CN104008139B - The creation method and device of video index table, the recommendation method and apparatus of video - Google Patents

The creation method and device of video index table, the recommendation method and apparatus of video Download PDF

Info

Publication number
CN104008139B
CN104008139B CN201410193565.6A CN201410193565A CN104008139B CN 104008139 B CN104008139 B CN 104008139B CN 201410193565 A CN201410193565 A CN 201410193565A CN 104008139 B CN104008139 B CN 104008139B
Authority
CN
China
Prior art keywords
video
ugc
videos
module
cluster
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201410193565.6A
Other languages
Chinese (zh)
Other versions
CN104008139A (en
Inventor
王敏
傅峰
傅一峰
周燕红
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing QIYI Century Science and Technology Co Ltd
Original Assignee
Beijing QIYI Century Science and Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing QIYI Century Science and Technology Co Ltd filed Critical Beijing QIYI Century Science and Technology Co Ltd
Priority to CN201410193565.6A priority Critical patent/CN104008139B/en
Publication of CN104008139A publication Critical patent/CN104008139A/en
Application granted granted Critical
Publication of CN104008139B publication Critical patent/CN104008139B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/70Information retrieval; Database structures therefor; File system structures therefor of video data
    • G06F16/71Indexing; Data structures therefor; Storage structures

Landscapes

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

Abstract

The invention provides a kind of creation method and device of video index table, the recommendation method and apparatus of video, to solve because the video of recommendation is unordered, missing, it is inaccurate the problem of.The creation method of video index table includes:Obtain the video ID of multiple UGC videos and each UGC videos;UGC videos are clustered, for each cluster distribution Video clustering ID;Description information is generated for the UGC videos in each cluster, description information is ranked up;UGC video recommendations lists are generated using the description information after sequence;First order concordance list is created using video ID and corresponding Video clustering ID, using Video clustering ID and corresponding UGC video recommendations list creating second level concordance list.Include the description information of the UGC videos after being clustered due to video recommendations list, therefore description information is more comprehensive, solves the problem of lacking some collection numbers, description information is sorted, solve the problem of order is chaotic.

