CN107613391A - Method is recommended in a kind of association advertisement based on video content - Google Patents
Method is recommended in a kind of association advertisement based on video content Download PDFInfo
- Publication number
- CN107613391A CN107613391A CN201610547325.0A CN201610547325A CN107613391A CN 107613391 A CN107613391 A CN 107613391A CN 201610547325 A CN201610547325 A CN 201610547325A CN 107613391 A CN107613391 A CN 107613391A
- Authority
- CN
- China
- Prior art keywords
- advertisement
- commodity
- association
- video
- video content
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Abstract
The invention discloses a kind of association advertisement based on video content to recommend method, comprises the following steps:Define video frequency program and advertisement/commodity are expressed with multi-C vector respectively;Define a learning sample and construct the matrix of video frequency program and the matrix of advertisement/commodity respectively according to the learning sample;By calculating the incidence matrix between acquisition video frequency program and advertisement/commodity;Mapping between video content and advertisement/commodity is created according to incidence matrix;The association advertisement that video content is carried out according to mapping is recommended;It is that the sequencing that advertisement/commodity are realized in video display process is recommended by building the association between video and its content and advertisement/commodity;Further, the association between time interval and advertisement/commodity is built;Realize that the sequencing of advertisement/commodity is recommended to combine the feature of place time interval in video display process.
Description
Technical field
The present invention relates to video display arts field, more particularly to a kind of association advertisement recommendation side based on video content
Method.
Background technology
Recommendation of the video to video is currently, there are, there is also the recommendation of commodity to commodity;But it is the absence of a kind of effective side
Method is based on video and its content to the recommendation between advertisement and its commodity;The means of existing associated video and advertisement and commodity are basic
All it is by the way of artificial layout, needs a kind of method for programming badly to automate the flow.
In addition, from the point of view of the selection of the period of advertisement/product promotion will be according to target audience, if advertisement is directed to
It is virgin, then to select the front and rear just relatively good of cartoon broadcast time, can so reach the arrival rate of advertisement to greatest extent, such as
The audient of fruit advertisement is working clan, then advertising broadcast time can not necessarily be selected in daytime, because daytime, people were on duty, not have
Having time sees TV, listens to the radio programme, paper of reading the newspaper.So the method for this sequencing allows for taking into full account and decision-making is when different
Between section be adapted to promote advertisement or commodity.
The content of the invention
In view of the above-mentioned deficiency that presently, there are, the present invention provides a kind of association advertisement based on video content and recommends method,
The association between video and its content and advertisement/commodity can be built, is the journey that advertisement/commodity are realized in video display process
Sequenceization is recommended.
To reach above-mentioned purpose, embodiments of the invention adopt the following technical scheme that:
Method is recommended in a kind of association advertisement based on video content, and method is recommended in the association advertisement based on video content
Comprise the following steps:
Define video frequency program and advertisement/commodity are expressed with multi-C vector respectively;
Define a learning sample and construct the matrix of video frequency program and the square of advertisement/commodity respectively according to the learning sample
Battle array;
By calculating the incidence matrix between acquisition video frequency program and advertisement/commodity;
Mapping between video content and advertisement/commodity is created according to incidence matrix;
The association advertisement that video content is carried out according to mapping is recommended.
According to one aspect of the present invention, the association advertisement based on video content recommends method to include:Pass through weight
Post-process further to adjust the value of each row of matrix.
According to one aspect of the present invention, each dimension of the multi-C vector of the video frequency program is that a kind of uniqueness is regarded
The expression of frequency marking label or video content gene.
According to one aspect of the present invention, each dimension of the multi-C vector of the advertisement/commodity is to a kind of unique
Advertisement or the expression of Commercial goods labelses or its attribution gene.
According to one aspect of the present invention, the association advertisement based on video content recommends method to include:
The definition period is expressed with multi-C vector;
Define a learning sample and the matrix of build time section and the matrix of advertisement/commodity are distinguished according to the learning sample;
By calculating the incidence matrix between acquisition period and advertisement/commodity;
According to the mapping between incidence matrix creation time section and advertisement/commodity;
The association advertisement carried out according to mapping in the period is recommended.
According to one aspect of the present invention, each dimension of the multi-C vector of the period is to a kind of unique
The expression of chronogeometry.
The advantages of present invention is implemented:Association advertisement of the present invention based on video content recommends method to include following step
Suddenly:Define video frequency program and advertisement/commodity are expressed with multi-C vector respectively;Define a learning sample and according to the study sample
This constructs the matrix of video frequency program and the matrix of advertisement/commodity respectively;Obtained by calculating between video frequency program and advertisement/commodity
Incidence matrix;Mapping between video content and advertisement/commodity is created according to incidence matrix;Video content is carried out according to mapping
Association advertisement recommend;It is real in video display process by building the association between video and its content and advertisement/commodity
The sequencing of existing advertisement/commodity is recommended;Further, the association between time interval and advertisement/commodity is built;To be broadcast in video
Feature during putting with reference to place time interval realizes that the sequencing of advertisement/commodity is recommended.
