CN106407241A - Video recommendation method and system - Google Patents
Video recommendation method and system Download PDFInfo
- Publication number
- CN106407241A CN106407241A CN201610162038.8A CN201610162038A CN106407241A CN 106407241 A CN106407241 A CN 106407241A CN 201610162038 A CN201610162038 A CN 201610162038A CN 106407241 A CN106407241 A CN 106407241A
- Authority
- CN
- China
- Prior art keywords
- video
- preferred contents
- contents label
- target recommended
- preference
- 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
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
Abstract
The invention provides a video recommendation method and system. The method comprises the following steps of: obtaining content label user preference statistic data of each video watched by a target recommended user, wherein each video corresponds to a plurality of content labels; determining preference content label information of the target recommended user according to the obtained content label user preference statistic data; and selecting and pushing matched videos to the target recommended user according to the preference content label information of the target recommended user. According to the method and system, each video corresponds to a plurality of content labels, the preference content label information of the target recommended user is determined according to the content label user preference statistic data of each video watched by the target recommended user, and the behavior of watching the videos by the user can be specifically analyzed, so that videos can be accurately recommended to the user.
Description
Technical field
The present invention relates to field of computer technology, more particularly, to a kind of video recommendation method and be
System.
Background technology
With the development of video technique, have daily a large amount of as animations, film, TV play, comprehensive
The videos such as skill program produce, and simultaneously with the development of Internet technology, are that people provide more
Convenient way is watching the video of substantial amounts.When the video website such as YouTube, youku.com are per
The video segment often carving equivalent uploads, and the problem of information overload of video becomes more and more obvious,
People are led to rapidly cannot therefrom to obtain the video meeting oneself hobby.Ask to tackle this
Topic, video recommendations technology is arisen at the historic moment, and has become as current solution music field information mistake
The very potential method of load problem.
It is a kind of Information Filtering Technology on video recommendations technological essence, it is by user's history row
For the analysis of the factor such as custom, user's social relations and user's local environment, help user from
Those unnecessary information are filtered out, thus recommending to meet it for user in ever-increasing data
Hobby and the video of custom.Also had the benefit that using video recommendations technology simultaneously：(1) carry
The viewing desire of the viewer of high current web page, thus the person that is translated into video consumer；(2)
Depth excavation is carried out to system user data, more user's requests can be opened up；(3) continuous
Increase customer satisfaction degree, make user form anaclisiss to website.
Existing video recommendations technology is broadly divided into two class technical schemes, and a class technical scheme is base
In recommendation mechanisms, for example, concrete behavior is praised etc. by user's collection point, constantly recommend a certain class
Video, goes to test user's taste, if user's video that these are recommended persistently point praise or
Collection, then can reduce the weight of such video；Or to number first in video library first in system
According to modeling, then similarity between content is found by metadata it is assumed that video A and video B,
Broadly fall into the content with a type, then video A and B can be considered similar video,
Then recommend user.
Another kind of technical scheme is based on marking mechanism, that is, according to play quantity, number of reviews,
Point the data such as is praised, shares and being weighted giving a mark, and is that a certain video determines its ranking in classification,
According to classification ranking by video recommendations to user.
But above-mentioned only closed with the video that browses in the past based in the technical scheme of recommendation mechanisms
Connection is recommended, and without making a concrete analysis of the viewing behavior to video itself for the user, and leads to not essence
Push the high-quality classification video precipitating through big data accurately, and the skill based on marking mechanism
Classification to video in art scheme, the tag along sort of only single dimension, system ranking granule mistake
Slightly it is impossible to screen the situation of user's brush point, and in the case that nobody is for participation, also cannot
Accurately recommend the really higher video of video quality score to user.
Content of the invention
In view of the above problems, the embodiment of the present invention provides a kind of video recommendation method and system,
More accurately to recommend video to targeted customer.
In order to solve above-mentioned technical problem, a kind of video recommendations provided in an embodiment of the present invention
Method, it includes：
Obtain the content tab user preference system of target recommended users each video of viewing
Count, wherein each video corresponds to multiple content tabs；
Determine that this target is recommended according to the content tab user preference statistical data of described acquisition
The preferred contents label information of user；
Preferred contents label information according to this target recommended users selects the video of coupling to push away
Give target recommended users.
Wherein, described content tab user preference statistical data is the viewing of each video
The votes of each content tab to this video for the user or point praise number；
Described content tab user preference statistical data according to described acquisition determines to should
The preferred contents label information of target recommended users specifically includes：
Each content mark is determined according to the content tab user preference statistical data of described acquisition
The votes signed or point praise number, the wherein phase of multiple videos of target recommended users viewing
Merge into a unified counting with content tab；
Votes according to each content tab determining or point are praised number and are rejected invalid content mark
Sign preferred contents label composition information and each the preferred contents mark obtaining target recommended users
The preference weight information signed.
Wherein, the video of described coupling is to become with the preferred contents label of target recommended users
Divide the video with preference weight coupling；
The video of coupling is pushed to target recommended users according to following priority：
And preferred contents label composition identical with preferred contents label composition and preference weight
Identical and difference preference's weight and preferred contents label component portion difference and with preference power
Weight and preferred contents label component portion difference and difference preference's weight.
Alternatively, also include：
Quality score is carried out to video；
The video of described coupling be preferred contents label composition with target recommended users and
The video of video quality demands coupling；
The video of coupling is pushed to target recommended users according to following priority：
And video quality score identical with preferred contents label composition be high and preferred contents mark
Sign that composition is identical and video quality score is low and preferred contents label component portion difference and
Video quality score height and preferred contents label component portion difference and video quality score
Low.
Wherein, described quality score is carried out to video specifically include：
Obtain the quality score S of video playback and effective degree E of video playback；
Set up the relation of the number of times coefficient D of video playback and effective degree E of video playback,
Obtain the number of times coefficient D of video playback；
Calculate quality total score T of video playback, T=S × D according to below equation；
Video quality score G, G=T/E are calculated according to below equation.
Wherein, the described quality score S obtaining video playback specifically includes following steps：
Obtain the playing duration of video, and described playing duration is modified obtaining video
Effective playing duration L, wherein L=L1+L2 ...+Li ...+Ln, wherein, Li be i-th
Effective playing duration of secondary broadcasting；
Set up between quality score Si and effective playing duration Li of video i ＆ lt broadcasting
Relation；
According to the relation between Si and Li, it is calculated the single matter of video i ＆ lt broadcasting
Amount scoring Si；
It is calculated n effective video according to the quality score Si that video i ＆ lt is play to broadcast
Quality score S=S1+S2+ ... the Si+ ...+Sn put.