Description

The creation method and device of video index table, the recommendation method and apparatus of video
Technical field
The present invention relates to multimedia technology field, the more particularly to a kind of creation method and device of video index table, with And a kind of recommendation method and apparatus of video.
Background technology
UGC full name be User Generated Content, that is, user-generated content the meaning.UGC concept is most It is early that oneself original content is shown by internet platform or other are supplied to originating from internet arena, i.e. user Users to share.UGC is accompanied by what is risen to advocate the personalized Web2.0 concepts for main feature.UGC is not a certain kind Specific business, but a kind of user uses the new paragon of internet, i.e., become to download and upload based on download by original Lay equal stress on.Good friend's social networks, video sharing network, Knowledge Sharing network, community, forum, microblogging etc. are all UGC main applications Form.
UGC videos refer to user oneself generation, and upload to major video sharing websites by Internet technology Video content.UGC videos have the feature of magnanimity, and user can be by searching in viewing other users in video sharing website The UGC videos of biography.When user watches some UGC video, the video sharing website can recommend to user some with it is current The related video of the UGC videos of viewing.
But, the UGC videos very flattening that a large number of users is uploaded in current major video sharing websites lacks rigorous Message structure.Therefore, when recommending correlation UGC videos to user, the video sequences recommended are chaotic, and may lack certain A little videos, or even the video recommended are not belonging to same classification with currently viewing video.For example, for some TV play The UGC videos of class, the collection of drama showed when recommending is sequentially chaotic, lack some collection numbers etc., so that user's needs are numerous and diverse Collection of drama positioning is carried out by search engine.
Therefore, there is the defects such as the video recommended is unordered, lack, video is inaccurate in current UGC video recommendations technologies, Cause to scan for position fixing process cumbersomely, virtually increase the turnover rate of user.
The content of the invention
The invention provides the recommendation method and dress of a kind of creation method and device of video index table, and a kind of video Put, to solve to exist in current UGC video recommendations technologies the unordered video recommended, missing, the defect such as video is inaccurate, cause Position fixing process is scanned for cumbersomely, the problem of virtually increasing the turnover rate of user.
In order to solve the above problems, the invention discloses a kind of creation method of video index table, it is characterised in that bag Include:
Obtain multiple user generated content (UGC) videos of client upload and the video ID of each UGC video;
The multiple UGC videos are clustered, and for each cluster distribution Video clustering ID;
Description information is generated respectively for the UGC videos in each described cluster, and the description information is ranked up;
UGC video recommendations lists are generated using the description information after the sequence;
First order concordance list is created using the video ID and corresponding Video clustering ID, and, it is poly- using the video Class ID and corresponding UGC video recommendations list creating second level concordance list.
Preferably, it is described acquisition client upload multiple UGC videos and each UGC video video ID the step of it Afterwards, in addition to:
The video metadata of each UGC video is extracted respectively.
Preferably, it is described to include the step of clustered to the multiple UGC videos:
Similarity Measure is carried out to the video metadata of each two UGC videos respectively;
The corresponding UGC videos of video metadata that similarity is located within default similarity interval are determined to belong to together One cluster.
Preferably, the video metadata includes video title;The UGC videos difference in each described cluster Description information is generated, and is included the step of be ranked up to the description information:
Video set number is extracted from video title according to preset rules respectively for the UGC videos in each described cluster;
The video set number of UGC videos is defined as the description information of the UGC videos;
Ascending sort or descending sort are carried out respectively for the video set number of the UGC videos in each described cluster.
Preferably, the video metadata includes video duration;In the step clustered to the multiple UGC videos After rapid, in addition to:
Video duration is obtained respectively for the UGC videos in each described cluster;
When outside the video duration of some UGC video is located at default duration interval, by the UGC videos from belonging to it Deleted in cluster.
Preferably, methods described also includes:
Third level concordance list is created using the video ID and corresponding video metadata.
According to another aspect of the present invention, a kind of recommendation method of video is also disclosed, it is characterised in that including:
The user generated content (UGC) video playback request that client is sent is received, the playing request is regarded including target UGC Frequency ID;
The corresponding Video clustering ID of the target UGC videos ID are searched in the first order concordance list being pre-created;It is described First order concordance list includes video ID and corresponding Video clustering ID;
The corresponding UGC video recommendations lists of the Video clustering ID are searched in the second level concordance list being pre-created, are made For target UGC video recommendations lists;The second level concordance list includes Video clustering ID and corresponding UGC video recommendations list; Wherein, the Video clustering ID is used for the cluster for characterizing UGC videos, and the UGC video recommendations list is by the UGC in each cluster Description information generation after video sequence;
The target UGC video recommendations lists are issued to the client.
Preferably, the description information of the UGC videos is video set number;
It is described the target UGC video recommendations lists are issued to the client the step of before, in addition to:
The corresponding video metadatas of the target UGC videos ID are searched in the third level concordance list being pre-created;It is described Third level concordance list includes video ID and corresponding video metadata, and the video metadata includes video title;
Video set number is extracted from the video title according to preset rules;
Part UGC videos before or after intercepting the video set number from the target UGC video recommendations lists are pushed away Recommend list;
It is described to include the step of the target UGC video recommendations lists are issued into the client:By the part UGC Video recommendations list is issued to the client.
Preferably, it is described to be ordered as ascending sort or descending sort;It is described to be cut from the target UGC video recommendations lists The step of taking the part UGC video recommendations lists before or after the video set number includes:
If described be ordered as ascending sort, the video set number is intercepted from the target UGC video recommendations lists Part UGC video recommendations lists afterwards;
If described be ordered as descending sort, the video set number is intercepted from the target UGC video recommendations lists Part UGC video recommendations lists before.
According to another aspect of the present invention, a kind of creating device of video index table is also disclosed, it is characterised in that bag Include:
Video acquiring module, multiple user generated content (UGC) videos and each UGC video for obtaining client upload Video ID;
Cluster module, for being clustered to the multiple UGC videos, and for each cluster distribution Video clustering ID;
Order module, for generating description information respectively for the UGC videos in each described cluster, and to the description Information is ranked up;
Generation module, for generating UGC video recommendations lists using the description information after the sequence;
First creation module, for creating first order concordance list using the video ID and corresponding Video clustering ID;
Second creation module, for using the Video clustering ID and the corresponding UGC video recommendations list creating second level Concordance list.
Preferably, described device also includes:
Abstraction module, multiple UGC videos and each UGC for obtaining client upload in the video acquiring module are regarded After the video ID of frequency, the video metadata of each UGC video is extracted respectively.
Preferably, the cluster module includes:
Calculating sub module, for carrying out Similarity Measure to the video metadata of each two UGC videos respectively;
Determination sub-module is clustered, it is corresponding for the video metadata by similarity within default similarity interval UGC videos are determined to belong to same cluster.
Preferably, the order module includes:
Extracting sub-module, for for the UGC videos in each described cluster respectively according to preset rules from video title Middle extraction video set number;
Information determination sub-module, the description information for the video set number of UGC videos to be defined as to the UGC videos;
Sorting sub-module, ascending sort is carried out for the video set number for the UGC videos in each described cluster respectively Or descending sort.
Preferably, the video metadata includes video duration;Described device also includes:
Duration acquisition module, after being clustered in the cluster module to the multiple UGC videos, for described UGC videos in each cluster obtain video duration respectively;
Video removing module, for when the video duration of some UGC video is outside default duration interval, by this UGC videos are deleted from the cluster belonging to it.
Preferably, described device also includes:
3rd creation module, for creating third level concordance list using the video ID and corresponding video metadata.
According to another aspect of the present invention, a kind of recommendation apparatus of video is also disclosed, it is characterised in that including:
Receiving module, the user generated content (UGC) video playback request for receiving client transmission, the playing request Including target UGC videos ID;
First searching modul, for searching the target UGC videos ID correspondences in the first order concordance list being pre-created Video clustering ID;The first order concordance list includes video ID and corresponding Video clustering ID;
Second searching modul, it is corresponding for searching the Video clustering ID in the second level concordance list being pre-created UGC video recommendations lists, are used as target UGC video recommendations lists;The second level concordance list is including Video clustering ID and correspondingly UGC video recommendations lists;Wherein, the Video clustering ID is used for the cluster for characterizing UGC videos, the UGC video recommendations row Table is generated by the description information after the UGC videos sequence in each cluster;