Brief description of the drawings
Technical scheme in order to illustrate the embodiments of the present invention more clearly, it will use below required in embodiment
Accompanying drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for ability
For the those of ordinary skill of domain, on the premise of not paying creative work, it can also be obtained according to these accompanying drawings other attached
Figure.
Fig. 1 is that method schematic diagram is recommended in a kind of association advertisement based on video content of the present invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other under the premise of creative work is not made
Embodiment, belong to the scope of protection of the invention.
As shown in figure 1, a kind of association advertisement recommendation method based on video content, the association based on video content are wide
Recommendation method is accused to comprise the following steps:
Step S1:Define video frequency program and advertisement/commodity are expressed with multi-C vector respectively;
In actual applications, each dimension of the multi-C vector of the video frequency program be to a kind of distinct video label or
The expression of video content gene.
In actual applications, each dimension of the multi-C vector of the advertisement/commodity is to a kind of unique ads or business
The expression of product label or its attribution gene.
Step S2:Define a learning sample and construct matrix and the advertisement/business of video frequency program respectively according to the learning sample
The matrix of product;
Step S3:By calculating the incidence matrix between acquisition video frequency program and advertisement/commodity;
In actual applications, the association advertisement based on video content recommends method to include:By weight post processing come
Further adjust the value of each row of matrix.
Step S4:Mapping between video content and advertisement/commodity is created according to incidence matrix;
Step S5:The association advertisement that video content is carried out according to mapping is recommended.
In actual applications, the association advertisement based on video content recommends method to include:
The definition period is expressed with multi-C vector;
Define a learning sample and the matrix of build time section and the matrix of advertisement/commodity are distinguished according to the learning sample;
By calculating the incidence matrix between acquisition period and advertisement/commodity;
According to the mapping between incidence matrix creation time section and advertisement/commodity;
The association advertisement carried out according to mapping in the period is recommended.
In actual applications, each dimension of the multi-C vector of the period is to a kind of unique chronogeometry
Expression.
1) association advertisement/commercial product recommending based on video content (label or content gene) of time factor is not considered;
Any one video frequency program can be expressed as the vector of a n dimension, i.e. a1=(a11, a12 ..., a1n);Its
In each dimension be a kind of expression to distinct video label or video content gene;
Any one advertisement/commodity can be expressed as the vector of a m dimension, i.e. b1=(b11, b12 ..., b1m);Its
In each dimension be a kind of expression to unique ads or Commercial goods labelses or its attribution gene;
Any one learning sample can be expressed as (ai, bj), i.e., in displaying video programs ai, while played wide
Accuse bj or carry out commodity bj popularization;
It is now assumed that the sample number of our learning sample set is N, then constructing
1) matrix A of a Nxn dimension, every a line of matrix is all a video frequency program;
2) matrix B of a Nxm dimension, every a line of matrix is all an advertisement/commodity;
It is that A transposition is multiplied by matrix B to calculate C=(A^t) * B, the matrixes of obtained nxm dimensions be exactly a video frequency program with
Incidence matrix between advertisement/commodity.Its transposition is exactly the incidence matrix of advertisement/between commodity and program.In order to eliminate difference
Arrange weight difference between (i.e. different advertisement/commodity genes) it is excessive caused by the diversified deficient problem of association, we can be with
The value of each row is further adjusted by weight post processing.
Example 1
Every a line of the matrix all represents a video frequency program, and N represents the sample number as training.
Example 2
Every a line of the matrix all represents an advertisement or commodity, and N represents the sample number as training.
Example 3
Every a line of the matrix all represents a video tab or content gene and the label or base of an advertisement or commodity
Therefore the degree of association between, each row of the matrix all represent an advertisement or the label or gene of commodity and a video tab or
The degree of association between content gene.
Example 4
For any one video frequency program, if its vector is expressed as
X=[x1, x2,...,xn]
Label/gene vectors of advertisement/commodity that so its mapping obtains are exactly
Y=[y1, y2,...,ym]=x*C
Example 5
For a given items list, if the items list has L commodity, then by calculating this L commodity
Value recommended with regard to each commodity can be obtained the distance between above-mentioned vectorial y;It can just obtain by value sequence and be specified in broadcasting
It is adapted to the items list recommended during video.