Wherein, described described playing duration is carried out described revise include accumulative process, institute
State accumulative process and specifically include following steps：
Setting playing duration correction interval (A, B)；
The actual play duration to described video for total duration Z and user of acquisition current video
U, and U is compared with Z,
As U≤Z andWhen, then the effective degree of described video playback regards E as one
Secondary, video playback is considered as playing first, and effective playing duration L1 play first is reality
Playing duration U,
As U≤Z and U ∈ (A, B), then effective degree E of described video playback is considered as one
Secondary, video playback is considered as playing first, and effective playing duration L1 play first is current
Total duration Z of video,
Work as U>During Z and U ∈ (A, B), then effective degree E of described video playback is considered as two
Secondary, video playback is considered as repeating playing, and effective playing duration L1 play first is current
Total duration Z of video, second effective playing duration L2 play be in multiple broadcasting
The playing duration once play afterwards,
Work as U>Z andWhen, then effective degree E of described video playback is considered as
[U/Z]+1, wherein, symbol [] represents round numbers, and video playback is considered as repeating playing, front [U/Z]
The effective playing duration L1～L [U/Z] of secondary broadcasting is total duration Z of current video, the
[U/Z]+1 time broadcasting effective playing duration L [U/Z)+1 be last in multiple broadcasting
The duration play.
Alternatively, when also including duplicate removal process to obtain actual play before accumulative process
Long U.
According to a further aspect in the invention, the embodiment of the present invention additionally provides a kind of video and pushes away
Recommend system, it includes：
User preference statistical data obtains processing module, for having obtained target recommended users
The content tab user preference statistical data of each video of viewing, wherein each video pair
Should multiple content tabs；
Determine processing module, for the content tab user preference statistics according to described acquisition
Data determines the preferred contents label information of this target recommended users；
Push processing module, for the preferred contents label letter according to this target recommended users
Breath selects the video push of coupling to target recommended users.
Wherein, described content tab user preference statistical data is the viewing of each video
The votes of each content tab to this video for the user or point praise number；
Described determination processing module specifically includes：
Determining module, for the content tab user preference statistical data according to described acquisition
Determine that the votes of each content tab or point praise number, wherein target recommended users are watched
The identical content Label Merging of multiple videos be one and unified count；
Preferred contents label information determining module, for according to each content tab determining
Votes or point praise number reject invalid content labels to obtain the preference of target recommended users
The preference weight information of content tab composition information and each preferred contents label.
Wherein, the video of described coupling is to become with the preferred contents label of target recommended users
Divide the video with preference weight coupling；
The described processing module that pushes is pushed to mesh to the video of coupling according to following priority
Mark recommended users：
And preferred contents label composition identical with preferred contents label composition and preference weight
Identical and difference preference's weight and preferred contents label component portion difference and with preference power
Weight and preferred contents label component portion difference and difference preference's weight.
Alternatively, also include：
Video quality score processing module, for carrying out quality score to video；
The video of described coupling be preferred contents label composition with target recommended users and
The video of video quality demands coupling；
The described processing module that pushes is pushed to mesh to the video of coupling according to following priority
Mark recommended users：
And video quality score identical with preferred contents label composition be high and preferred contents mark
Sign that composition is identical and video quality score is low and preferred contents label component portion difference and
Video quality score height and preferred contents label component portion difference and video quality score
Low.
Wherein, described video quality score processing module specifically includes：
Acquisition module, for obtaining the quality score S of video playback and having of video playback
Effect number of times E；
Relation sets up module, for setting up number of times coefficient D and the video playback of video playback
Effective degree E relation；
Computing module, for being calculated D according to the relation of D and E, according to T=S
× D is calculated quality total score T of video playback and obtains video quality according to G=T/E
Scoring G.
Alternatively, described video quality score processing module also includes correcting module；
Described acquisition module is additionally operable to obtain the playing duration of video；
Described correcting module, has for being modified to described playing duration obtaining video
Effect playing duration L, wherein L=L1+L2 ...+Li ...+Ln, wherein, Li broadcasts for i ＆ lt
The effective playing duration put；
Described relation is set up module and is additionally operable to set up the single quality of video i ＆ lt broadcasting and comments
Divide the relation between Si and effective playing duration Li；
Described computing module be additionally operable to according to Si and Li between relation, be calculated video
The single quality score Si that i ＆ lt is play, and the single quality play according to video i ＆ lt
Scoring Si is calculated the quality score of n effective video broadcasting
S=S1+S2+ ... Si+ ...+Sn, wherein, Si is that the quality of i ＆ lt video single play is commented
Point.
Alternatively, described video quality score processing module also includes setup module and judgement
Module；
Described setup module, for arranging playing duration correction interval (A, B)；
Described acquisition module is additionally operable to obtain total duration Z of current video and user to described
The actual play duration U of video；
Described judge module, for being compared U with Z, and is entered according to comparative result
Row judges,
As U≤Z andWhen, then the effective degree of described video playback regards E as one
Secondary, video playback is considered as playing first, and effective playing duration L1 play first is reality
Playing duration U,
As U≤Z and U ∈ (A, B), then effective degree E of described video playback is considered as one
Secondary, video playback is considered as playing first, and effective playing duration L1 play first is current
Total duration Z of video,
Work as U>During Z and U ∈ (A, B), then effective degree E of described video playback is considered as two
Secondary, video playback is considered as repeating playing, and effective playing duration L1 play first is current
Total duration Z of video, second effective playing duration L2 play be in multiple broadcasting
The playing duration once play afterwards,
Work as U>Z andWhen, then effective degree E of described video playback is considered as
[U/Z]+1, wherein symbol [] represent round numbers, and video playback is considered as repeating playing, front [U/Z]
The effective playing duration L1～L [U/Z] of secondary broadcasting is total duration Z of current video, the
[U/Z]+1 time broadcasting effective playing duration L [U/Z)+1 be last in multiple broadcasting
The duration play.
Alternatively, described video quality score processing module also includes deduplication module, is used for
Processed by duplicate removal and obtain actual play duration U.
The video recommendation method providing according to embodiments of the present invention and system, it passes through to obtain
The content tab user preference statistical data of each video of target recommended users viewing,
Wherein each video corresponds to multiple content tabs；Content tab user according to described acquisition
Preference statistical data determines the preferred contents label information of this target recommended users；According to this
The preferred contents label information of target recommended users from select coupling video push to target
Recommended users, because each video is to should have multiple content tabs, and recommend according to target
The content tab user preference statistical data of each video of user's viewing determines that target pushes away
Recommend the preferred contents label information of user, that is, make a concrete analysis of the viewing row to video itself for the user
For therefore, accurately video can be recommended to user.
Brief description
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art,
The accompanying drawing of required use in embodiment or description of the prior art simply will be situated between below
Continue it should be apparent that, drawings in the following description be only the present invention described in some
Embodiment, for those of ordinary skill in the art, can also obtain according to these accompanying drawings
Obtain other accompanying drawings.