Module is issued, for the target UGC video recommendations lists to be issued into the client.
Preferably, the description information of the UGC videos is video set number;Described device also includes:
3rd searching modul, for issuing module described by the target UGC video recommendations lists and being issued to the visitor Before the end of family, the corresponding video metadatas of the target UGC videos ID are searched in the third level concordance list being pre-created;It is described Third level concordance list includes video ID and corresponding video metadata, and the video metadata includes video title;
Extraction module, for extracting video set number from the video title according to preset rules;
Interception module, for intercepting the video set number from the target UGC video recommendations lists before or after Part UGC video recommendations lists;
It is described to issue module, specifically for the part UGC video recommendations lists are issued into the client.
Preferably, it is described to be ordered as ascending sort or descending sort;The interception module includes:
First interception submodule, for it is described be ordered as ascending sort when, from the target UGC video recommendations lists Intercept the part UGC video recommendations lists after the video set number;
Second interception submodule, for it is described be ordered as descending sort when, from the target UGC video recommendations lists Intercept the part UGC video recommendations lists before the video set number.
Compared with prior art, the present invention includes advantages below:
Multiple UGC videos that client is uploaded can be clustered first in the present invention, and for each cluster distribution Video clustering ID;Then description information is generated respectively for the UGC videos in each cluster, and description information is ranked up, UGC video recommendations lists are generated using the description information after sequence;It is finally poly- using the video ID and corresponding video of UGC videos Class ID creates first order concordance list, and, indexed using Video clustering ID and the corresponding UGC video recommendations list creating second level Table.After the request of UGC video playbacks is received, you can find target UGC in first order concordance list and second level concordance list The corresponding target UGC video recommendations lists of video ID, and it is issued to client.Due to include in video recommendations list into The description information of UGC videos after row cluster, therefore these description informations are more comprehensive, solve and lack asking for some collection numbers Topic;And these description informations are sorted, therefore solve the problem of order is chaotic.
Brief description of the drawings
Fig. 1 is a kind of flow chart of the creation method of video index table of the embodiment of the present invention one;
Fig. 2 is a kind of flow chart of the recommendation method of video of the embodiment of the present invention one;
Fig. 3 is a kind of flow chart of the creation method of video index table of the embodiment of the present invention two;
Fig. 4 is a kind of flow chart of the recommendation method of video of the embodiment of the present invention two;
Fig. 5 is a kind of structured flowchart of the creating device of video index table of the embodiment of the present invention three;
Fig. 6 is a kind of structured flowchart of the recommendation apparatus of video of the embodiment of the present invention three.
Embodiment
In order to facilitate the understanding of the purposes, features and advantages of the present invention, it is below in conjunction with the accompanying drawings and specific real Applying mode, the present invention is further detailed explanation.
When user watches some UGC video in video sharing website, the website can recommend some and current sight to user The video for the UGC videos correlation seen.But, the video sequences that current video recommendations technology is recommended are chaotic, and may Lack some videos, or even the video recommended is not belonging to same classification with currently viewing video, results in the need for cumbersomely Position fixing process is scanned for, virtually increases the turnover rate of user.In view of the above-mentioned problems, the present invention proposes a kind of video index The creation method and device of table, and a kind of video recommendation method and apparatus, so as to ensure to recommend the video of user more Comprehensively, in order, Consumer's Experience is improved.
Below, the recommendation side of creation method and device of each the following embodiment to video index table, and video is passed through Method and device describe in detail respectively.
Embodiment one:
Reference picture 1, shows a kind of flow chart of the creation method of video index table of the embodiment of the present invention one, this method Specifically it may comprise steps of:
Step 101, multiple UGC videos of client upload and the video ID of each UGC video are obtained.
UGC videos refer to user oneself generation, and upload to major video sharing websites by Internet technology Video content.UGC videos have a feature of magnanimity, user can by searched in video sharing website viewing oneself or other The UGC videos that user uploads.After user uploads UGC videos by client, server can preserve these UGC videos, and And, each UGC videos also have a unique video labeling (ID)., can be to being preserved in server in the embodiment of the present invention UGC videos integrated, so as to provide condition for video recommendations process, multiple UGC of client upload can be obtained first The video ID of video and each UGC video.
Step 102, the multiple UGC videos are clustered, and for each cluster distribution Video clustering ID.
After multiple UGC videos of client upload are got, you can the multiple UGC videos are clustered, because This, which can generate to include at least one cluster, each cluster, belongs at least one same category of UGC video, then also Can be for each cluster distribution Video clustering ID, to identify each cluster.
Step 103, description information is generated respectively for the UGC videos in each described cluster, and to the description information It is ranked up.
After at least one above-mentioned cluster is generated, can also respectively it be given birth to further directed to the UGC videos in each cluster Into description information, the description information is used for the feature for describing corresponding UGC videos, can be by some according to the description information UGC videos make a distinction with others UGC videos.
Step 104, UGC video recommendations lists are generated using the description information after the sequence.
After generating description information respectively for the UGC videos in each described cluster, the description information can be entered Row sequence, the description information after sequence can characterize the order of corresponding UGC videos, then can use retouching after the sequence State information generation UGC video recommendations lists.I.e. for a cluster, the description information of UGC videos therein can be arranged Sequence, and generate a corresponding UGC video recommendations list.
The UGC recommends the description information generation of UGC video of the list of videos after being clustered, therefore included UGC videos are more comprehensive, and description information is sorted, therefore generation is that orderly UGC recommends list of videos.
Step 105, first order concordance list is created using the video ID and corresponding Video clustering ID, and, using institute State Video clustering ID and corresponding UGC video recommendations list creating second level concordance list.
According to each above-mentioned step, UGC videos are clustered, at least one UGC video can be included in each cluster, Each UGC videos have a video ID, and each cluster is allocated a Video clustering ID.It is therefore possible to use described regard Frequency ID and corresponding Video clustering ID creates first order concordance list, and a video ID can correspond to a Video clustering ID, and this is regarded Frequency cluster ID is the Video clustering ID of cluster belonging to the corresponding UGC videos of video ID;One Video clustering ID can be corresponded to One or more video ID, video ID are the video ID of the UGC videos in the corresponding clusters of Video clustering ID.
For each cluster, also the description information of UGC videos therein is ranked up, and using the description letter after sequence Breath generation UGC video recommendations lists.It is therefore possible to use the Video clustering ID and corresponding UGC video recommendations list creating Second level concordance list, a Video clustering ID can correspond to a UGC video recommendations list, and the UGC video recommendations lists are The UGC video recommendations lists of the description information generation of UGC videos in the corresponding clusters of Video clustering ID.
Generating above-mentioned UGC video recommendations list, and after creating first order concordance list and second level concordance list, you can When carrying out UGC video recommendations, the corresponding UGC video recommendations list of table search is indexed according to first order concordance list and the second level, and It is shown to user.
Below, the video index table progress that the creation method of the video index table according to above-mentioned Fig. 1 is created is introduced The method of video recommendations.
Reference picture 2, shows a kind of flow chart of the recommendation method of video of the embodiment of the present invention one, and this method specifically may be used To comprise the following steps:
Step 201, the UGC video playbacks request that client is sent is received, the playing request includes target UGC videos ID。
User is preserved when accessing video sharing website in the server that can ask the broadcasting video sharing website UGC videos, now can send UGC video playbacks by the user end to server of the video sharing website and ask, the broadcasting Video ID, i.e. target UGC videos ID that user wants the target UGC videos of broadcasting can be included in request.
Step 202, the corresponding Video clusterings of the target UGC videos ID are searched in the first order concordance list being pre-created ID。
After above-mentioned UGC video playbacks request is received, you can search and play in the first order concordance list being pre-created The included corresponding Video clustering ID of target UGC videos ID in request.The first order concordance list can think above-mentioned Fig. 1 institutes The first order concordance list created in the creation method of the video index table shown, the first order concordance list can include video ID and Corresponding Video clustering ID, therefore, its corresponding Video clustering ID can be found according to the target UGC videos ID.
