiments, the further feature of the present invention and aspect being become Clear.
Accompanying drawing explanation
The accompanying drawing of the part comprising in the description and constituting description together illustrates the present invention's with description Exemplary embodiment, feature and aspect, and for explaining the principle of the present invention.
Fig. 1 illustrates the flowchart of the recommendation method of multimedia resource according to an embodiment of the invention；
Fig. 2 illustrates an exemplary tool of the recommendation method step S101 of multimedia resource according to an embodiment of the invention Body flowchart；
Fig. 3 illustrates the another exemplary of the recommendation method step S101 of multimedia resource according to an embodiment of the invention Implement flow chart；
Fig. 4 illustrates an exemplary tool of the recommendation method step S102 of multimedia resource according to an embodiment of the invention Body flowchart；
Fig. 5 illustrates in the recommendation method step S103 of multimedia resource according to an embodiment of the invention for each second Class multimedia resource, searches the first kind multimedia resource mated with this Equations of The Second Kind multimedia resource, calculates the many matchmakers of this Equations of The Second Kind What body resource was exemplary with the one of the matching degree of the first kind multimedia resource mated implements flow chart；
Fig. 6 illustrates an exemplary tool of the recommendation method step S104 of multimedia resource according to an embodiment of the invention Body flowchart；
Fig. 7 illustrates the structured flowchart of the recommendation apparatus of multimedia resource according to another embodiment of the present invention；
Fig. 8 illustrates an exemplary structural frames of the recommendation apparatus of multimedia resource according to another embodiment of the present invention Figure；
Fig. 9 shows the structured flowchart of the recommendation apparatus of a kind of multimedia resource of an alternative embodiment of the invention.
Detailed description of the invention
Various exemplary embodiments, feature and the aspect of the present invention is described in detail below with reference to accompanying drawing.In accompanying drawing identical Reference represent the same or analogous element of function.Although the various aspects of embodiment shown in the drawings, but remove Non-specifically is pointed out, it is not necessary to accompanying drawing drawn to scale.
The most special word " exemplary " means " as example, embodiment or illustrative ".Here as " exemplary " Illustrated any embodiment should not necessarily be construed as preferred or advantageous over other embodiments.
It addition, in order to better illustrate the present invention, detailed description of the invention below gives numerous details. It will be appreciated by those skilled in the art that do not have some detail, the present invention equally implements.In some instances, for Method well known to those skilled in the art, means, element and circuit are not described in detail, in order to highlight the purport of the present invention.
Embodiment 1
Fig. 1 illustrates the flowchart of the recommendation method of multimedia resource according to an embodiment of the invention.Such as Fig. 1 institute Showing, the method specifically includes that
In step S101, multimedia resource to be recommended is divided into first kind multimedia resource and Equations of The Second Kind multimedia Resource.
In embodiments of the present invention, multimedia can be the comprehensive of media, such as, include text, sound and image. Such as, the multimedia resource of the embodiment of the present invention can be video, in this no limit.
As an example of the embodiment of the present invention, before step S101, the method also includes: from multimedia resources database In filter out multimedia resource to be recommended.Such as, multimedia resource to be recommended can be not comprise in multimedia resources database The multimedia resource of sensitive information.Multimedia resource to be recommended can also be according to other screening requirements from multimedia resources database Screening, in this no limit.
In embodiments of the present invention, all multimedia resources to be recommended can be divided into first kind multimedia resource and Equations of The Second Kind multimedia resource, and different method meters is respectively adopted for first kind multimedia resource and Equations of The Second Kind multimedia resource Calculate its click volume discreet value.
In step s 102, for each first kind multimedia resource, according to the history point of this first kind multimedia resource Hit data and determine the click volume discreet value of this first kind multimedia resource.
In step s 103, for each Equations of The Second Kind multimedia resource, lookup is mated with this Equations of The Second Kind multimedia resource First kind multimedia resource, calculates this Equations of The Second Kind multimedia resource and the matching degree of the first kind multimedia resource mated, and Click volume discreet value according to matching degree with the first kind multimedia resource mated calculates the point of this Equations of The Second Kind multimedia resource The amount of hitting discreet value.
In embodiments of the present invention, for each first kind multimedia resource, respectively according to this first kind multimedia resource History click data determine the click volume discreet value of this first kind multimedia resource.For each Equations of The Second Kind multimedia resource, Search the first kind multimedia resource mated with this Equations of The Second Kind multimedia resource respectively.Such as, for Equations of The Second Kind multimedia resource v1If finding and this Equations of The Second Kind multimedia resource v1The first kind multimedia resource v of coupling2, then Equations of The Second Kind multimedia money is calculated Source v1With first kind multimedia resource v2Matching degree, further according to matching degree and first kind multimedia resource v2Click volume pre- Valuation calculates Equations of The Second Kind multimedia resource v1Click volume discreet value.
In step S104, according to the click volume discreet value of multimedia resource to be recommended to multimedia resource to be recommended Recommend.
Wherein, multimedia resource to be recommended includes first kind multimedia resource and Equations of The Second Kind multimedia resource.
In a kind of possible implementation, can be in conjunction with the click volume discreet value of multimedia resource to be recommended and application Multimedia resource to be recommended is recommended by scene.Wherein, application scenarios can include application platform information and channel information. Such as, application platform can include page end and APP (Application, application) end.According to the difference of application platform, it is used for Carry out the possible difference of number of the multimedia resource to be recommended recommended.Channel can include news channel, XATV-6 and body Ssd channel etc..According to the difference of channel, it is likely to difference for the multimedia resource to be recommended carrying out recommending.
Fig. 2 illustrates an exemplary tool of the recommendation method step S101 of multimedia resource according to an embodiment of the invention Body flowchart.As in figure 2 it is shown, multimedia resource to be recommended to be divided into first kind multimedia resource and the many matchmakers of Equations of The Second Kind Body resource, including:
In step s 201, the history click data of each multimedia resource to be recommended is obtained respectively.
In step S202, for each multimedia resource to be recommended, according to going through of this multimedia resource to be recommended History click data judges whether the current click volume of this multimedia resource to be recommended is in the decline phase, the most then wait to push away by this The multimedia resource recommended is defined as first kind multimedia resource, otherwise the multimedia resource that this is to be recommended is defined as Equations of The Second Kind many Media resource.
Wherein, the current click volume of multimedia resource to be recommended is in the decline phase and can refer to this multimedia to be recommended Resource its click volume that postpones over time presents downward trend.
Fig. 3 illustrates the another exemplary of the recommendation method step S101 of multimedia resource according to an embodiment of the invention Implement flow chart.As it is shown on figure 3, multimedia resource to be recommended is divided into first kind multimedia resource and Equations of The Second Kind is many Media resource, including:
In step S301, obtain the uplink time of each multimedia resource to be recommended respectively.