Fig. 1 is the specific embodiment flow chart according to video recommendation method of the present invention；
Fig. 2 is the specific embodiment overall schematic according to video recommendation system of the present invention；
Fig. 3 is concrete according to determine processing module in video recommendation system of the present invention one
Embodiment composition schematic diagram；
Fig. 4 is another specific embodiment integrally group according to video recommendation system of the present invention
Become schematic diagram.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, to the skill in the embodiment of the present invention
Art scheme is clearly and completely described it is clear that described embodiment is only this
Invent a part of embodiment, rather than whole embodiments.Based on the enforcement in the present invention
Example, the every other embodiment that those of ordinary skill in the art are obtained, broadly fall into this
The scope of bright protection.
Refer to Fig. 1, it is the specific embodiment flow process according to video recommendation method of the present invention
Figure.
As shown, in the present embodiment, video recommendation method mainly comprises the steps：
Step S101, obtains the content tab of target recommended users each video of viewing
User preference statistical data, wherein each video correspond to multiple content tabs；
When implementing, in the present embodiment each video can set up 5-10 label or its
The content tab of his quantity, is not specifically limited here, and can support that User Defined is mended
Fill, that is, user can submit self-defining content tab the content tab as this video to,
In addition, the statistical data of content tab user preference described in the present embodiment can be for example every
Individual video watch each content tab to this video for the user votes or point praise number,
Votes or point praise the order of accuarcy that number can be used for determining this content tab, when certain video
Content tab votes when changing, all users playing this video are to should content
The content tab user preference statistical data of label can be carried out when next measurement period starts
Corresponding variation, is not specifically limited here；
Step S102, the content tab user preference statistical data according to described acquisition determines
The preferred contents label information of this target recommended users；
When implementing, for example, can be counted according to the content tab user preference of described acquisition
Data determine each content tab votes or point praise number, then according to determine each
The votes of content tab or point are praised several invalid content labels of rejecting into and are obtained target recommendation use
The preferred contents label composition information at family and the preference weight information of each preferred contents label,
For example, votes or composition weight are less than preset value (the such as composition of single internal standard label
Weight percentage is less than 5%) content tab can be considered invalid tag, reject not as inclined
Good content tab composition, calculates without in preference weight.
Need explanation, in the present embodiment, synonymous label is considered as with a label, backstage
After configuration, all form and votes etc. of representing of synonymous label all merge into a process,
And the identical content label for multiple videos of target recommended users viewing is also merged into
One unified to count, in addition, repeat the video watched for user, this video each
The votes of content or point praise number need to be multiplied by broadcasting time or one of broadcasting time pre-
Fixed coefficient, is not specifically limited here；And the composition weight of each content tab is for example
Formula can be calculated as follows calculated, that is,：
Single label composition weight percentage=(the ∑ list label votes/total votes of ∑) * 100%
And preference weight can be calculated according to the following equation, that is,：
Single label preference weight percentage ratio=(the ∑ list label counter poll/total ticket of ∑ enumerator
Number) * 100%
It is exemplified below.
Assume in the present embodiment that video M has 4 content tabs, i.e. content tab a,
Votes are 10 tickets, content tab b, and votes are 20 tickets, content tab c, ballot
Number is 30 tickets, content tab d, and votes are 40 tickets, and video N has in 2
Hold label, i.e. content tab a, votes are 1 ticket, content tab e, votes are 20
Ticket, in addition the preset value of invalid content label is 5%；
Embodiment 1,
Certain target recommended users completely plays 1 video M, the composition label power of video M
Percentage ratio is again：
Content tab a：10/ (10+20+30+40) %=10%, content tab b：20/(10+
20+30+40) %=20%, content tab c：30/ (10+20+30+40) %=30%, content
Label d：40/ (10+20+30+40) %=40%, due to the composition weight of each content tab
Both greater than preset value (being more than 5%), that is, do not have invalid content label, therefore, preference
Weight percentage is identical numerical value with the composition weight percentage of video M, that is, in preference
The preference weight percentage ratio holding label a is 10%, the preference weight of preferred contents label b
Percentage ratio is 20%, and the preference weight percentage ratio of preferred contents label c is 30%, preference
The preference weight percentage ratio of content tab d is 40%.
Embodiment 2
Certain target recommended users completely plays 1 video N, video N composition label weight
Percentage is：
Content tab a：1/ (1+20)=4.76%, that is, be less than preset value, therefore this content mark
Sign as invalid content label
Content tab e：20/20=100%, i.e. user preference content tab content tab e,
The preference weight percentage ratio of this content tab e is 100%.
Embodiment 3
Certain target recommended users is complete to play 2 video M and 1 video N, due to regarding
In frequency N, content tab a is invalid tag, therefore, rejects this invalid tag, video is inclined
Good weight calculation is as follows, that is,：
The preference weight of preferred contents label a is：(10+10)/(10+10+20+20+30+30
+ 40+40+20)=9.09%, the preference weight of preferred contents label b is：(20+20)/(10+
10+20+20+30+30+40+40+20)=18.18%, the preference weight of preferred contents label c
For：(30+30)/(10+10+20+20+30+30+40+40+20)=27.27%, preferred contents
The preference weight of label d is：(40+40)/(10+10+20+20+30+30+40+40+20)=
36.36%, the preference weight of preferred contents label e is：20/(10+10+20+20+30+30
+ 40+40+20)=9.09%.
Step S103, according to the preferred contents label information selection of this target recommended users
The video push joined is to target recommended users.
When implementing, the video play by user will not push to this user,
And can recommend to use to target using the video push that various ways select coupling when specifically pushing
Family, for example, as a specific embodiment, the video of described coupling is to recommend with target
The preferred contents label composition of user and the video of preference weight coupling, and the video mating
Target recommended users can be pushed to according to following priority：
Identical with preferred contents label composition and preference weight become split-phase with preferred contents label
With and difference preference's weight and preferred contents label component portion difference and same preference weight,
With preferred contents label component portion difference and difference preference's weight.
Need explanation, in the present embodiment, preference weight identical inclusion preference weight is identical
Situation, also include weight difference range situation within limits, such as one
Preference weight percentage ratio be 30%, can by with preference weight about 30% 5% interval model
Enclose (i.e. between 25%-35%) be also considered as identical with preference weight, in practice can basis
Concrete condition sets and preference weight identical critical field, is not specifically limited here.
The video recommendation method being provided according to this above-described embodiment, it passes through to obtain target recommendation
The content tab user preference statistical data of each video of user's viewing, wherein each
Video corresponds to multiple content tabs；Content tab user preference statistics according to described acquisition
Data determines the preferred contents label information of this target recommended users；Recommended according to this target
The preferred contents label information of user from select coupling video push to target recommended users,
Because each video is to should have multiple content tabs, and watched according to target recommended users
The content tab user preference statistical data of each video determine the inclined of target recommended users
Good content tab information, that is, make a concrete analysis of the viewing behavior to video itself for the user, therefore,
Accurately video can be recommended to user.