Step 203, the corresponding UGC video recommendations of the Video clustering ID are searched in the second level concordance list being pre-created List, is used as target UGC video recommendations lists.
, can also be in the second level being pre-created after the corresponding Video clustering ID of the target UGC videos ID are found The corresponding UGC video recommendations lists of the Video clustering ID are searched in concordance list.The first order concordance list can think above-mentioned The second level concordance list created in the creation method of video index table shown in Fig. 1, the second level concordance list can include regarding Frequency cluster ID and corresponding UGC video recommendations list;Wherein, the Video clustering ID is used for the cluster for characterizing UGC videos, described The list of UGC video recommendations is generated by the description information after the UGC videos sequence in each cluster.Therefore, it is poly- according to the video Class ID can find its corresponding UGC video recommendations list, and the UGC video recommendations lists found are the target The corresponding UGC video recommendations lists of UGC videos ID, therefore it is used as target UGC video recommendations lists.
Step 204, the target UGC video recommendations lists are issued to the client.
Find after target UGC video recommendations lists, you can target UGC video recommendations lists are issued to the client End, therefore target UGC video recommendations lists can be shown on the display interface of client, user can be according to the target Corresponding recommendation video is played in the list of UGC video recommendations.
Include the description information of the UGC videos after being clustered in video recommendations list in the embodiment of the present invention, Therefore these description informations are more comprehensive, solve the problem of lacking some collection numbers;And these description informations are arranged Sequence, therefore solve the problem of order is chaotic.
Below, a kind of creation method of video index table and a kind of recommendation side of video are specifically described by embodiment two Method.
Embodiment two:
Reference picture 3, shows a kind of flow chart of the creation method of video index table of the embodiment of the present invention two, this method Specifically it may comprise steps of:
Step 301, multiple UGC videos of client upload and the video ID of each UGC video are obtained.
During user is uploaded onto the server UGC videos by client, server can be protected to these UGC videos Deposit, wherein, each UGC videos are to that should have a video ID.First, server can obtain UGC videos that it is preserved and The video ID of each UGC video, then carries out relevant treatment to it, to generate corresponding video index table.
Step 302, the video metadata of each UGC video is extracted respectively.
After above-mentioned multiple UGC videos are got, the video metadata of each UGC video is extracted respectively, then according to this A little video metadatas can be clustered to the multiple UGC videos.
Step 303, the multiple UGC videos are clustered, and for each cluster distribution Video clustering ID.
In the embodiment of the present invention, the video metadata can include video title, video duration, video uploader, regard The information such as frequency type, can be clustered according to these video metadatas to the multiple UGC videos, so as to will belong to same class Other UGC videos flock together.If for example, UGC videos be TV play class UGC videos, can be by same portion's TV Acute each collection flocks together;If UGC videos are the UGC videos of variety show class, can be by same variety show Each phase program aggregation is together, etc..
Due to for some UGC videos, only by its video title, video duration, video uploader, video type these Information can not be clustered exactly.For example, for the UGC videos of TV play class, the TV play of same title is potentially included Multiple versions, such as this TV play of entitled " Hero Shooting Vulture ", its potentially include 83 editions, 94 editions, 03 edition, 08 Multiple versions such as version, if only can not be by " Hero Shooting Vulture " TV play of same version by information such as above-mentioned video titles Each collection flocks together exactly.Therefore, in the embodiment of the present invention, the video metadata can also include video director, regard The information such as frequency performer, video director, the video performer of the TV play of different editions are different, therefore according to these video informations Can a certain version of unique mark TV play, so as to improve the accuracy of Video clustering.
In one preferred embodiment of the invention, the step 303 can specifically include following sub-step:
Sub-step a1, carries out Similarity Measure to the video metadata of each two UGC videos respectively;
Sub-step a2, the corresponding UGC videos of video metadata similarity being located within default similarity interval are true It is set to and belongs to same cluster.
Similarity Measure is carried out to the video metadata of each two UGC videos respectively, if similarity is located at default phase Like spending within interval, then it can illustrate that corresponding two UGC videos belong to same cluster.Set for the similarity is interval Put, it is ensured that the similarity for belonging to the video metadata of the UGC videos of same cluster can neither be too small, can not be too big, such as Fruit is too big, then illustrates that the relevant information difference of two UGC videos is very big, be not belonging to same classification, if too small, illustrate two The relevant information difference very little of individual UGC videos, it may belong to identical UGC videos.Those skilled in the art can be according to reality Border experience sets the interval corresponding numerical value of similarity, and the embodiment of the present invention is not any limitation as to specific numerical value.
In one preferred embodiment of the invention, in the similarity position for judging the video metadata of two UGC videos Outside default similarity is interval, and when being less than the smaller boundary value in the interval, for example similarity interval is [a, b], when When judging that similarity is less than boundary value a, because the two UGC videos may be considered identical UGC videos in the case of this kind, Therefore one belonged in identical UGC videos can also be further deleted, so as to avoid video from repeating, takes video resource Problem.
In the embodiment of the present invention, above-mentioned sub-step a1 can include following sub-step:
Sub-step a11, a corresponding vector is converted into by the metadata of each UGC videos;
Sub-step a12, carries out similarity meter according to corresponding vector to the video metadata of each two UGC videos respectively Calculate.
In the embodiment of the present invention, sub-step a12 can be by calculating two vectorial Euclidean distances, manhatton distance, folders The methods such as angle cosine value carry out Similarity Measure to the video metadata of each two UGC videos.For example, can calculate each two to Euclidean distance between amount, when the Euclidean distance is located within the default first distance interval, determines that two vectors are right The similarity for the video metadata answered is located within default similarity interval;The Man Ha between each two vector can also be calculated Distance, when the manhatton distance is located within default second distance interval, determines the corresponding video of two vectors The similarity of metadata is located within default similarity interval;The cosine value of the angle of each two vector can also be calculated, when When the cosine value of the angle is located within default cosine value interval, the phase of the corresponding video metadata of two vectors is determined It is located at like degree within default similarity interval.It is interval and cosine value interval for the wherein first distance interval, second distance Concrete numerical value, is utilized technical staff and carries out related setting according to practical experience, the embodiment of the present invention is to specific numerical value It is not any limitation as.
Euclidean distance is a kind of distance calculating method for being easiest to understand, the distance from point-to-point transmission in Euclidean space is public Formula.Two n-dimensional vector a (x11, x12..., x1n) and b (x21, x22..., x2n) between Euclidean distance be:The form for being expressed as vector operation can also be used:
Manhatton distance is also referred to as city block distance.Two n-dimensional vector a (x11, x12..., x1n) and b (x21, x22..., x2n) between manhatton distance be:
Included angle cosine can be used to weigh the difference of two vector directions in geometry, and this concept is borrowed in machine learning to weigh Difference between sample vector.For two n-dimensional vector a (x11, x12..., x1n) and b (x21, x22..., x2n), it can use similar The similarity degree between them is weighed in the concept of included angle cosine:I.e. Included angle cosine span is [- 1,1].Included angle cosine is bigger to represent that two vectorial angles are smaller, and included angle cosine is smaller to be represented The angle of two vectors is bigger, and when two vectorial directions are overlapped, included angle cosine takes maximum 1, when two vectorial directions are complete Opposing angles cosine takes minimum value -1.
By said process, it can will belong to same category of UGC videos and flock together, therefore can be given birth to after processing Into multiple clusters, a Video clustering ID then can be distributed for each cluster, with one cluster of unique mark.
Step 304, video duration is obtained respectively for the UGC videos in each described cluster, when regarding for some UGC video When frequency duration is located at outside default duration interval, the UGC videos are deleted from the cluster belonging to it.
In one preferred embodiment of the invention, after each cluster is got, it can also gather for described each UGC videos in class carry out Denoising disposal.Obtain video duration respectively first, and it is default to judge whether video duration is located at Duration interval within, for example, for the cluster of a certain portion's TV play " expert detective Di Ren outstanding ", each of which collection (i.e. each UGC Video) duration all at 40 minutes or so, therefore can now set duration interval be (35,45), if some UGC video Duration be located at the interval outside, then the UGC videos may not be a certain collection in the TV play, for example some UGC video When a length of 10 minutes, then the UGC videos may be film clips, now the UGC videos can then be deleted.Therefore, as some UGC When the video duration of video is located at outside default duration interval, the UGC videos can be deleted from the cluster belonging to it, from And further improve the accuracy of UGC videos in each cluster.
Step 305, description information is generated respectively for the UGC videos in each described cluster, and to the description information It is ranked up.