In step s 302, for each multimedia resource to be recommended, it is judged that this multimedia resource to be recommended upper Whether the time span passing time gap present system time is more than the first preset value, the most then by this multimedia to be recommended Resource is defined as first kind multimedia resource, otherwise the multimedia resource that this is to be recommended is defined as Equations of The Second Kind multimedia resource.
Such as, the first preset value can be 48 hours, in this no limit.
Fig. 4 illustrates an exemplary tool of the recommendation method step S102 of multimedia resource according to an embodiment of the invention Body flowchart.As shown in Figure 4, for each first kind multimedia resource, according to the history of this first kind multimedia resource Click data determines the click volume discreet value of this first kind multimedia resource, including:
In step S401, for each first kind multimedia resource, obtain the history point of this first kind multimedia resource Hit data.
In step S402, this first kind multimedia is trained to provide according to the history click data of this first kind multimedia resource The time attenuation quotient in source.
In step S403, history click data and time attenuation quotient according to this first kind multimedia resource calculate and are somebody's turn to do The click volume discreet value of first kind multimedia resource.
In a kind of possible implementation, according to the history click data of this first kind multimedia resource train this first The time attenuation quotient of class multimedia resource, including: according to the history click data of first kind multimedia resource i, employing formula 1 is instructed Practicing the time attenuation quotient θ of first kind multimedia resource i, wherein, the history of first kind multimedia resource i is clicked on packet and is included the One class multimedia resource i click volume before appointed day；
F (t)=Vi0×eθtFormula 1；
Wherein, Vi0Represent the first kind multimedia resource i click volume at appointed day；T=0 represents appointed day, f (0) =Vi0；As t ＞ 0, t represents the t days before appointed day, and f (t) represents that first kind multimedia resource i is before appointed day The click volume of the t days；As t ＜ 0, t represents the-t sky after appointed day, and f (t) represents first kind multimedia resource i The click volume discreet value in-t sky after appointed day.
Such as, the history of first kind multimedia resource i is clicked on packet and is included first kind multimedia resource i at appointed day Click volume and first kind multimedia resource i click volume of 15 days before appointed day.T=1 represents the previous of appointed day My god, f (1) represents the first kind multimedia resource i click volume in the previous day of appointed day, and t=2 represents front the two of appointed day My god, f (2) represents the first kind multimedia resource i click volume a few days ago at appointed day, and by that analogy, t=15 represents appointment First 15 days of date, f (15) represents that first kind multimedia resource i was the click volume of first 15 days of appointed day.Instruct in employing formula 1 When practicing the time attenuation quotient θ of first kind multimedia resource i, for t ＞ 0, the f (t) in formula 1, Vi0It is known with t, according to f (t)、Vi0The time attenuation quotient θ obtaining first kind multimedia resource i can be trained with t.According to first kind multimedia resource i's Time attenuation quotient θ can also be corrected by history click data.
History click data and time attenuation quotient according to this first kind multimedia resource calculate this first kind multimedia The click volume discreet value of resource, including: use the formula 1 after training calculate first kind multimedia resource i the after appointed day- The click volume discreet value of t days.
T=-1 represents the previous day of appointed day, and f (-1) represents previous at appointed day of first kind multimedia resource i It click volume.After training obtains the time attenuation quotient θ of first kind multimedia resource i, for t ＜ 0, the V in formula 1i0、θ It is known with t, it is thus possible to calculate the click volume discreet value of certain day after appointed day.Such as, the formula 1 after training is used Calculate the click volume discreet value in the first kind multimedia resource i-t sky after appointed day, Ke Yiwei: use the formula after training 1 calculates f (-1), and f (-1) is defined as the click volume discreet value of first kind multimedia resource i.
In a kind of possible implementation, according to the history click data of this first kind multimedia resource train this first The time attenuation quotient of class multimedia resource, including: according to the history click data of first kind multimedia resource i, employing formula 2 is instructed Practicing the time attenuation quotient θ of first kind multimedia resource i, wherein, the history of first kind multimedia resource i is clicked on packet and is included the One class multimedia resource i click volume before appointed day；
G (t)=lg (Vi0×eθt) formula 2；
Wherein, Vi0Represent the first kind multimedia resource i click volume at appointed day；T=0 represents appointed day, g (0) =lgVi0；As t ＞ 0, t represents the t days before appointed day, and g (t) represents that first kind multimedia resource i is at appointed day The logarithm value of the click volume of front the t days；As t ＜ 0, t represents the-t sky after appointed day, and g (t) represents that the first kind is many The logarithm value of the click volume discreet value in the media resource i-t sky after appointed day.
History click data and time attenuation quotient according to this first kind multimedia resource calculate this first kind multimedia The click volume discreet value of resource, including: use the formula 2 after training calculate first kind multimedia resource i the after appointed day- The logarithm value of the click volume discreet value of t days.
In this example, click volume is taken the logarithm, thus reduce numerical value, it is simple to calculate and store.In this example, according to The logarithm value of the click volume discreet value in the formula 2 calculated first kind multimedia resource i-t sky after appointed day is permissible Rank (sequence) value as first kind multimedia resource i.By calculating the rank value of the multimedia resource newly uploaded, it is possible to allow The multimedia resource of more high-qualitys newly uploaded is exposed.
Fig. 5 illustrates in the recommendation method step S103 of multimedia resource according to an embodiment of the invention for each second Class multimedia resource, searches the first kind multimedia resource mated with this Equations of The Second Kind multimedia resource, calculates the many matchmakers of this Equations of The Second Kind What body resource was exemplary with the one of the matching degree of the first kind multimedia resource mated implements flow chart.As it is shown in figure 5, For each Equations of The Second Kind multimedia resource, search the first kind multimedia resource mated with this Equations of The Second Kind multimedia resource, calculate The matching degree of this Equations of The Second Kind multimedia resource and the first kind multimedia resource mated, including:
In step S501, for each Equations of The Second Kind multimedia resource, believe according to the appointment of this Equations of The Second Kind multimedia resource Breath searches the first kind multimedia resource mated with this Equations of The Second Kind multimedia resource；Appointment information include following at least one: on Biography time, time span and uploader information.
In step S502, appointment information and the first kind multimedia mated according to this Equations of The Second Kind multimedia resource provide The appointment information in source calculates this Equations of The Second Kind multimedia resource and the matching degree of the first kind multimedia resource mated.
Wherein, the matching degree of Equations of The Second Kind multimedia resource and the first kind multimedia resource mated more than 0 and less than or Equal to 1.
As an example of the embodiment of the present invention, Equations of The Second Kind multimedia resource and when uploading of first kind multimedia resource Between closer to, matching degree the most between the two is the highest.
As an example of the embodiment of the present invention, Equations of The Second Kind multimedia resource is long with the time of first kind multimedia resource Spend closer to, matching degree the most between the two is the highest.Time span refers to the time span of multimedia resource to be recommended, example As, if this multimedia resource to be recommended is video, then the time span of this multimedia resource to be recommended also refers to regard Frequently length.