In addition it is recommended that being also required to when video is to target recommended users consider video quality, that is,
Video can be carried out in the present embodiment with quality score, and the video mating is to recommend with target
The preferred contents label composition of user and the video of video quality demands coupling, coupling
Video can be pushed to target recommended users according to following priority：
And video quality score identical with preferred contents label composition be high and preferred contents label
Composition is identical and video quality score low with preferred contents label component portion difference and regard
With preferred contents label component portion difference and video quality score is low for frequency quality score height.
Need explanation, the high score of the video quality of the present embodiment and low point can be according to reality
Situation sets corresponding standard it is possible to dynamic adjust, and is not specifically limited here.
It is illustrated below：
If the preferred contents label composition of target recommended users X be preferred contents label A,
Preferred contents label B and preferred contents label C, wherein preferred contents label A is inclined
Good weight percentage is 33%, and the preference weight percentage ratio of preferred contents label B is 33%,
The preference weight percentage ratio of preferred contents label C is 34%；
And the video of coupling is as follows in video library：
The quality score of video 1 is 100 points, in addition, the content tab A of this video 1
Composition weight percentage be 33%, the composition weight percentage of content tab B is 28%,
The composition weight percentage of content tab C is 39%, and video 1 and preferred contents label
Composition and preference weight are identical, and in addition video 1 is identical with preferred contents label composition and regards
Frequency quality score is high；
The quality score of video 2 is 40 points, the in addition one-tenth of the content tab A of this video 2
Weight percentage is divided to be 50%, the composition weight percentage of content tab B is 30%, interior
The composition weight percentage holding label D is 20%, and video 2 is become with preferred contents label
Divide part variation and same preference weight；In addition, video 2 becomes branch with preferred contents label
Divide difference and video quality score is low；
The quality score of video 3 is 80 points, in addition, the content tab A of this video 3
Composition weight percentage is 90%, and the composition weight percentage of content tab C is 10%,
And video 3 and preferred contents label component portion difference and difference preference's weight；In addition,
With preferred contents label component portion difference and video quality score is high for video 3；
The quality score of video 4 is 30 points, in addition, the one-tenth of the content tab A of this video
Weight percentage is divided to be 5%, the composition weight percentage of content tab B is 5%, content
The composition weight percentage of label C is 90%, and video 4 and preferred contents label composition
Identical and difference preference's weight；In addition, video 4 identical with preferred contents label composition and
Video quality score is low；
During concrete push, the video of coupling can be pushed to target according to following priority to be recommended
User X, that is,：
Identical with preferred contents label composition and preference weight become split-phase with preferred contents label
With and difference preference's weight and preferred contents label component portion difference and same preference weight,
With preferred contents label component portion difference and difference preference's weight, that is,：
Video 1 has precedence over video 4, and video 4 has precedence over video 2, and video 2 has precedence over and regards
Frequently 3.
In addition, the video of coupling can be pushed to target recommended users according to another kind of priority
X, that is,：
And video quality score identical with preferred contents label composition be high and preferred contents label
Composition is identical and video quality score low with preferred contents label component portion difference and regard
With preferred contents label component portion difference and video quality score is low for frequency quality score height,
I.e.：
Video 1 has precedence over video 4, and video 4 has precedence over video 3, and video 3 has precedence over and regards
Frequently 2.
Need explanation, can also carry out otherwise for pushing priority, example
As, the size order according to the content tab of user's preference enters row major and pushes, or
Fault-tolerant interval domain of walker of adjustment coupling etc. all can be adjusted to pushing video, here
It is not specifically limited.
In addition, can adopt in various manners for video is carried out with quality score in the present embodiment,
In order that video quality score more conforms to user's specification, as a preferred embodiment,
The present embodiment carries out quality score and can adopt following manner, that is, to video：
S1, the acquisition quality score S of video playback and effective degree E of video playback；
Wherein, video single play refers to once broadcasting in effective broadcasting time for the video, obtains
The quality score S taking video playback specifically includes following steps：
S11, the playing duration of acquisition video, and described playing duration is modified being regarded
Effective playing duration L of frequency；
Described correction includes adding up to process processing with duplicate removal, and described accumulative process specifically includes following
Step：
S111, setting playing duration correction interval (A, B), wherein A and B value on demand；
S112, when obtaining total duration Z of current video and user to the actual play of described video
Long U,
As U≤Z andWhen, then the effective degree of described video playback regards E as once,
Video playback is considered as playing first, and effective playing duration is considered as actual play duration U,
As U≤Z and U ∈ (A, B), then effective degree E of described video playback is considered as once,
Video playback is considered as playing first, and effective playing duration is considered as total duration Z of current video,
Work as U>During Z and U ∈ (A, B), then effective degree E of described video playback is considered as twice,
Video playback is considered as repeating playing, the effective playing duration play first be current video total when
Long Z, effective playing duration that second is play is the playing duration of second broadcasting,
Work as U>Z andWhen, then effective degree E of described video playback is considered as U/Z+1,
Wherein, U/Z round numbers, video playback is considered as repeating playing, effective broadcasting of U/Z broadcasting
Total duration Z of Shi Changwei current video, effective playing duration of the U/Z+1 time broadcasting is multiple
The last duration play in broadcasting.
Described duplicate removal processes and specifically includes following steps：
S113, acquisition user play time interval I repeatedly play during certain video first, and set
The correction of putting time interval interval (A ', B ')；
S114, accumulative playing duration is compared with total duration Z of described video for W,
If W>Z, and I ∈ (A ', B ') when, then described video is carried out duplicate removal process, if repeatedly
A length of R during the video segment repeating in broadcasting, then actual play duration U is W-R；
After playing first, if I ∈ (A ', B ') when, then described video is carried out duplicate removal process,
If a length of R during the video segment repeating in the multiple broadcasting after playing first, actual play
Duration U is W-R；
After playing first, ifWhen, then do not carry out duplicate removal process；
S12, set up the quality score S of video single playiWith effective playing duration LiBetween
Relation, described SiWith LiBetween relation be：
When effective playing duration L of videoi<During playing duration preset value P, calculate public affairs using following
Formula：
Si=k1Li
When effective playing duration L of videoiDuring >=playing duration preset value P, calculate public affairs using following
Formula：
Wherein, p0It is to work as LiVideo quality score during=P, k1It is constant with a；
S13, according to SiWith LiBetween relation, the quality being calculated video single play comments
Divide Si；
S14, the quality score S according to video single playiIt is calculated n effective video to broadcast
The quality score S=S put1+S2+…Si+…+Sn, wherein, SiFor i ＆ lt video single play
Quality score.