After above-mentioned processing is carried out to each cluster, then can respectively it be given birth to for the UGC videos in each described cluster It is ranked up into description information, and to the description information, to generate the corresponding UGC video recommendations list of the cluster.
In one preferred embodiment of the invention, the step 305 can include following sub-step:
Sub-step b1, is extracted according to preset rules from video title respectively for the UGC videos in each described cluster Video set number;
Sub-step b2, the description information of the UGC videos is defined as by the video set number of UGC videos;
Sub-step b3, ascending sort or descending are carried out for the video set number of the UGC videos in each described cluster respectively Sequence.
In one preferred embodiment of the invention, video name and video set number can be included in the video title, Video set number can be extracted from video title according to the orderly pattern base pre-established.Pattern is to refer to enter video title A set of criterion of row cutting, is usually what is formed according to data analysis, and constantly adds and safeguard.According to data analysis, Most collection of drama class video council is just like the naming method of " xxx yy collection " or " xxxyy ", and " xxx " is word, and " yy " is Numeral or date, therefore " xxx " therein can be regarded as video name, " yy " therein is regarded as into video set number.Cause This, can extract numeral or date " yy " in the embodiment of the present invention according to which from video title, i.e., as video set number.
After extracting video set number in video title, you can the video set number is defined as into retouching for correspondence UGC videos Information is stated, the description information can either make a distinction the UGC videos with other UGC videos in being clustered belonging to it, can be with Characterize the correspondence order of UGC videos.And hence it is also possible to enter respectively for the video set number of the UGC videos in each described cluster Row ascending sort or descending sort.
In the embodiment of the present invention, can according in cluster UGC videos video type determine be carry out ascending sort or Descending sort.If for example, the UGC videos in some cluster are TV play class, to the video set of the UGC videos in the cluster Number carries out ascending sort, such as the description information after sorting is the 1st collection, the 2nd collection, the 3rd collection ....If in some cluster UGC videos are variety show class, then carry out descending sort to the video set number of the UGC videos in the cluster, such as after sorting Description information is in October, 2013, in September, 2013, in August, 2013 ....Certainly, for TV play class UGC videos video Collection number can also carry out descending sort, and ascending sort, this hair can also be carried out for the video set number of variety show class UGC videos Bright embodiment is not any limitation as to this.
Step 306, UGC video recommendations lists are generated using the description information after the sequence.
After the description information of the UGC videos in for each cluster is ranked up, you can using after the sequence Description information generates the corresponding UGC video recommendations list of the acute class.The list of UGC video recommendations can be exposed to user, user According to the UGC video recommendations list video therein can be selected to play out.
Step 307, first order concordance list is created using the video ID and corresponding Video clustering ID, using the video ID and corresponding UGC video recommendations list creating second level concordance list are clustered, using the video ID and corresponding video element number According to establishment third level concordance list.
UGC videos are clustered, at least one UGC video can be included in each cluster, each UGC videos have One video ID, each cluster is allocated a Video clustering ID, it is therefore possible to use the video ID and corresponding video are poly- Class ID creates first order concordance list.One video ID can correspond to a Video clustering ID, and a Video clustering ID can be corresponded to One or more video ID.
For each cluster, also the description information of UGC videos therein is ranked up, and using the description letter after sequence Breath generation UGC video recommendations lists, it is therefore possible to use the Video clustering ID and corresponding UGC video recommendations list creating Second level concordance list.One Video clustering ID can correspond to a UGC video recommendations list.
Each UGC videos have a video ID, and each UGC videos also have corresponding video metadata, therefore, can To create third level concordance list using the video ID and corresponding video metadata.One video ID can correspond to a video Metadata.
After first order concordance list, second level concordance list and third level concordance list is created, you can pushed away carrying out UGC videos When recommending, the corresponding UGC video recommendations list of table search is indexed according to first order concordance list second level concordance list and the third level, and show Show to user.
Below, the video index table progress that the creation method of the video index table according to above-mentioned Fig. 3 is created is introduced The method of video recommendations.
Reference picture 4, shows a kind of flow chart of the recommendation method of video of the embodiment of the present invention two, the first order therein Concordance list, second level concordance list and third level concordance list can be the index set up by the method shown in above-mentioned Fig. 1 or Fig. 3 Table, UGC video recommendations list can be the video recommendations list that is generated by the method shown in above-mentioned Fig. 1 or Fig. 3.This method Specifically it may comprise steps of:
Step 401, the UGC video playbacks request that client is sent is received, the playing request includes target UGC videos ID。
User is preserved when accessing video sharing website in the server that can ask the broadcasting video sharing website UGC videos, the video that the UGC videos asked can upload for active user oneself, or what other users were uploaded regards Frequently.For example, user can input the title of target UGC videos in the search box in video sharing web station interface, and click is searched Rope, now can then send UGC video playbacks and ask to server, and target UGC videos ID can be included in the playing request.
Step 402, the corresponding Video clusterings of the target UGC videos ID are searched in the first order concordance list being pre-created ID。
Because first order concordance list includes video ID and corresponding Video clustering ID, therefore regarded receiving above-mentioned UGC After frequency playing request, you can target UGC videos included in playing request are searched in the first order concordance list being pre-created The corresponding Video clustering ID of ID.
Step 403, the corresponding UGC video recommendations of the Video clustering ID are searched in the second level concordance list being pre-created List, is used as target UGC video recommendations lists.
Because second level concordance list includes Video clustering ID and corresponding UGC video recommendations list, therefore finding After the corresponding Video clustering ID of the target UGC videos ID, it can also be searched in the second level concordance list being pre-created described The corresponding UGC video recommendations lists of Video clustering ID.
Step 404, the corresponding video element numbers of the target UGC videos ID are searched in the third level concordance list being pre-created According to.
In one preferred embodiment of the invention, further it can also be searched in the third level concordance list being pre-created The corresponding video metadata of the target UGC videos ID.The third level concordance list includes video ID and corresponding video element Data.
Step 405, video set number is extracted from the video title according to preset rules;Pushed away from the target UGC videos Recommend the part UGC video recommendations lists before or after the video set number is intercepted in list.
Video title can be included in the video metadata, the corresponding video element numbers of target UGC videos ID are being found According to rear, video set number is extracted from the video title according to preset rules, the video set number is the target UGC that user is asked The video set number of video.In the embodiment of the present invention, in order to further reduce the occupancy for recommending position, save and recommend position money Source, the part UGC videos before or after can also intercepting the video set number from the target UGC video recommendations lists are pushed away Recommend list.
It is described intercept the video set number from the target UGC video recommendations lists before or after part UGC regard The step of frequency recommendation list, can include:If described be ordered as ascending sort, from the target UGC video recommendations lists Intercept the part UGC video recommendations lists after the video set number;If described be ordered as descending sort, from the target The part UGC video recommendations lists before the video set number are intercepted in UGC video recommendations lists.
Above-mentioned interception way is the interception way for being accustomed to setting according to the viewing of user, if for example, user currently please The 4th collection for some TV play asked, then the collection of drama that the user may be subsequently watched after the 4th collection, and for TV play The corresponding UGC video recommendations list of video of class would generally carry out ascending sort, therefore can intercept the part after the 4th collection UGC video recommendations lists;If user's current request for this phase of in April, 2013 of some variety show, the user The collection of drama that may subsequently watch after this phase of in April, 2013, and pushed away for the corresponding UGC videos of video of variety show class Descending sort would generally be carried out by recommending list, therefore can intercept the part UGC video recommendations lists before in April, 2013.
Certainly, in the embodiment of the present invention, when being ordered as ascending sort, it can also be arranged from the target UGC video recommendations The part UGC video recommendations lists before the video set number are intercepted in table;, can also be from described when being ordered as descending sort The part UGC video recommendations lists after the video set number, those skilled in the art are intercepted in target UGC video recommendations lists Relevant treatment is carried out according to practical experience, the embodiment of the present invention is not any limitation as to this.
Step 406, the part UGC video recommendations lists are issued to the client.
After above-mentioned part UGC video recommendations lists are truncated to, you can only issue part UGC video recommendations lists To the client, so as to reduce the occupancy for recommending position in client, save and recommend position resource.Certainly, in order to by comprehensively The whole target UGC video recommendations lists found can also be issued to the client by UGC video recommendations to user, The embodiment of the present invention is not any limitation as to this.