As an example of the embodiment of the present invention, if the uploading of Equations of The Second Kind multimedia resource and first kind multimedia resource Person is identical, then Equations of The Second Kind multimedia resource is higher with the matching degree of first kind multimedia resource.
As an example of the embodiment of the present invention, Equations of The Second Kind multimedia resource and the first kind multimedia resource mated Matching degree determine according to uplink time, time span and uploader information, wherein, uplink time for the weight of matching degree is λ1, time span is λ for the weight of matching degree2, uploader information is λ for the weight of matching degree3。
In a kind of possible implementation, pre-with the click volume of the first kind multimedia resource mated according to matching degree Valuation calculates the click volume discreet value of this Equations of The Second Kind multimedia resource, particularly as follows: by many with the first kind mated for matching degree matchmakers The product of the click volume discreet value of body resource is as the click volume discreet value of this Equations of The Second Kind multimedia resource.
Fig. 6 illustrates an exemplary tool of the recommendation method step S104 of multimedia resource according to an embodiment of the invention Body flowchart.As shown in Figure 6, according to the click volume discreet value of multimedia resource to be recommended, multimedia to be recommended is provided Source is recommended, including:
In step s 601, according to the click volume discreet value of multimedia resource to be recommended order from high to low to all Multimedia resource to be recommended is ranked up.
In step S602, taking out from the multimedia resource all to be recommended after sequence sorts preceding M waits to push away The multimedia resource recommended is recommended, and wherein, the size of M pre-sets.
As an example of the embodiment of the present invention, treat according to the click volume discreet value of multimedia resource to be recommended and push away The multimedia resource recommended is recommended, including: the click volume discreet value of multimedia resource to be recommended, root it is calculated by formula 1 According to the click volume discreet value order from high to low of multimedia resource to be recommended, all multimedia resources to be recommended are carried out Sequence, then from the multimedia resource all to be recommended after sequence, take out preceding M the multimedia resource to be recommended that sort Recommend.
As another example of the embodiment of the present invention, treat according to the click volume discreet value of multimedia resource to be recommended The multimedia resource recommended is recommended, including: the click volume discreet value of multimedia resource to be recommended it is calculated by formula 2 Logarithm value, according to the logarithm value of the click volume discreet value of multimedia resource to be recommended order from high to low to all to be recommended Multimedia resource be ranked up, then from the multimedia resource all to be recommended after sequence, take out the preceding M that sorts treat The multimedia resource recommended is recommended.
So, by multimedia resource to be recommended being divided into first kind multimedia resource and Equations of The Second Kind multimedia money Source, determines the click volume discreet value of first kind multimedia resource according to the history click data of first kind multimedia resource, according to The click volume discreet value of the first kind multimedia resource mated with Equations of The Second Kind multimedia resource and matching degree meter between the two Calculate the click volume discreet value of Equations of The Second Kind multimedia resource, treat pushing away further according to the click volume discreet value of multimedia resource to be recommended The multimedia resource recommended is recommended, and the recommendation method of multimedia resource according to embodiments of the present invention can make newly to upload many Media resource is also had an opportunity recommended, which thereby enhances the recommendation accuracy of multimedia resource, thus improves multimedia resource Recommendation effect.
Embodiment 2
Fig. 7 illustrates the structured flowchart of the recommendation apparatus of multimedia resource according to another embodiment of the present invention.This device can Recommendation method for the multimedia resource shown in service chart 1.For convenience of description, illustrate only in the figure 7 and the present invention The part that embodiment is relevant.
As it is shown in fig. 7, this device includes: divide module 71, for multimedia resource to be recommended is divided into the first kind Multimedia resource and Equations of The Second Kind multimedia resource；First click volume discreet value determines module 72, for for each described first Class multimedia resource, determines described first kind multimedia resource according to the history click data of described first kind multimedia resource Click volume discreet value；Second click volume discreet value determines module 73, for for each described Equations of The Second Kind multimedia resource, searches The first kind multimedia resource mated with described Equations of The Second Kind multimedia resource, calculates described Equations of The Second Kind multimedia resource and is mated The matching degree of first kind multimedia resource, and according to the click volume of described matching degree with the first kind multimedia resource mated Discreet value calculates the click volume discreet value of described Equations of The Second Kind multimedia resource；Recommending module 74, for according to described to be recommended Described multimedia resource to be recommended is recommended by the click volume discreet value of multimedia resource.
Fig. 8 illustrates an exemplary structural frames of the recommendation apparatus of multimedia resource according to another embodiment of the present invention Figure.This device may be used for the recommendation method of the multimedia resource shown in service chart 1 to Fig. 6.For convenience of description, in fig. 8 Illustrate only the part relevant to this example.The assembly that in Fig. 8, label is identical with Fig. 7 has an identical function, for simplicity's sake, Omit the detailed description to these assemblies.As shown in Figure 8:
In a kind of possible implementation, described division module 71 includes: history clicks on data acquisition submodule 711, For obtaining the history click data of each multimedia resource to be recommended respectively；First divides submodule 712, for for often Individual described multimedia resource to be recommended, waits to push away according to the history click data judgement of described multimedia resource to be recommended Whether the current click volume of the multimedia resource recommended is in the decline phase, determined by the most described multimedia resource to be recommended For first kind multimedia resource, otherwise described multimedia resource to be recommended is defined as Equations of The Second Kind multimedia resource.
In a kind of possible implementation, described division module 71 includes: uplink time obtains submodule 713, is used for Obtain the uplink time of each multimedia resource to be recommended respectively；Second divides submodule 714, for treating described in each The multimedia resource recommended, it is judged that the time of the uplink time distance present system time of described multimedia resource to be recommended is long Described multimedia resource to be recommended whether more than the first preset value, is the most then defined as first kind multimedia resource by degree, no Then described multimedia resource to be recommended is defined as Equations of The Second Kind multimedia resource.
In a kind of possible implementation, described first click volume discreet value determines that module 72 includes: history hits According to obtaining submodule 721, for for each described first kind multimedia resource, obtain going through of described first kind multimedia resource History click data；Time attenuation quotient training submodule 722, for the history hits according to described first kind multimedia resource According to the time attenuation quotient training described first kind multimedia resource；First click volume discreet value calculating sub module 723, for root History click data and described time attenuation quotient according to described first kind multimedia resource calculate described first kind multimedia money The click volume discreet value in source.
In a kind of possible implementation, described time attenuation quotient training submodule 722 specifically for:
According to the history click data of first kind multimedia resource i, employing formula 1 trains described first kind multimedia resource i Time attenuation quotient θ, wherein, the history of described first kind multimedia resource i is clicked on packet and is included described first kind multimedia Resource i click volume before appointed day；
F (t)=Vi0×eθtFormula 1；
Wherein, Vi0Represent the described first kind multimedia resource i click volume at described appointed day；T=0 represents described finger Fix the date, f (0)=Vi0；As t ＞ 0, t represents the t days before described appointed day, and f (t) represents the many matchmakers of the described first kind The click volume of before described appointed day the t days of body resource i；As t ＜ 0, t represents the-t after described appointed day My god, f (t) represents the click volume discreet value in the described first kind multimedia resource i-t sky after described appointed day；
Described first click volume discreet value calculating sub module 723 specifically for:
The formula 1 after training is used to calculate the point in the described first kind multimedia resource i-t sky after described appointed day The amount of hitting discreet value.
In a kind of possible implementation, described time attenuation quotient training submodule 722 specifically for:
According to the history click data of first kind multimedia resource i, employing formula 2 trains described first kind multimedia resource i Time attenuation quotient θ, wherein, the history of described first kind multimedia resource i is clicked on packet and is included described first kind multimedia Resource i click volume before appointed day；
G (t)=lg (Vi0×eθt) formula 2；
Wherein, Vi0Represent the described first kind multimedia resource i click volume at described appointed day；T=0 represents described finger Fix the date, g (0)=lgVi0；As t ＞ 0, t represents the t days before described appointed day, and g (t) represents that the described first kind is many The logarithm value of the click volume of before described appointed day the t days of media resource i；As t ＜ 0, t represents at described appointed day After-t sky, g (t) represents that the click volume in the described first kind multimedia resource i-t sky after described appointed day is estimated The logarithm value of value；
Described first click volume discreet value calculating sub module 723 specifically for:
The formula 2 after training is used to calculate the point in the described first kind multimedia resource i-t sky after described appointed day The logarithm value of the amount of hitting discreet value.
In a kind of possible implementation, described second click volume discreet value determines that module 73 includes: matched sub-block 731, for for each described Equations of The Second Kind multimedia resource, according to the appointment information searching of described Equations of The Second Kind multimedia resource with The first kind multimedia resource of described Equations of The Second Kind multimedia resource coupling；Described appointment information include following at least one: upload Time, time span and uploader information；Matching degree calculating sub module 732, for according to described Equations of The Second Kind multimedia resource Appointment information calculates described Equations of The Second Kind multimedia resource with the appointment information of the first kind multimedia resource mated and is mated The matching degree of first kind multimedia resource.
In a kind of possible implementation, described second click volume discreet value determines that module 73 includes:
Second click volume discreet value calculating sub module 733, for by described matching degree and the first kind multimedia mated The product of the click volume discreet value of resource is as the click volume discreet value of described Equations of The Second Kind multimedia resource.
In a kind of possible implementation, described recommending module 74 includes: sorting sub-module 741, for according to described All described multimedia resources to be recommended are entered by the click volume discreet value of multimedia resource to be recommended order from high to low Row sequence；Recommend submodule 742, preceding for taking out sequence from all described multimedia resource to be recommended after sequence M described multimedia resource to be recommended is recommended, and wherein, the size of M pre-sets.
It should be noted that so, by multimedia resource to be recommended being divided into first kind multimedia resource and Two class multimedia resources, determine the click volume of first kind multimedia resource according to the history click data of first kind multimedia resource Discreet value, according to the click volume discreet value of the first kind multimedia resource mated with Equations of The Second Kind multimedia resource and between the two Matching degree calculate Equations of The Second Kind multimedia resource click volume discreet value, the click volume further according to multimedia resource to be recommended is pre- Multimedia resource to be recommended is recommended by valuation, and the recommendation apparatus of multimedia resource according to embodiments of the present invention can make The multimedia resource newly uploaded also is had an opportunity recommended, which thereby enhances the recommendation accuracy of multimedia resource, thus improves The recommendation effect of multimedia resource.
Embodiment 3
Fig. 9 shows the structured flowchart of the recommendation apparatus of a kind of multimedia resource of an alternative embodiment of the invention.Institute Stating the recommendation apparatus 1100 of multimedia resource can be to possess the host server of computing capability, personal computer PC or can The portable computer carried or terminal etc..Calculating node is not implemented and limits by the specific embodiment of the invention.
The recommendation apparatus 1100 of described multimedia resource includes processor (processor) 1110, communication interface (Communications Interface) 1120, memorizer (memory) 1130 and bus 1140.Wherein, processor 1110, Communication interface 1120 and memorizer 1130 complete mutual communication by bus 1140.
Communication interface 1120 is used for and network device communications, and wherein the network equipment includes such as Virtual Machine Manager center, is total to Enjoy storage etc..
Processor 1110 is used for performing program.Processor 1110 is probably a central processor CPU, or special collection Become circuit ASIC (Application Specific Integrated Circuit), or be configured to implement the present invention One or more integrated circuits of embodiment.
Memorizer 1130 is used for depositing file.Memorizer 1130 may comprise high-speed RAM memorizer, it is also possible to also includes non- Volatile memory (non-volatile memory), for example, at least one disk memory.Memorizer 1130 can also be to deposit Memory array.Memorizer 1130 is also possible to by piecemeal, and described piece can be by certain rule sets synthesis virtual volume.
In a kind of possible embodiment, said procedure can be the program code including computer-managed instruction.This journey Sequence is particularly used in: realize the operation of each step in embodiment 1.
Those of ordinary skill in the art are it is to be appreciated that each exemplary cell in embodiment described herein and algorithm Step, it is possible to being implemented in combination in of electronic hardware or computer software and electronic hardware.These functions are actually with hardware also It is that software form realizes, depends on application-specific and the design constraint of technical scheme.Professional and technical personnel can be for Specific application selects different methods to realize described function, but this realization is it is not considered that exceed the model of the present invention Enclose.
If using the form of computer software realize described function and as independent production marketing or use time, then exist To a certain extent it is believed that all or part of (part such as contributed prior art) of technical scheme is Embody in form of a computer software product.This computer software product is generally stored inside the non-volatile of embodied on computer readable In storage medium, including some instructions with so that computer equipment (can be that personal computer, server or network set Standby etc.) perform all or part of step of various embodiments of the present invention method.And aforesaid storage medium include USB flash disk, portable hard drive, Read only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic The various medium that can store program code such as dish or CD.
The above, the only detailed description of the invention of the present invention, but protection scope of the present invention is not limited thereto, and any Those familiar with the art, in the technical scope that the invention discloses, can readily occur in change or replace, should contain Cover within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with described scope of the claims.