2) consider that weight calculation is recommended in the advertisement of time factor;
The period of one day is divided into n decile, such as with 30 minutes for granularity, then 24 hours are exactly 48 areas
Between, define time segment description amount, t1=(t11, t12 ..., t1n);Each of which dimension is all to a kind of unique time
The expression of gene;
Any one advertisement/commodity can be expressed as the vector of a m dimension, i.e. b1=(b11, b12 ..., b1m);Its
In each dimension be to a kind of unique advertisement or the expression of Commercial goods labelses or its attribution gene;
Any one learning sample can be expressed as (ti, bj), i.e., in time period t i, played advertisement or to commodity bj
Promoted;
It is now assumed that the sample number of our learning sample set is N, then constructing
1.1) the matrix T of a Nxn dimension, every a line of matrix is all a period description vectors;
1.2) matrix B of a Nxm dimension, every a line of matrix is all an advertisement or commodity;
Calculate the transposition that C'=(T^t) * B are T and be multiplied by matrix B, the matrix of obtained nxm dimensions, be exactly in terms of capable dimension
Some period, there is every series advertisements/Commercial goods labelses or the probability of its attribution gene;In terms of the dimension of row be exactly some advertisement/
The weight that Commercial goods labelses or its attribution gene are played in different time sections.
Example 1
The time interval (gene) that every a line of the matrix all represents an advertisement or commodity are played, such as one week
7x24 hours be used as unit interval using 30 minutes, then have 336 unit intervals once week, i.e., n herein is 336.N is represented
Sample number as training.
Example 2
Every a line of the matrix all represents an advertisement or commodity, and N represents the sample number as training.
Example 3
Every a line of the matrix all represent a time interval (gene) and the label or gene of an advertisement or commodity it
Between the degree of association, each row of the matrix all represent an advertisement or the label or gene of commodity and a time interval (gene)
Between the degree of association.
Example 4
For any one time interval, if its vector is expressed as
X=[x1, x2,...,xn]
Label/gene vectors of advertisement/commodity that so its mapping obtains are exactly,
Y=[y1, y2,...,ym]=x*C
Example 5
For a given items list, if the items list has L commodity,
So by calculating the distance between this L commodity and above-mentioned vectorial y with regard to each commodity can be obtained in preset time
The recommended value in section;By value sequence just can obtain specified time section be adapted to the items list of recommendation.
When there is a video playback, the vector and video playback of video tab or its content gene can be passed through
Chronogeometry, obtain being best suitable for association advertisement or the business played when at the appointed time section plays certain video by above-mentioned matrix operation
Product are recommended.
The advantages of present invention is implemented:Association advertisement of the present invention based on video content recommends method to include following step
Suddenly:Define video frequency program and advertisement/commodity are expressed with multi-C vector respectively;Define a learning sample and according to the study sample
This constructs the matrix of video frequency program and the matrix of advertisement/commodity respectively;Obtained by calculating between video frequency program and advertisement/commodity
Incidence matrix;Mapping between video content and advertisement/commodity is created according to incidence matrix;Video content is carried out according to mapping
Association advertisement recommend;It is real in video display process by building the association between video and its content and advertisement/commodity
The sequencing of existing advertisement/commodity is recommended;Further, the association between time interval and advertisement/commodity is built;To be broadcast in video
Feature during putting with reference to place time interval realizes that the sequencing of advertisement/commodity is recommended.
The foregoing is only a specific embodiment of the invention, but protection scope of the present invention is not limited thereto, any
Those skilled in the art is in technical scope disclosed by the invention, the change or replacement that can readily occur in, all should
It is included within the scope of the present invention.Therefore, protection scope of the present invention should using the scope of the claims as
It is accurate.
Claims (6)
1. method is recommended in a kind of association advertisement based on video content, it is characterised in that the association based on video content is wide
Recommendation method is accused to comprise the following steps:
Define video frequency program and advertisement/commodity are expressed with multi-C vector respectively;
Define a learning sample and construct the matrix of video frequency program and the matrix of advertisement/commodity respectively according to the learning sample;
By calculating the incidence matrix between acquisition video frequency program and advertisement/commodity;
Mapping between video content and advertisement/commodity is created according to incidence matrix;
The association advertisement that video content is carried out according to mapping is recommended.
2. method is recommended in the association advertisement according to claim 1 based on video content, it is characterised in that described to be based on regarding
The association advertisement of frequency content recommends method to include:The value of each row of matrix is further adjusted by weight post processing.
3. method is recommended in the association advertisement according to claim 1 based on video content, it is characterised in that the video section
Each dimension of purpose multi-C vector is the expression to a kind of distinct video label or video content gene.
4. method is recommended in the association advertisement according to claim 1 based on video content, it is characterised in that the advertisement/
Each dimension of the multi-C vector of commodity is the expression to a kind of unique ads or Commercial goods labelses or its attribution gene.