S2, set up the number of times coefficient D of video playback and the pass of effective degree E of video playback
System, obtains the number of times coefficient D of video playback, the relation between described D and E is：
As the effective degree E≤effective degree preset value Q of video playback, using following calculating
Formula：
D=bE+d1-1
Effective degree E when video playback>During effective degree preset value Q, using following calculating
Formula：
Wherein, d1It is number of times coefficient when playing first for video, d2It is secondary as E=Q
Number system number, b, d1And d2For constant.
S3, quality total score T according to below equation calculating video playback, T=S × D；
S4, according to below equation calculate video quality score G, G=T/E.
It is illustrated below：
Obtain the instantiation of effective playing duration of video：
Setting playing duration correction interval (A, B), A=(1-20%) * Z, B=(1+20%) * Z,
Assume total duration Z=600s of current video, then revising interval is (480-720s).
The video that certain user is 600s to video total duration plays out, and plays for the first time the
The video segment of 0-60s, second plays the video segment of 30-90s, then for the first time and the
The playing duration (adding up playing duration) of secondary accumulative broadcasting is 120s, due to 120s<600s
AndThen effective degree E of described video playback is considered as once, and this regards
Frequency is play and is considered as playing first, and effective playing duration L is considered as accumulative playing duration 120s.
Calculate the instantiation of the quality score of video playback：
Hypothesis playing duration preset value is 500s, then effective playing duration 120s of this video<Play
Duration preset value 500s, using formula S=k1L calculates the quality score S of video single play,
S=k1* 120, because video is effectively play 1 time, then S=S=k1* 120, wherein, k1Be to
Permanent number.
Calculate the instantiation of the number of times coefficient of video playback：
Assume that effective degree preset value is 2 times, then the effective degree 1 of this video playback<Effectively
Number of times preset value 2, using formula D=bE+d1The number of times coefficient of -1 calculating video playback,
D=b+d1- 1, wherein, b, d1For giving constant.
Calculate the instantiation of the quality total score of video playback：
Quality total score T=S*D of this video playback, therefore, T=k1*120*(b+d1-1).
Calculate the instantiation of video quality score：
Video quality score G=T/E, due to E=1, then G=T=k1*120*(b+d1-1).
According to the quality score method of the present embodiment, can effectively filter the impact of user's brush point,
Algorithm amount of calculation is less, can mitigate the burden of server, and before playing duration preset value,
Video quality score increases more slowly, after playing duration preset value, plays all within each second
More more valuable than before, increasing degree is larger, and this kind of mode can effectively filter out user's brush point
Impact, in addition, according to the quality score method of the present embodiment, effective degree preset value it
Before, play each time all than front once play more valuable, after effective degree preset value,
Play each time and be not all once more worth than front, the video quality score obtaining more conforms to use
Family standard.
Need explanation, above-mentioned is only a kind of mode that video quality is scored, this
Can not do concrete here in the way of other score to video quality in embodiment
Limit.
Another aspect of the present invention is described below.
With reference to Fig. 2, this figure is overall according to the specific embodiment of video recommendation system of the present invention
Schematic diagram.
As shown, the video recommendation system of the present embodiment mainly includes：
User preference statistical data obtains processing module 1, for having obtained target recommended users
The content tab user preference statistical data of each video of viewing, wherein each video pair
Should multiple content tabs；
Determine processing module 2, for the content tab user preference statistics according to described acquisition
Data determines the preferred contents label information of this target recommended users；
It is to be appreciated that the statistical data of content tab user preference described in the present embodiment is every
Individual video watch each content tab to this video for the user votes or point praise number,
With reference to Fig. 3, as a preferred embodiment, described determination processing module 2 is specifically wrapped
Include：
Determining module 21, for the content tab user preference statistical number according to described acquisition
Praise number according to the votes determining each content tab or point, wherein target recommended users are seen
The identical content Label Merging of the multiple videos seen is a unified counting；
Preferred contents label information determining module 22, for according to each content mark determining
The votes signed or point are praised number and are rejected invalid content label to obtain the inclined of target recommended users
Good content tab composition information and the preference weight information of each preferred contents label.
Push processing module 3, for the preferred contents label letter according to this target recommended users
Breath selects the video push of coupling to target recommended users from video library.
When implementing, can be pushed away to target using the video push that various ways select coupling
Recommend user, for example, as a specific embodiment, the video of described coupling is and target
The preferred contents label composition of recommended users and the video of preference weight coupling, described push
Processing module is pushed to target recommended users to the video of coupling according to following priority：
And preferred contents label composition identical with preferred contents label composition and preference weight
Identical and difference preference's weight and preferred contents label component portion difference and with preference power
Weight and preferred contents label component portion difference and difference preference's weight.
The video recommendation system being provided according to the present embodiment, it passes through to obtain target recommendation use
The content tab user preference statistical data of each video of family viewing, wherein each regards
The corresponding multiple content tabs of frequency；Content tab user preference statistical number according to described acquisition
According to the preferred contents label information determining this target recommended users；Recommend to use according to this target
The preferred contents label information at family from select coupling video push to target recommended users,
Because each video is to should have multiple content tabs, and watched according to target recommended users
The content tab user preference statistical data of each video determine the inclined of target recommended users
Good content tab information, that is, make a concrete analysis of the viewing behavior to video itself for the user, therefore,
Accurately video can be recommended to user.
Need explanation it is recommended that be also required to when video is to target recommended users consider video matter
Amount, with reference to Fig. 4, this figure is to be embodied as according to another of video recommendation system of the present invention
Example overall schematic.
As shown, unlike the embodiments above be the present embodiment video recommendation system in
Also include：Video quality score processing module 4, for carrying out quality score to video；And
The video of described coupling is the preferred contents label composition and video with target recommended users
The video of prescription coupling；
The described processing module 3 that pushes can push according to following priority to the video of coupling
To target recommended users：
And video quality score identical with preferred contents label composition be high and preferred contents mark
Sign that composition is identical and video quality score is low and preferred contents label component portion difference and
Video quality score height and preferred contents label component portion difference and video quality score
Low.
In the present embodiment for video is carried out quality score can adopt in various manners, in order to
Video quality score is made to more conform to user's specification, as a preferred embodiment, described
Video quality score processing module 4 may include：
Acquisition module, for obtaining the quality score S of video playback and having of video playback
Effect number of times E；
Relation sets up module, for setting up number of times coefficient D and the video playback of video playback
Effective degree E relation；
Computing module, for being calculated D according to the relation of D and E, according to T=S
× D is calculated quality total score T of video playback and obtains video quality according to G=T/E
Scoring G.