Claims (18)

1. the recommendation method of a multimedia resource, it is characterised in that including:
Multimedia resource to be recommended is divided into first kind multimedia resource and Equations of The Second Kind multimedia resource；
For each described first kind multimedia resource, determine institute according to the history click data of described first kind multimedia resource State the click volume discreet value of first kind multimedia resource；
For each described Equations of The Second Kind multimedia resource, search the first kind multimedia mated with described Equations of The Second Kind multimedia resource Resource, calculates described Equations of The Second Kind multimedia resource and the matching degree of the first kind multimedia resource mated, and according to described Degree of joining calculates the click volume of described Equations of The Second Kind multimedia resource with the click volume discreet value of the first kind multimedia resource mated Discreet value；
Described multimedia resource to be recommended is recommended by the click volume discreet value according to described multimedia resource to be recommended.
Method the most according to claim 1, it is characterised in that multimedia resource to be recommended is divided into the many matchmakers of the first kind Body resource and Equations of The Second Kind multimedia resource, including:
Obtain the history click data of each multimedia resource to be recommended respectively；
For each described multimedia resource to be recommended, sentence according to the history click data of described multimedia resource to be recommended Whether the current click volume of disconnected described multimedia resource to be recommended is in the decline phase, the most then by described many matchmakers to be recommended Body resource is defined as first kind multimedia resource, and described multimedia resource to be recommended is otherwise defined as Equations of The Second Kind multimedia money Source.
Method the most according to claim 1, it is characterised in that multimedia resource to be recommended is divided into the many matchmakers of the first kind Body resource and Equations of The Second Kind multimedia resource, including:
Obtain the uplink time of each multimedia resource to be recommended respectively；
For each described multimedia resource to be recommended, it is judged that the uplink time distance of described multimedia resource to be recommended is worked as Whether the time span of front system time more than the first preset value, is the most then defined as described multimedia resource to be recommended First kind multimedia resource, is otherwise defined as Equations of The Second Kind multimedia resource by described multimedia resource to be recommended.
Method the most according to claim 1, it is characterised in that for each described first kind multimedia resource, according to institute The history click data stating first kind multimedia resource determines the click volume discreet value of described first kind multimedia resource, including:
For each described first kind multimedia resource, obtain the history click data of described first kind multimedia resource；
History click data according to described first kind multimedia resource trains the time decay of described first kind multimedia resource Coefficient；
History click data and the described first kind of described time attenuation quotient calculating according to described first kind multimedia resource are many The click volume discreet value of media resource.
Method the most according to claim 4, it is characterised in that according to the history hits of described first kind multimedia resource According to train described first kind multimedia resource time attenuation quotient, including:
According to the history click data of first kind multimedia resource i, employing formula 1 train described first kind multimedia resource i time Between attenuation quotient θ, wherein, the history of described first kind multimedia resource i is clicked on packet and is included described first kind multimedia resource i Click volume before appointed day；
F (t)=Vi0×eθtFormula 1；
Wherein, Vi0Represent the described first kind multimedia resource i click volume at described appointed day；T=0 represents described appointment day Phase, f (0)=Vi0；As t ＞ 0, t represents the t days before described appointed day, and f (t) represents that described first kind multimedia provides The click volume of before described appointed day the t days of source i；As t ＜ 0, t represents the-t sky after described appointed day, f T () represents the click volume discreet value in the described first kind multimedia resource i-t sky after described appointed day；
History click data and the described first kind of described time attenuation quotient calculating according to described first kind multimedia resource are many The click volume discreet value of media resource, including:
The formula 1 after training is used to calculate the click volume in the described first kind multimedia resource i-t sky after described appointed day Discreet value.
Method the most according to claim 4, it is characterised in that according to the history hits of described first kind multimedia resource According to train described first kind multimedia resource time attenuation quotient, including:
According to the history click data of first kind multimedia resource i, employing formula 2 train described first kind multimedia resource i time Between attenuation quotient θ, wherein, the history of described first kind multimedia resource i is clicked on packet and is included described first kind multimedia resource i Click volume before appointed day；
G (t)=lg (Vi0×eθt) formula 2；
Wherein, Vi0Represent the described first kind multimedia resource i click volume at described appointed day；T=0 represents described appointment day Phase, g (0)=lgVi0；As t ＞ 0, t represents the t days before described appointed day, and g (t) represents described first kind multimedia The logarithm value of the click volume of before described appointed day the t days of resource i；As t ＜ 0, t represents after described appointed day -t sky, g (t) represents the click volume discreet value in the described first kind multimedia resource i-t sky after described appointed day Logarithm value；
History click data and the described first kind of described time attenuation quotient calculating according to described first kind multimedia resource are many The click volume discreet value of media resource, including:
The formula 2 after training is used to calculate the click volume in the described first kind multimedia resource i-t sky after described appointed day The logarithm value of discreet value.
Method the most according to claim 1, it is characterised in that for each described Equations of The Second Kind multimedia resource, search with The first kind multimedia resource of described Equations of The Second Kind multimedia resource coupling, calculates described Equations of The Second Kind multimedia resource and is mated The matching degree of first kind multimedia resource, including:
For each described Equations of The Second Kind multimedia resource, according to the appointment information searching of described Equations of The Second Kind multimedia resource with described The first kind multimedia resource of Equations of The Second Kind multimedia resource coupling；Described appointment information include following at least one: uplink time, Time span and uploader information；
Appointment information according to described Equations of The Second Kind multimedia resource and the appointment information meter of first kind multimedia resource mated The matching degree of the first kind multimedia resource calculating described Equations of The Second Kind multimedia resource and mated.
8. according to the method described in claim 1 to 7 any one, it is characterised in that according to described matching degree with mated The click volume discreet value of first kind multimedia resource calculates the click volume discreet value of described Equations of The Second Kind multimedia resource, particularly as follows:
Using the product of described matching degree and the click volume discreet value of the first kind multimedia resource mated as described Equations of The Second Kind The click volume discreet value of multimedia resource.
9. according to the method described in claim 1 to 7 any one, it is characterised in that according to described multimedia money to be recommended Described multimedia resource to be recommended is recommended by the click volume discreet value in source, including:
Click volume discreet value according to described multimedia resource to be recommended order from high to low is to all described to be recommended Multimedia resource is ranked up；