5. method is recommended in the association advertisement based on video content according to one of Claims 1-4, it is characterised in that institute
Stating the association advertisement based on video content recommends method to include:
The definition period is expressed with multi-C vector;
Define a learning sample and the matrix of build time section and the matrix of advertisement/commodity are distinguished according to the learning sample;
By calculating the incidence matrix between acquisition period and advertisement/commodity;
According to the mapping between incidence matrix creation time section and advertisement/commodity;
The association advertisement carried out according to mapping in the period is recommended.
6. method is recommended in the association advertisement according to claim 5 based on video content, it is characterised in that the period
Each dimension of multi-C vector be to a kind of expression of unique chronogeometry.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610547325.0A CN107613391A (en) | 2016-07-12 | 2016-07-12 | Method is recommended in a kind of association advertisement based on video content |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610547325.0A CN107613391A (en) | 2016-07-12 | 2016-07-12 | Method is recommended in a kind of association advertisement based on video content |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107613391A true CN107613391A (en) | 2018-01-19 |
Family
ID=61054995
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610547325.0A Pending CN107613391A (en) | 2016-07-12 | 2016-07-12 | Method is recommended in a kind of association advertisement based on video content |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107613391A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108259949A (en) * | 2018-02-11 | 2018-07-06 | 北京未来媒体科技股份有限公司 | Method, apparatus and electronic equipment are recommended in a kind of advertisement |
CN111629273A (en) * | 2020-04-14 | 2020-09-04 | 北京奇艺世纪科技有限公司 | Video management method, device, system and storage medium |
CN112135193A (en) * | 2020-09-24 | 2020-12-25 | 湖南快乐阳光互动娱乐传媒有限公司 | Video recommendation method and device |
CN112348566A (en) * | 2020-10-15 | 2021-02-09 | 北京捷通华声科技股份有限公司 | Method and device for determining recommended advertisements and storage medium |
-
2016
- 2016-07-12 CN CN201610547325.0A patent/CN107613391A/en active Pending
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108259949A (en) * | 2018-02-11 | 2018-07-06 | 北京未来媒体科技股份有限公司 | Method, apparatus and electronic equipment are recommended in a kind of advertisement |
CN111629273A (en) * | 2020-04-14 | 2020-09-04 | 北京奇艺世纪科技有限公司 | Video management method, device, system and storage medium |
CN111629273B (en) * | 2020-04-14 | 2022-02-11 | 北京奇艺世纪科技有限公司 | Video management method, device, system and storage medium |
CN112135193A (en) * | 2020-09-24 | 2020-12-25 | 湖南快乐阳光互动娱乐传媒有限公司 | Video recommendation method and device |
CN112135193B (en) * | 2020-09-24 | 2022-06-07 | 湖南快乐阳光互动娱乐传媒有限公司 | Video recommendation method and device |
CN112348566A (en) * | 2020-10-15 | 2021-02-09 | 北京捷通华声科技股份有限公司 | Method and device for determining recommended advertisements and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107613391A (en) | Method is recommended in a kind of association advertisement based on video content | |
CN104869173B (en) | For the information-pushing method and device of outdoor large billboard | |
Pfeiffer et al. | Can old media enhance new media?: How traditional advertising pays off for an online social network | |
US11758242B2 (en) | Targeting television advertisements based on automatic optimization of demographic information | |
CN107798560A (en) | A kind of retail shop's individual character advertisement intelligent method for pushing and system | |
CN106709765A (en) | Advertisement delivery management method | |
CN103814576A (en) | Method and apparatus for delivering targeted content to television viewers | |
CN105448214A (en) | Advertisement display method and device of controllable electronic screen | |
KR20090089919A (en) | Network node ad targeting | |
KR20190075141A (en) | Multi-dimensional advertisement bidding | |
CN106547767A (en) | Determine the method and device of video front cover picture | |
US20130085822A1 (en) | Mutually Advertising System, Advertisement Distribution Planning Apparatus and Method, and Computer Program | |
CN109543132A (en) | Content recommendation method, device, electronic equipment and storage medium | |
CN104700288A (en) | Internet advertising method and device | |
CN107025570B (en) | Method for pushing advertisements in cinema | |
CN103209341B (en) | The display packing of a kind of volume bar, system and digital TV terminal | |
CN112184338B (en) | Advertisement transaction system based on big data | |
Song et al. | Conjoint analysis for IPTV service | |
US20220108354A1 (en) | Autonomous system and method of combining of digital signage and mobile media content | |
CN103379161A (en) | Method, system and device for displaying medium information | |
US20090177547A1 (en) | Method for providing personalized advertisements | |
CN105072204A (en) | Information pushing method and device | |
CN105741147A (en) | APP (Application) shop promotion system | |
Zahrádka et al. | What is the price of outdoor advertising: A case study of the Czech republic? | |
CN105992015A (en) | Information processing method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180119 |
|
WD01 | Invention patent application deemed withdrawn after publication |