Client is after UGC video recommendations lists are received, you can the UGC videos in selection UGC video recommendations lists Play out, each UGC videos also correspond to a broadcast address, client is played out according to the broadcast address, for Specific playing process, the embodiment of the present invention is no longer discussed in detail.
The embodiment of the present invention changes information organizational form originally unordered in internet, and magnanimity information is tied as far as possible Structure, ordering, are finally naturally presented to user, reduce user mutual, strengthen Consumer's Experience, for video website, Can effectively it reduce because of customer loss caused by the defect such as content missing, unordered.
For foregoing each method embodiment, in order to be briefly described, therefore it is all expressed as to a series of combination of actions, but It is that those skilled in the art should know, the present invention is not limited by described sequence of movement, because according to the present invention, certain A little steps can be carried out sequentially or simultaneously using other.Secondly, those skilled in the art should also know, be retouched in specification The embodiment stated belongs to preferred embodiment, and involved action and the module not necessarily present invention are necessary.
Embodiment three:
Reference picture 5, shows a kind of structured flowchart of the creating device of video index table of the embodiment of the present invention three, the dress Putting can specifically include with lower module:
Video acquiring module 501, multiple user generated content (UGC) videos and each UGC for obtaining client upload The video ID of video;
Cluster module 502, for being clustered to the multiple UGC videos, and for each cluster distribution Video clustering ID;
Order module 503, for generating description information respectively for the UGC videos in each described cluster, and to described Description information is ranked up;
Generation module 504, for generating UGC video recommendations lists using the description information after the sequence;
First creation module 505, for creating first order concordance list using the video ID and corresponding Video clustering ID;
Second creation module 506, for using the Video clustering ID and corresponding UGC video recommendations list creating second Level concordance list.
In one preferred embodiment of the invention, described device can also include with lower module:
Abstraction module, multiple UGC videos and each UGC for obtaining client upload in the video acquiring module are regarded After the video ID of frequency, the video metadata of each UGC video is extracted respectively;
The video metadata can include video title, video duration, video uploader, video type, video director, The information such as video performer.
Duration acquisition module, after being clustered in the cluster module to the multiple UGC videos, for described UGC videos in each cluster obtain video duration respectively;
Video removing module, for when the video duration of some UGC video is outside default duration interval, by this UGC videos are deleted from the cluster belonging to it;
3rd creation module, for creating third level concordance list using the video ID and corresponding video metadata.
In one preferred embodiment of the invention, the cluster module can include following submodule:
Calculating sub module, for carrying out Similarity Measure to the video metadata of each two UGC videos respectively;
Determination sub-module is clustered, it is corresponding for the video metadata by similarity within default similarity interval UGC videos are determined to belong to same cluster.
The order module can include following submodule:
Extracting sub-module, for for the UGC videos in each described cluster respectively according to preset rules from video title Middle extraction video set number;
Information determination sub-module, the description information for the video set number of UGC videos to be defined as to the UGC videos;
Sorting sub-module, ascending sort is carried out for the video set number for the UGC videos in each described cluster respectively Or descending sort.
, can also be according to the first order concordance list, second level concordance list and the third level of above-mentioned establishment in the embodiment of the present invention Concordance list, carries out UGC video recommendations.Reference picture 6, shows a kind of structure of the recommendation apparatus of video of the embodiment of the present invention three Block diagram, the device can specifically include with lower module:
Receiving module 601, the user generated content (UGC) video playback request for receiving client transmission is described to play Request includes target UGC videos ID;
First searching modul 602, for searching ID pairs of the target UGC videos in the first order concordance list being pre-created The Video clustering ID answered;The first order concordance list includes video ID and corresponding Video clustering ID;
Second searching modul 603, for searching the Video clustering ID correspondences in the second level concordance list being pre-created UGC video recommendations lists, be used as target UGC video recommendations lists;The second level concordance list includes Video clustering ID and right The UGC video recommendations lists answered;Wherein, the Video clustering ID is used for the cluster for characterizing UGC videos, the UGC video recommendations List is generated by the description information after the UGC videos sequence in each cluster;The description information of the UGC videos is video set Number;
Module 604 is issued, for the target UGC video recommendations lists to be issued into the client.
In one preferred embodiment of the invention, described device can also include with lower module:
3rd searching modul, for issuing module described by the target UGC video recommendations lists and being issued to the visitor Before the end of family, the corresponding video metadatas of the target UGC videos ID are searched in the third level concordance list being pre-created;It is described Third level concordance list includes video ID and corresponding video metadata, and the video metadata includes video title;
Extraction module, for extracting video set number from the video title according to preset rules;
Interception module, for intercepting the video set number from the target UGC video recommendations lists before or after Part UGC video recommendations lists;
It is described to issue module, specifically for the part UGC video recommendations lists are issued into the client.
In the embodiment of the present invention, the sequence can be ascending sort or descending sort;The interception module can include Following submodule:
First interception submodule, for it is described be ordered as ascending sort when, from the target UGC video recommendations lists Intercept the part UGC video recommendations lists after the video set number;
Second interception submodule, for it is described be ordered as descending sort when, from the target UGC video recommendations lists Intercept the part UGC video recommendations lists before the video set number.
Multiple UGC videos that client is uploaded can be clustered first in the embodiment of the present invention, and it is poly- for each Class distribution Video clustering ID;Then description information is generated respectively for the UGC videos in each cluster, and description information is carried out Sequence, UGC video recommendations lists are generated using the description information after sequence;Finally using the video ID of UGC videos and corresponding Video clustering ID creates first order concordance list, and, using Video clustering ID and corresponding UGC video recommendations list creating second Level concordance list.After the request of UGC video playbacks is received, you can find mesh in first order concordance list and second level concordance list The corresponding target UGC video recommendations lists of UGC videos ID are marked, and are issued to client.Due to being wrapped in video recommendations list The description information of the UGC videos after being clustered has been included, therefore these description informations are more comprehensive, solve and lack some collection The problem of number;And these description informations are sorted, therefore solve the problem of order is chaotic.
For device embodiment, because it is substantially similar to embodiment of the method, so description is fairly simple, it is related Part illustrates referring to the part of embodiment of the method.
Each embodiment in this specification is described by the way of progressive, what each embodiment was stressed be with Between the difference of other embodiment, each embodiment identical similar part mutually referring to.
The present invention can be described in the general context of computer executable instructions, such as program Module.Usually, program module includes performing particular task or realizes routine, program, object, the group of particular abstract data type Part, data structure etc..The present invention can also be put into practice in a distributed computing environment, in these DCEs, by Remote processing devices connected by communication network perform task.In a distributed computing environment, program module can be with Positioned at including in the local and remote computer-readable storage medium including storage device.
Finally, in addition it is also necessary to explanation, herein, such as first and second or the like relational terms be used merely to by One entity or operation make a distinction with another entity or operation, and not necessarily require or imply these entities or operation Between there is any this actual relation or order.Moreover, term " comprising ", "comprising" or its any other variant meaning Covering including for nonexcludability, so that process, method, commodity or equipment including a series of key elements not only include that A little key elements, but also other key elements including being not expressly set out, or also include be this process, method, commodity or The intrinsic key element of equipment.In the absence of more restrictions, the key element limited by sentence "including a ...", is not arranged Except also there is other identical element in the process including the key element, method, commodity or equipment.
Above to a kind of creation method and device of video index table provided by the present invention, and a kind of recommendation of video Method and apparatus, are described in detail, and specific case used herein is carried out to the principle and embodiment of the present invention Illustrate, the explanation of above example is only intended to the method and its core concept for helping to understand the present invention;Simultaneously for this area Those skilled in the art, according to the present invention thought, will change in specific embodiments and applications, to sum up Described, this specification content should not be construed as limiting the invention.