Preceding M the described multimedia to be recommended that sort is taken out from all described multimedia resource to be recommended after sequence Resource is recommended, and wherein, the size of M pre-sets.
10. the recommendation apparatus of a multimedia resource, it is characterised in that including:
Divide module, for multimedia resource to be recommended being divided into first kind multimedia resource and Equations of The Second Kind multimedia money Source；
First click volume discreet value determines module, for for each described first kind multimedia resource, according to the described first kind The history click data of multimedia resource determines the click volume discreet value of described first kind multimedia resource；
Second click volume discreet value determines module, for for each described Equations of The Second Kind multimedia resource, searches and described second The first kind multimedia resource of class multimedia resource coupling, calculates described Equations of The Second Kind multimedia resource many with the first kind mated The matching degree of media resource, and calculate with the click volume discreet value of the first kind multimedia resource mated according to described matching degree The click volume discreet value of described Equations of The Second Kind multimedia resource；
Recommending module, is used for the click volume discreet value according to described multimedia resource to be recommended to described multimedia to be recommended Resource is recommended.
11. devices according to claim 10, it is characterised in that described division module includes:
History clicks on data acquisition submodule, for obtaining the history click data of each multimedia resource to be recommended respectively；
First divides submodule, for for each described multimedia resource to be recommended, according to described multimedia to be recommended The history click data of resource judges whether the current click volume of described multimedia resource to be recommended is in the decline phase, if so, Then described multimedia resource to be recommended is defined as first kind multimedia resource, otherwise by described multimedia resource to be recommended It is defined as Equations of The Second Kind multimedia resource.
12. devices according to claim 10, it is characterised in that described division module includes:
Uplink time obtains submodule, for obtaining the uplink time of each multimedia resource to be recommended respectively；
Second divides submodule, for for each described multimedia resource to be recommended, it is judged that described multimedia to be recommended Whether the time span of the uplink time distance present system time of resource more than the first preset value, the most then pushes away described waiting The multimedia resource recommended is defined as first kind multimedia resource, otherwise described multimedia resource to be recommended is defined as Equations of The Second Kind Multimedia resource.
13. devices according to claim 10, it is characterised in that described first click volume discreet value determines that module includes:
History clicks on data acquisition submodule, for for each described first kind multimedia resource, obtains the described first kind many The history click data of media resource；
Time attenuation quotient training submodule, trains described for the history click data according to described first kind multimedia resource The time attenuation quotient of first kind multimedia resource；
First click volume discreet value calculating sub module, for the history click data according to described first kind multimedia resource and institute State time attenuation quotient and calculate the click volume discreet value of described first kind multimedia resource.
14. devices according to claim 13, it is characterised in that described time attenuation quotient training submodule is specifically used In:
According to the history click data of first kind multimedia resource i, employing formula 1 train described first kind multimedia resource i time Between attenuation quotient θ, wherein, the history of described first kind multimedia resource i is clicked on packet and is included described first kind multimedia resource i Click volume before appointed day；
F (t)=Vi0×eθtFormula 1；
Wherein, Vi0Represent the described first kind multimedia resource i click volume at described appointed day；T=0 represents described appointment day Phase, f (0)=Vi0；As t ＞ 0, t represents the t days before described appointed day, and f (t) represents that described first kind multimedia provides The click volume of before described appointed day the t days of source i；As t ＜ 0, t represents the-t sky after described appointed day, f T () represents the click volume discreet value in the described first kind multimedia resource i-t sky after described appointed day；
Described first click volume discreet value calculating sub module specifically for:
The formula 1 after training is used to calculate the click volume in the described first kind multimedia resource i-t sky after described appointed day Discreet value.
15. devices according to claim 13, it is characterised in that described time attenuation quotient training submodule is specifically used In:
According to the history click data of first kind multimedia resource i, employing formula 2 train described first kind multimedia resource i time Between attenuation quotient θ, wherein, the history of described first kind multimedia resource i is clicked on packet and is included described first kind multimedia resource i Click volume before appointed day；
G (t)=lg (Vi0×eθt) formula 2；
Wherein, Vi0Represent the described first kind multimedia resource i click volume at described appointed day；T=0 represents described appointment day Phase, g (0)=lgVi0；As t ＞ 0, t represents the t days before described appointed day, and g (t) represents described first kind multimedia The logarithm value of the click volume of before described appointed day the t days of resource i；As t ＜ 0, t represents after described appointed day -t sky, g (t) represents the click volume discreet value in the described first kind multimedia resource i-t sky after described appointed day Logarithm value；
Described first click volume discreet value calculating sub module specifically for:
The formula 2 after training is used to calculate the click volume in the described first kind multimedia resource i-t sky after described appointed day The logarithm value of discreet value.
16. devices according to claim 10, it is characterised in that described second click volume discreet value determines that module includes:
Matched sub-block, for for each described Equations of The Second Kind multimedia resource, according to the finger of described Equations of The Second Kind multimedia resource Determine the first kind multimedia resource that information searching mates with described Equations of The Second Kind multimedia resource；Described appointment information include with down to One item missing: uplink time, time span and uploader information；
Matching degree calculating sub module is many for the appointment information according to described Equations of The Second Kind multimedia resource and the first kind mated The appointment information of media resource calculates the matching degree of described Equations of The Second Kind multimedia resource and the first kind multimedia resource mated.
17. according to the device described in claim 10 to 16 any one, it is characterised in that described second click volume discreet value is true Cover half block includes:
Second click volume discreet value calculating sub module, for by the point of described matching degree with the first kind multimedia resource mated The product of the amount of hitting discreet value is as the click volume discreet value of described Equations of The Second Kind multimedia resource.
18. according to the device described in claim 10 to 16 any one, it is characterised in that described recommending module includes:
Sorting sub-module, is used for the order from high to low of the click volume discreet value according to described multimedia resource to be recommended to institute Described multimedia resource to be recommended is had to be ranked up；
Recommend submodule, for taking out preceding M the institute that sort from all described multimedia resource to be recommended after sequence Stating multimedia resource to be recommended to recommend, wherein, the size of M pre-sets.
CN201610594958.7A 2016-07-26 2016-07-26 Multimedia resource recommendation method and device Expired - Fee Related CN106227834B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610594958.7A CN106227834B (en) 2016-07-26 2016-07-26 Multimedia resource recommendation method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610594958.7A CN106227834B (en) 2016-07-26 2016-07-26 Multimedia resource recommendation method and device