In addition, described acquisition module can be additionally used in obtaining the playing duration of video, described regard
Frequency quality score processing module 4 may also include correcting module；Described correcting module is mainly used
In described playing duration is modified obtaining with effective playing duration L of video, wherein
L=L1+L2 ...+Li ...+Ln, wherein, effective playing duration that Li plays for i ＆ lt；
Described relation is set up module and is additionally operable to set up the single quality of video i ＆ lt broadcasting and comments
Divide the relation between Si and effective playing duration Li；
Described computing module be additionally operable to according to Si and Li between relation, be calculated video
The single quality score Si that i ＆ lt is play, and the single quality play according to video i ＆ lt
Scoring Si is calculated the quality score of n effective video broadcasting
S=S1+S2+ ... Si+ ...+Sn, wherein, Si is that the quality of i ＆ lt video single play is commented
Point.
In addition, as an alternatively embodiment, described video quality score processing module 4
May also include setup module and judge module；Wherein said setup module, broadcasts for setting
Put time scale modification interval (A, B)；And described acquisition module is additionally operable to obtain current video
The total duration Z and user actual play duration U to described video；
Described judge module, for being compared U with Z, and is entered according to comparative result
Row judges,
As U≤Z andWhen, then the effective degree of described video playback regards E as one
Secondary, video playback is considered as playing first, and effective playing duration L1 play first is reality
Playing duration U,
As U≤Z and U ∈ (A, B), then effective degree E of described video playback is considered as one
Secondary, video playback is considered as playing first, and effective playing duration L1 play first is current
Total duration Z of video,
Work as U>During Z and U ∈ (A, B), then effective degree E of described video playback is considered as two
Secondary, video playback is considered as repeating playing, and effective playing duration L1 play first is current
Total duration Z of video, second effective playing duration L2 play be in multiple broadcasting
The playing duration once play afterwards,
Work as U>Z andWhen, then effective degree E of described video playback is considered as
[U/Z]+1, wherein symbol [] represent round numbers, and video playback is considered as repeating playing, front [U/Z]
The effective playing duration L1～L [U/Z] of secondary broadcasting is total duration Z of current video, the
[U/Z]+1 time broadcasting effective playing duration L [U/Z)+1 be last in multiple broadcasting
The duration play.
In addition, as an optional embodiment, described video quality score processing module 4
May also include deduplication module, obtain actual play duration U for processing by duplicate removal, this
In repeat no more.
According to the video recommendation system of the present embodiment, when it is to video quality score, can be effective
Filter the impact of user's brush point, algorithm amount of calculation is less, can mitigate the burden of server, and
Before playing duration preset value, video quality score increases more slowly, pre- in playing duration
If after value, play within each second all more more valuable than before, increasing degree is larger, this kind of mode
The impact of user's brush point can effectively be filtered out, in addition, the quality score side according to the present embodiment
Method, before effective degree preset value, play each time all than front once play more valuable,
After effective degree preset value, play each time and be not all once more worth than front, obtain
Video quality score more conforms to user's specification.
In above-mentioned provided description, illustrate a large amount of details.However, energy
Enough understanding, embodiments of the invention can be put into practice in the case of not having these details.
In some instances, known method, structure and technology are not been shown in detail, so as not to
Obscure the understanding of this description.
Similarly it will be appreciated that in order to simplify the disclosure and help understand each invented party
One or more of face, in the description to the exemplary embodiment of the present invention above,
Each feature of the present invention is grouped together into single embodiment, figure or to it sometimes
Description in.However, the method for the disclosure should be construed to reflect following intention：
I.e. the present invention for required protection requires than the spy being expressly recited in each claim
Levy more features.More precisely, as the following claims reflect,
Inventive aspect is all features less than single embodiment disclosed above.Therefore, abide by
The claims following specific embodiment are thus expressly incorporated in this specific embodiment,
Wherein each claim itself is as the separate embodiments of the present invention.
It should be noted that above-described embodiment the present invention will be described rather than to the present invention
Limited, and those skilled in the art are without departing from scope of the following claims
In the case of can design alternative embodiment.
Claims (10)
1. a kind of video recommendation method is it is characterised in that include：
Obtain the content tab user preference statistical data of target recommended users each video of viewing, wherein each video corresponds to multiple content tabs；
The preferred contents label information of this target recommended users is determined according to the content tab user preference statistical data of described acquisition；
Preferred contents label information according to this target recommended users selects the video push of coupling to target recommended users.
2. video recommendation method according to claim 1 it is characterised in that described content tab user preference statistical data be each video watch each content tab to this video for the user votes or point praise number.
3. video recommendation method according to claim 2 is it is characterised in that described content tab user preference statistical data according to described acquisition determines to should the preferred contents label information of target recommended users specifically include：
Determine that the votes of each content tab or point praise number according to the content tab user preference statistical data of described acquisition, the wherein identical content Label Merging of multiple videos of target recommended users viewing is a unified counting；
Votes according to each content tab determining or point are praised number rejecting invalid content label and are obtained the preferred contents label composition information of target recommended users and the preference weight information of each preferred contents label.
4. video recommendation method according to claim 3 is it is characterised in that the video of described coupling is the video mating with preferred contents label composition and the preference weight of target recommended users；
The video of coupling is pushed to target recommended users according to following priority：
Identical with preferred contents label composition and preference weight identical with preferred contents label composition and difference preference's weight and preferred contents label component portion difference and same preference weight and preferred contents label component portion difference and difference preference's weight.
5. video recommendation method according to claim 3 is it is characterised in that also include：
Quality score is carried out to video；
The video of described coupling is the video mating with preferred contents label composition and the video quality demands of target recommended users；
The video of coupling is pushed to target recommended users according to following priority：
And video quality score identical with preferred contents label composition be high identical with preferred contents label composition and video quality score low with preferred contents label component portion difference and video quality score high with preferred contents label component portion difference and video quality score is low.
6. a kind of video recommendation system is it is characterised in that include：
User preference statistical data obtains processing module, and for obtaining the content tab user preference statistical data of each video of target recommended users viewing, wherein each video corresponds to multiple content tabs；
Determine processing module, for determining the preferred contents label information of this target recommended users according to the content tab user preference statistical data of described acquisition；
Push processing module, select the video push of coupling to target recommended users for the preferred contents label information according to this target recommended users.
7. video recommendation system according to claim 6 it is characterised in that described content tab user preference statistical data be each video watch each content tab to this video for the user votes or point praise number.
8. video recommendation system according to claim 7 is it is characterised in that described determination processing module specifically includes：
Determining module, the votes or point for determining each content tab according to the content tab user preference statistical data of described acquisition praise number, and the wherein identical content Label Merging of multiple videos of target recommended users viewing is a unified counting；
Preferred contents label information determining module, praises number for the votes according to each content tab determining or point and rejects invalid content label to obtain the preferred contents label composition information of target recommended users and the preference weight information of each preferred contents label.