Claims (14)

1. a kind of creation method of video index table, it is characterised in that including:
Obtain multiple user generated content (UGC) videos of client upload and the video ID of each UGC video;
The video metadata of each UGC video is extracted respectively;The multiple UGC videos are clustered, and for each cluster Distribute Video clustering ID;
Description information is generated respectively for the UGC videos in each described cluster, and the description information is ranked up;Using Description information generation UGC video recommendations lists after the sequence;
First order concordance list is created using the video ID and corresponding Video clustering ID, and, using the Video clustering ID With corresponding UGC video recommendations list creating second level concordance list;
Wherein, the video metadata includes video title;The UGC videos in each described cluster are generated respectively to be retouched The step of stating information, and be ranked up to the description information includes:
Video set number is extracted from video title according to preset rules respectively for the UGC videos in each described cluster;
The video set number of UGC videos is defined as the description information of the UGC videos;
Ascending sort or descending sort are carried out respectively for the video set number of the UGC videos in each described cluster.
2. according to the method described in claim 1, it is characterised in that described the step of clustered to the multiple UGC videos Including:
Similarity Measure is carried out to the video metadata of each two UGC videos respectively;
The corresponding UGC videos of video metadata that similarity is located within default similarity interval are determined to belong to same Cluster.
3. according to the method described in claim 1, it is characterised in that the video metadata includes video duration;Described right After the step of the multiple UGC videos are clustered, in addition to:
Video duration is obtained respectively for the UGC videos in each described cluster;
When the video duration of some UGC video be located at default duration it is interval outside when, by the UGC videos from the cluster belonging to it It is middle to delete.
4. according to the method described in claim 1, it is characterised in that also include:
Third level concordance list is created using the video ID and corresponding video metadata.
5. a kind of recommendation method of video, it is characterised in that including:
The user generated content (UGC) video playback request that client is sent is received, the playing request includes target UGC videos ID;
The corresponding Video clustering ID of the target UGC videos ID are searched in the first order concordance list being pre-created;Described first Level concordance list includes video ID and corresponding Video clustering ID;
The corresponding UGC video recommendations lists of the Video clustering ID are searched in the second level concordance list being pre-created, mesh is used as Mark UGC video recommendations lists;The second level concordance list includes Video clustering ID and corresponding UGC video recommendations list;Wherein, The Video clustering ID is used for the cluster for characterizing UGC videos, and the UGC video recommendations list is by the UGC videos in each cluster Description information generation after sequence;
The corresponding video metadatas of the target UGC videos ID are searched in the third level concordance list being pre-created;Described 3rd Level concordance list includes video ID and corresponding video metadata, and the video metadata includes video title;
Video set number is extracted from the video title according to preset rules;
The part UGC video recommendations row before or after the video set number are intercepted from the target UGC video recommendations lists Table;
The target UGC video recommendations lists are issued to the client.
6. method according to claim 5, it is characterised in that the description information of the UGC videos is video set number;
It is described to include the step of the target UGC video recommendations lists are issued into the client:By the part UGC videos Recommendation list is issued to the client.
7. method according to claim 5, it is characterised in that described to be ordered as ascending sort or descending sort;It is described from The step of the part UGC video recommendations lists before or after the video set number is intercepted in the target UGC video recommendations lists Suddenly include:
If described be ordered as ascending sort, after intercepting the video set number from the target UGC video recommendations lists Part UGC video recommendations lists;
If described be ordered as descending sort, before intercepting the video set number from the target UGC video recommendations lists Part UGC video recommendations lists.
8. a kind of creating device of video index table, it is characterised in that including:
Video acquiring module, for obtaining multiple user generated content (UGC) videos of client upload and regarding for each UGC video Frequency ID;
Abstraction module, for obtaining multiple UGC videos and each UGC video that client is uploaded in the video acquiring module After video ID, the video metadata of each UGC video is extracted respectively;Wherein, the video metadata includes video title;
Cluster module, for being clustered to the multiple UGC videos, and for each cluster distribution Video clustering ID;
Order module, for generating description information respectively for the UGC videos in each described cluster, and to the description information It is ranked up;Wherein, the order module includes:Extracting sub-module, information determination sub-module, sorting sub-module;
The extracting sub-module, for for the UGC videos in each described cluster respectively according to preset rules from video title Middle extraction video set number;
Described information determination sub-module, the description information for the video set number of UGC videos to be defined as to the UGC videos;
The sorting sub-module, ascending sort is carried out for the video set number for the UGC videos in each described cluster respectively Or descending sort;
Generation module, for generating UGC video recommendations lists using the description information after the sequence;
First creation module, for creating first order concordance list using the video ID and corresponding Video clustering ID;
Second creation module, for being indexed using the Video clustering ID and the corresponding UGC video recommendations list creating second level Table.
9. device according to claim 8, it is characterised in that the cluster module includes:
Calculating sub module, for carrying out Similarity Measure to the video metadata of each two UGC videos respectively;
Determination sub-module is clustered, for similarity to be located to the corresponding UGC of video metadata within default similarity interval Video is determined to belong to same cluster.
10. device according to claim 8, it is characterised in that the video metadata includes video duration;Described device Also include:
Duration acquisition module, after being clustered in the cluster module to the multiple UGC videos, for it is described each UGC videos in cluster obtain video duration respectively;
Video removing module, for when the video duration of some UGC video is outside default duration interval, by the UGC Video is deleted from the cluster belonging to it.
11. device according to claim 8, it is characterised in that described device also includes:
3rd creation module, for creating third level concordance list using the video ID and corresponding video metadata.
12. a kind of recommendation apparatus of video, it is characterised in that including:
Receiving module, the user generated content (UGC) video playback request for receiving client transmission, the playing request includes Target UGC videos ID;
First searching modul, for being searched in the first order concordance list being pre-created, the target UGC videos ID is corresponding to be regarded Frequency cluster ID;The first order concordance list includes video ID and corresponding Video clustering ID;
Second searching modul, is regarded for searching the corresponding UGC of the Video clustering ID in the second level concordance list being pre-created Frequency recommendation list, is used as target UGC video recommendations lists;The second level concordance list includes Video clustering ID and corresponding UGC Video recommendations list;Wherein, the Video clustering ID is used for the cluster for characterizing UGC videos, and the UGC video recommendations list is by each The description information generation after the sequence of UGC videos in individual cluster;
3rd searching modul, for issuing module described by the target UGC video recommendations lists and being issued to the client Before, the corresponding video metadatas of the target UGC videos ID are searched in the third level concordance list being pre-created;Described 3rd Level concordance list includes video ID and corresponding video metadata, and the video metadata includes video title;
Extraction module, for extracting video set number from the video title according to preset rules;
Interception module, for intercepting the part before or after the video set number from the target UGC video recommendations lists UGC video recommendations lists;
Module is issued, for the target UGC video recommendations lists to be issued into the client.
13. device according to claim 12, it is characterised in that the description information of the UGC videos is video set number;Institute Stating device also includes:
It is described to issue module, specifically for the part UGC video recommendations lists are issued into the client.
14. device according to claim 12, it is characterised in that described to be ordered as ascending sort or descending sort;It is described Interception module includes:
First interception submodule, for it is described be ordered as ascending sort when, from the target UGC video recommendations lists intercept Part UGC video recommendations lists after the video set number;
Second interception submodule, for it is described be ordered as descending sort when, from the target UGC video recommendations lists intercept Part UGC video recommendations lists before the video set number.
CN201410193565.6A 2014-05-08 2014-05-08 The creation method and device of video index table, the recommendation method and apparatus of video Active CN104008139B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410193565.6A CN104008139B (en) 2014-05-08 2014-05-08 The creation method and device of video index table, the recommendation method and apparatus of video