Publications (2)

Publication Number Publication Date
CN106227834A true CN106227834A (en) 2016-12-14
CN106227834B CN106227834B (en) 2020-03-27

Family

ID=57534156

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610594958.7A Expired - Fee Related CN106227834B (en) 2016-07-26 2016-07-26 Multimedia resource recommendation method and device

Country Status (1)

Country Link
CN (1) CN106227834B (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107748787A (en) * 2017-10-27 2018-03-02 姜俊 A kind of heatrate processing method and processing device
CN108804439A (en) * 2017-04-26 2018-11-13 合信息技术（北京）有限公司 The recommendation method and device of multimedia resource
CN108804452A (en) * 2017-04-28 2018-11-13 合信息技术（北京）有限公司 Multimedia resource cover methods of exhibiting and device
CN109068158A (en) * 2018-09-18 2018-12-21 苏州商信宝信息科技有限公司 A kind of short-sighted frequency value estimate recommender system based on the network platform
WO2019024670A1 (en) * 2017-08-03 2019-02-07 优酷信息技术（北京）有限公司 Multimedia resource recommendation method and device
CN110909176A (en) * 2019-11-20 2020-03-24 腾讯科技（深圳）有限公司 Data recommendation method and device, computer equipment and storage medium
CN110990690A (en) * 2019-11-12 2020-04-10 上海易点时空网络有限公司 Post recommendation method and device
CN111199412A (en) * 2018-11-19 2020-05-26 腾讯科技（深圳）有限公司 Information processing method, information processing device, computer equipment and storage medium
CN112948402A (en) * 2021-01-15 2021-06-11 浙江大华技术股份有限公司 Database updating method and device, electronic equipment and computer readable storage medium
CN113487368A (en) * 2021-07-29 2021-10-08 宿迁市盛邦信息科技有限公司 Scientific and technological information push service system based on big data
CN116389444A (en) * 2023-04-10 2023-07-04 北京智享嘉网络信息技术有限公司 Traffic scheduling method and system based on user web application

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050288954A1 (en) * 2000-10-19 2005-12-29 Mccarthy John Method, system and personalized web content manager responsive to browser viewers' psychological preferences, behavioral responses and physiological stress indicators
CN103164481A (en) * 2011-12-16 2013-06-19 盛乐信息技术（上海）有限公司 Recommendation method and system of video with largest rising trend
CN104520881A (en) * 2012-06-22 2015-04-15 谷歌公司 Ranking nearby destinations based on visit likelihoods and predicting future visits to places from location history
CN105184618A (en) * 2015-10-20 2015-12-23 广州唯品会信息科技有限公司 Commodity individual recommendation method for new users and system
CN105447045A (en) * 2014-09-02 2016-03-30 阿里巴巴集团控股有限公司 Information ordering method and device and information providing method and system
CN105760530A (en) * 2016-03-03 2016-07-13 北京百度网讯科技有限公司 Information card sequencing method and device based on artificial intelligence

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050288954A1 (en) * 2000-10-19 2005-12-29 Mccarthy John Method, system and personalized web content manager responsive to browser viewers' psychological preferences, behavioral responses and physiological stress indicators
CN103164481A (en) * 2011-12-16 2013-06-19 盛乐信息技术（上海）有限公司 Recommendation method and system of video with largest rising trend
CN104520881A (en) * 2012-06-22 2015-04-15 谷歌公司 Ranking nearby destinations based on visit likelihoods and predicting future visits to places from location history
CN105447045A (en) * 2014-09-02 2016-03-30 阿里巴巴集团控股有限公司 Information ordering method and device and information providing method and system
CN105184618A (en) * 2015-10-20 2015-12-23 广州唯品会信息科技有限公司 Commodity individual recommendation method for new users and system
CN105760530A (en) * 2016-03-03 2016-07-13 北京百度网讯科技有限公司 Information card sequencing method and device based on artificial intelligence

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108804439A (en) * 2017-04-26 2018-11-13 合信息技术（北京）有限公司 The recommendation method and device of multimedia resource
CN108804452B (en) * 2017-04-28 2021-06-04 阿里巴巴（中国）有限公司 Multimedia resource cover display method and device
CN108804452A (en) * 2017-04-28 2018-11-13 合信息技术（北京）有限公司 Multimedia resource cover methods of exhibiting and device
WO2019024670A1 (en) * 2017-08-03 2019-02-07 优酷信息技术（北京）有限公司 Multimedia resource recommendation method and device
CN109388739A (en) * 2017-08-03 2019-02-26 合信息技术（北京）有限公司 The recommended method and device of multimedia resource
CN107748787A (en) * 2017-10-27 2018-03-02 姜俊 A kind of heatrate processing method and processing device
CN109068158A (en) * 2018-09-18 2018-12-21 苏州商信宝信息科技有限公司 A kind of short-sighted frequency value estimate recommender system based on the network platform
CN109068158B (en) * 2018-09-18 2021-06-22 上海众引文化传播股份有限公司 Short video value estimation recommendation system based on network platform
CN111199412A (en) * 2018-11-19 2020-05-26 腾讯科技（深圳）有限公司 Information processing method, information processing device, computer equipment and storage medium
CN110990690A (en) * 2019-11-12 2020-04-10 上海易点时空网络有限公司 Post recommendation method and device
CN110909176A (en) * 2019-11-20 2020-03-24 腾讯科技（深圳）有限公司 Data recommendation method and device, computer equipment and storage medium
CN112948402A (en) * 2021-01-15 2021-06-11 浙江大华技术股份有限公司 Database updating method and device, electronic equipment and computer readable storage medium
CN113487368A (en) * 2021-07-29 2021-10-08 宿迁市盛邦信息科技有限公司 Scientific and technological information push service system based on big data
CN116389444A (en) * 2023-04-10 2023-07-04 北京智享嘉网络信息技术有限公司 Traffic scheduling method and system based on user web application
CN116389444B (en) * 2023-04-10 2023-09-15 北京智享嘉网络信息技术有限公司 Traffic scheduling method and system based on user web application

Also Published As

Publication number Publication date
CN106227834B (en) 2020-03-27

Similar Documents

Publication Publication Date Title
CN106227834A (en) The recommendation method and device of multimedia resource
US10405016B2 (en) Recommending media items based on take rate signals
US11151203B2 (en) Interest embedding vectors
US8990241B2 (en) System and method for recommending queries related to trending topics based on a received query
CN103377232B (en) Headline keyword recommendation method and system
US11294974B1 (en) Golden embeddings
EP2438539B1 (en) Co-selected image classification
US11023506B2 (en) Query pattern matching
US8892554B2 (en) Automatic word-cloud generation
US20180246973A1 (en) User interest modeling
EP3021264A1 (en) Information recommendation method and apparatus in social media
US20180246899A1 (en) Generate an index for enhanced search based on user interests
US9183499B1 (en) Evaluating quality based on neighbor features
CN106294830A (en) The recommendation method and device of multimedia resource
US9171045B2 (en) Recommending queries according to mapping of query communities
US20180246972A1 (en) Enhanced search to generate a feed based on a user's interests
EP2480995A1 (en) Searching for information based on generic attributes of the query
CN108132963A (en) Resource recommendation method and device, computing device and storage medium
CN103678336B (en) Method and device for identifying entity words
CN104217030A (en) Method and device for classifying users according to search log data of server
US11789946B2 (en) Answer facts from structured content
CN106354867A (en) Multimedia resource recommendation method and device
US20160012130A1 (en) Aiding composition of themed articles about popular and novel topics and offering users a navigable experience of associated content
CN109819002B (en) Data pushing method and device, storage medium and electronic device
CN107784127A (en) A kind of focus localization method and device

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CP03 Change of name, title or address

Address after: 100080 Beijing Haidian District city Haidian street A Sinosteel International Plaza No. 8 block 5 layer A, C

Patentee after: Youku network technology (Beijing) Co.,Ltd.

Address before: 100080 area a and C, 5 / F, block a, Sinosteel International Plaza, No. 8, Haidian Street, Haidian District, Beijing

Patentee before: 1VERGE INTERNET TECHNOLOGY (BEIJING) Co.,Ltd.

CP03 Change of name, title or address
TR01 Transfer of patent right

Effective date of registration: 20200520

Address after: 310052 room 508, floor 5, building 4, No. 699, Wangshang Road, Changhe street, Binjiang District, Hangzhou City, Zhejiang Province

Patentee after: Alibaba (China) Co.,Ltd.

Address before: 100080 Beijing Haidian District city Haidian street A Sinosteel International Plaza No. 8 block 5 layer A, C

Patentee before: Youku network technology (Beijing) Co.,Ltd.

TR01 Transfer of patent right
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20200327

Termination date: 20200726

CF01 Termination of patent right due to non-payment of annual fee