9. video recommendation system according to claim 8 is it is characterised in that the video of described coupling is the video mating with preferred contents label composition and the preference weight of target recommended users；
The described processing module that pushes is pushed to target recommended users to the video of coupling according to following priority：
Identical with preferred contents label composition and preference weight identical with preferred contents label composition and difference preference's weight and preferred contents label component portion difference and same preference weight and preferred contents label component portion difference and difference preference's weight.
10. video recommendation system according to claim 8 is it is characterised in that also include：
Video quality score processing module, for carrying out quality score to video；
The video of described coupling is the video mating with preferred contents label composition and the video quality demands of target recommended users；
The described processing module that pushes is pushed to target recommended users to the video of coupling according to following priority：
And video quality score identical with preferred contents label composition be high identical with preferred contents label composition and video quality score low with preferred contents label component portion difference and video quality score high with preferred contents label component portion difference and video quality score is low.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610162038.8A CN106407241A (en) | 2016-03-21 | 2016-03-21 | Video recommendation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610162038.8A CN106407241A (en) | 2016-03-21 | 2016-03-21 | Video recommendation method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106407241A true CN106407241A (en) | 2017-02-15 |
Family
ID=58007190
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610162038.8A Pending CN106407241A (en) | 2016-03-21 | 2016-03-21 | Video recommendation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106407241A (en) |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107786898A (en) * | 2017-09-28 | 2018-03-09 | 南京林洋电力科技有限公司 | A kind of electric power wisdom business hall media intelligent player method with compensation mechanism |
| CN107948263A (en) * | 2017-11-15 | 2018-04-20 | 福建中金在线信息科技有限公司 | The information-pushing method and its relevant device of a kind of finance product |
| CN108322827A (en) * | 2018-01-31 | 2018-07-24 | 武汉斗鱼网络科技有限公司 | Method, system and the computer readable storage medium of the video preference of measure user |
| CN108376166A (en) * | 2018-02-24 | 2018-08-07 | 宁夏玲杰科技有限公司 | Intelligence learning method, apparatus and electronic equipment |
| CN109657138A (en) * | 2018-12-10 | 2019-04-19 | 深圳墨世科技有限公司 | A kind of video recommendation method, device, electronic equipment and storage medium |
| CN109729395A (en) * | 2018-12-14 | 2019-05-07 | 广州市百果园信息技术有限公司 | Video quality evaluation method, device, storage medium and computer equipment |
| CN109726329A (en) * | 2018-12-29 | 2019-05-07 | 百度在线网络技术（北京）有限公司 | Content recommendation method, device and equipment |
| CN109831678A (en) * | 2019-02-26 | 2019-05-31 | 中国联合网络通信集团有限公司 | Short method for processing video frequency and system |
| CN110059221A (en) * | 2019-03-11 | 2019-07-26 | 咪咕视讯科技有限公司 | Video recommendation method, electronic equipment and computer readable storage medium |
| CN111353068A (en) * | 2020-02-28 | 2020-06-30 | 腾讯音乐娱乐科技（深圳）有限公司 | Video recommendation method and device |
| CN111490929A (en) * | 2020-03-27 | 2020-08-04 | 威比网络科技(上海)有限公司 | Video clip pushing method and device, electronic equipment and storage medium |
| CN111522993A (en) * | 2020-04-08 | 2020-08-11 | 咪咕文化科技有限公司 | Label validity management method and device, network equipment and storage medium |
| CN111726649A (en) * | 2020-06-28 | 2020-09-29 | 百度在线网络技术（北京）有限公司 | Video stream processing method, device, computer equipment and medium |
| CN111723237A (en) * | 2020-06-12 | 2020-09-29 | 腾讯科技（深圳）有限公司 | Media content access control method |
| CN111722766A (en) * | 2020-06-04 | 2020-09-29 | 北京达佳互联信息技术有限公司 | Multimedia resource display method and device |
| CN112653907A (en) * | 2020-12-15 | 2021-04-13 | 泰康保险集团股份有限公司 | Video recommendation method and device |
| CN112800275A (en) * | 2021-01-26 | 2021-05-14 | 广州欢网科技有限责任公司 | Short video feed stream real-time calculation pushing method, device and equipment |
| CN113312511A (en) * | 2021-06-11 | 2021-08-27 | 北京百度网讯科技有限公司 | Method, apparatus, device and computer-readable storage medium for recommending content |
| CN113449147A (en) * | 2021-07-06 | 2021-09-28 | 乐视云计算有限公司 | Video recommendation method and device based on theme |
| CN114513687A (en) * | 2022-02-14 | 2022-05-17 | 青岛聚看云科技有限公司 | Server and media asset recommendation method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101719167A (en) * | 2010-01-15 | 2010-06-02 | 北京暴风网际科技有限公司 | Interactive movie searching method |
| CN102073677A (en) * | 2010-12-01 | 2011-05-25 | 北京开心人信息技术有限公司 | Comment method and system based on tag |
| CN103136275A (en) * | 2011-12-02 | 2013-06-05 | 盛乐信息技术（上海）有限公司 | System and method for recommending personalized video |
| CN103324758A (en) * | 2013-07-10 | 2013-09-25 | 苏州大学 | News classifying method and system |
| CN104156472A (en) * | 2014-08-25 | 2014-11-19 | 四达时代通讯网络技术有限公司 | Video recommendation method and system |
-
2016
- 2016-03-21 CN CN201610162038.8A patent/CN106407241A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101719167A (en) * | 2010-01-15 | 2010-06-02 | 北京暴风网际科技有限公司 | Interactive movie searching method |
| CN102073677A (en) * | 2010-12-01 | 2011-05-25 | 北京开心人信息技术有限公司 | Comment method and system based on tag |
| CN103136275A (en) * | 2011-12-02 | 2013-06-05 | 盛乐信息技术（上海）有限公司 | System and method for recommending personalized video |
| CN103324758A (en) * | 2013-07-10 | 2013-09-25 | 苏州大学 | News classifying method and system |
| CN104156472A (en) * | 2014-08-25 | 2014-11-19 | 四达时代通讯网络技术有限公司 | Video recommendation method and system |
Cited By (31)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107786898A (en) * | 2017-09-28 | 2018-03-09 | 南京林洋电力科技有限公司 | A kind of electric power wisdom business hall media intelligent player method with compensation mechanism |
| CN107948263A (en) * | 2017-11-15 | 2018-04-20 | 福建中金在线信息科技有限公司 | The information-pushing method and its relevant device of a kind of finance product |
| CN108322827A (en) * | 2018-01-31 | 2018-07-24 | 武汉斗鱼网络科技有限公司 | Method, system and the computer readable storage medium of the video preference of measure user |
| CN108322827B (en) * | 2018-01-31 | 2021-02-02 | 武汉斗鱼网络科技有限公司 | Method, system, and computer-readable storage medium for measuring video preferences of a user |
| CN108376166A (en) * | 2018-02-24 | 2018-08-07 | 宁夏玲杰科技有限公司 | Intelligence learning method, apparatus and electronic equipment |
| CN109657138A (en) * | 2018-12-10 | 2019-04-19 | 深圳墨世科技有限公司 | A kind of video recommendation method, device, electronic equipment and storage medium |
| CN109657138B (en) * | 2018-12-10 | 2021-02-26 | 深圳大宇无限科技有限公司 | Video recommendation method and device, electronic equipment and storage medium |
| CN109729395A (en) * | 2018-12-14 | 2019-05-07 | 广州市百果园信息技术有限公司 | Video quality evaluation method, device, storage medium and computer equipment |
| CN109729395B (en) * | 2018-12-14 | 2022-02-08 | 广州市百果园信息技术有限公司 | Video quality evaluation method and device, storage medium and computer equipment |
| CN109726329A (en) * | 2018-12-29 | 2019-05-07 | 百度在线网络技术（北京）有限公司 | Content recommendation method, device and equipment |
| CN109831678A (en) * | 2019-02-26 | 2019-05-31 | 中国联合网络通信集团有限公司 | Short method for processing video frequency and system |
| CN110059221A (en) * | 2019-03-11 | 2019-07-26 | 咪咕视讯科技有限公司 | Video recommendation method, electronic equipment and computer readable storage medium |
| CN110059221B (en) * | 2019-03-11 | 2023-10-20 | 咪咕视讯科技有限公司 | Video recommendation method, electronic device and computer readable storage medium |
| CN111353068A (en) * | 2020-02-28 | 2020-06-30 | 腾讯音乐娱乐科技（深圳）有限公司 | Video recommendation method and device |
| CN111490929A (en) * | 2020-03-27 | 2020-08-04 | 威比网络科技(上海)有限公司 | Video clip pushing method and device, electronic equipment and storage medium |
| CN111490929B (en) * | 2020-03-27 | 2022-07-15 | 深圳市企鹅网络科技有限公司 | Video clip pushing method and device, electronic equipment and storage medium |
| CN111522993B (en) * | 2020-04-08 | 2023-08-15 | 咪咕文化科技有限公司 | Tag validity management method, device, network equipment and storage medium |
| CN111522993A (en) * | 2020-04-08 | 2020-08-11 | 咪咕文化科技有限公司 | Label validity management method and device, network equipment and storage medium |
| CN111722766A (en) * | 2020-06-04 | 2020-09-29 | 北京达佳互联信息技术有限公司 | Multimedia resource display method and device |
| CN111723237A (en) * | 2020-06-12 | 2020-09-29 | 腾讯科技（深圳）有限公司 | Media content access control method |
| CN111726649B (en) * | 2020-06-28 | 2021-12-28 | 百度在线网络技术（北京）有限公司 | Video stream processing method, device, computer equipment and medium |
| WO2022000826A1 (en) * | 2020-06-28 | 2022-01-06 | 百度在线网络技术（北京）有限公司 | Video stream processing method and apparatus, and computer device and medium |
| CN111726649A (en) * | 2020-06-28 | 2020-09-29 | 百度在线网络技术（北京）有限公司 | Video stream processing method, device, computer equipment and medium |
| CN112653907A (en) * | 2020-12-15 | 2021-04-13 | 泰康保险集团股份有限公司 | Video recommendation method and device |
| CN112653907B (en) * | 2020-12-15 | 2022-11-15 | 泰康保险集团股份有限公司 | Video recommendation method and device |
| CN112800275A (en) * | 2021-01-26 | 2021-05-14 | 广州欢网科技有限责任公司 | Short video feed stream real-time calculation pushing method, device and equipment |
| CN113312511A (en) * | 2021-06-11 | 2021-08-27 | 北京百度网讯科技有限公司 | Method, apparatus, device and computer-readable storage medium for recommending content |
| CN113312511B (en) * | 2021-06-11 | 2024-02-27 | 北京百度网讯科技有限公司 | Method, apparatus, device and computer readable storage medium for recommending content |
| CN113449147A (en) * | 2021-07-06 | 2021-09-28 | 乐视云计算有限公司 | Video recommendation method and device based on theme |
| CN114513687A (en) * | 2022-02-14 | 2022-05-17 | 青岛聚看云科技有限公司 | Server and media asset recommendation method |
| CN114513687B (en) * | 2022-02-14 | 2024-02-27 | 青岛聚看云科技有限公司 | Server and media asset recommendation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106407241A (en) | Video recommendation method and system | |
| CN103559206B (en) | A kind of information recommendation method and system | |
| CN105512331B (en) | A kind of video recommendation method and device | |
| CN107483982B (en) | Anchor recommendation method and device | |
| CN109672939B (en) | Method and device for marking video content popularity | |
| CN103729785B (en) | Video user gender classification method and device for method | |
| CN100426860C (en) | Method and apparatus for recommending items of interest to a user based on recommendations for one or more third parties | |
| CN109446419B (en) | Method and device for recommending video | |
| US8027943B2 (en) | Systems and methods for observing responses to invitations by users in a web-based social network | |
| US8555173B2 (en) | Recommendation engine | |
| CN103984741B (en) | Customer attribute information extracting method and system thereof | |
| US8893012B1 (en) | Visual indicator based on relative rating of content item | |
| CN105721899B (en) | A kind of method and system of video quality score | |
| EP2076891A2 (en) | Online entertainment network for user-contributed content | |
| CN106454536B (en) | The determination method and device of information recommendation degree | |
| CN102929966B (en) | A kind of for providing the method and system of personalized search list | |
| KR20180072793A (en) | Push Information Approximate Selection Alignment Method, Device and Computer Storage Medium | |
| CN109543132A (en) | Content recommendation method, device, electronic equipment and storage medium | |
| CN105338408B (en) | Video recommendation method based on time factor | |
| CN108363730B (en) | Content recommendation method, system and terminal equipment | |
| CN107273391A (en) | Document recommends method and apparatus | |
| CN105897847A (en) | Information push method and device | |
| Bae et al. | Is the electronic word of mouth effect always positive on the movie? | |
| CN106202570A (en) | A kind of user information acquiring method and device | |
| CN108471544A (en) | A kind of structure video user portrait method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application 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 Applicant after: Alibaba (China) Co.,Ltd. Address before: 100029, B, 6 floor, World Trade Center, No. 36 East Third Ring Road, Dongcheng District, Beijing Applicant before: Transmission network technology (Shanghai) Co., Ltd |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170215 |