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410193565.6A CN104008139B (en) 2014-05-08 2014-05-08 The creation method and device of video index table, the recommendation method and apparatus of video

Publications (2)

Publication Number Publication Date
CN104008139A CN104008139A (en) 2014-08-27
CN104008139B true CN104008139B (en) 2017-09-22

Family

ID=51368796

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410193565.6A Active CN104008139B (en) 2014-05-08 2014-05-08 The creation method and device of video index table, the recommendation method and apparatus of video

Country Status (1)

Country Link
CN (1) CN104008139B (en)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104333813A (en) * 2014-08-28 2015-02-04 无锡天脉聚源传媒科技有限公司 Video playing method, device and system
CN105786857B (en) * 2014-12-24 2019-12-10 Tcl集团股份有限公司 Method and system for improving video aggregation efficiency
CN105373600B (en) * 2015-10-30 2019-02-22 北京奇艺世纪科技有限公司 Video broadcasts single sequence method and device
CN108205557A (en) * 2016-12-19 2018-06-26 顺丰科技有限公司 Searching terminal, server-side and information communication search method and system
WO2018113673A1 (en) * 2016-12-23 2018-06-28 北京奇虎科技有限公司 Method and apparatus for pushing search result of variety show query
CN106649740A (en) * 2016-12-23 2017-05-10 北京奇虎科技有限公司 Method and device for recommending UGC (User Generated Content) data of computers, communication and consumer electronics based on search
CN109388729A (en) * 2017-08-14 2019-02-26 阿里巴巴集团控股有限公司 Search method, device and the audio query system of audio sub fingerprint
CN110020145A (en) * 2017-11-23 2019-07-16 北京搜狗科技发展有限公司 A kind of information recommendation method and device
CN110633379B (en) * 2019-08-29 2023-04-28 北京睿企信息科技有限公司 Graphics Processing Unit (GPU) -parallel-operation-based graphics searching system and method
CN111294636B (en) * 2020-01-21 2022-05-17 北京字节跳动网络技术有限公司 Video data adjusting method and device, electronic equipment and computer readable medium
CN111274446A (en) * 2020-03-02 2020-06-12 Oppo广东移动通信有限公司 Video processing method and related device

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103186550A (en) * 2011-12-27 2013-07-03 盛乐信息技术(上海)有限公司 Method and system for generating video-related video list

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2007015228A1 (en) * 2005-08-02 2007-02-08 Mobixell Networks Content distribution and tracking
US9009589B2 (en) * 2007-03-30 2015-04-14 Google Inc. Conversion of portable program modules for constrained displays

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103186550A (en) * 2011-12-27 2013-07-03 盛乐信息技术(上海)有限公司 Method and system for generating video-related video list

Also Published As

Publication number Publication date
CN104008139A (en) 2014-08-27

Similar Documents

Publication Publication Date Title
CN104008139B (en) The creation method and device of video index table, the recommendation method and apparatus of video
US11921805B2 (en) Web document enhancement
US10572565B2 (en) User behavior models based on source domain
US10671687B2 (en) Managing access rights to content using social media
US11601510B1 (en) Method and system for topic disambiguation and classification
JP2021103543A (en) Use of machine learning for recommending live-stream content
CN106326391B (en) Multimedia resource recommendation method and device
US11558480B2 (en) Tracking content use via social media
US9208471B2 (en) Matching users with similar interests
CN112364202A (en) Video recommendation method and device and electronic equipment
US10977677B2 (en) Contact importer
US20150237056A1 (en) Media dissemination system
CN105956131A (en) Resource search method and device
US11966433B1 (en) Subscribe to people in videos
US20150356191A1 (en) Web document enhancement
CN103412880A (en) Method and device for determining implicit associated information between multimedia resources
AU2018429394B2 (en) Media source measurement for incorporation into a censored media corpus
US20160021051A1 (en) Providing Access To Content Via Social Media
US11562014B1 (en) Generating visual media collections for a dynamic social networking account
CN104504074B (en) It is a kind of to the other processing method and processing device of multimedia class
Wang et al. Exploiting content relevance and social relevance for personalized ad recommendation on internet TV
Papadakis et al. Movie score: Personalized movie recommendation on mobile devices
CN112085390A (en) Method and system for evaluating film and television work propagation effect
KR101984773B1 (en) Classifier recall estimation for sparse topics
CN103336784A (en) Method and equipment for confirming preferred resource description information of resource

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant