CN104008139A - Method and device for creating video index table and method and device for recommending video - Google Patents

Method and device for creating video index table and method and device for recommending video Download PDF

Info

Publication number
CN104008139A
CN104008139A CN201410193565.6A CN201410193565A CN104008139A CN 104008139 A CN104008139 A CN 104008139A CN 201410193565 A CN201410193565 A CN 201410193565A CN 104008139 A CN104008139 A CN 104008139A
Authority
CN
China
Prior art keywords
video
ugc
cluster
recommendation list
module
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201410193565.6A
Other languages
Chinese (zh)
Other versions
CN104008139B (en
Inventor
王敏
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

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
    • G06F16/71—Indexing; Data structures therefor; Storage structures

Abstract

The invention provides a method and device for creating a video index table and a method and device for recommending video. The problem that recommended video is unordered, is lost and is inaccurate is solved. The method for creating the video index table comprises the steps that a plurality of pieces of UGC video and video IDs of the UGC video are obtained; the UGC video is clustered, and video clustering IDs are distributed to all clusters; description information is generated according to the UGC video in all the clusters and is ordered; a UGC video recommendation list is generated by adopting the ordered description information; a first-stage index table is created by adopting the video IDs and the corresponding video clustering IDs, and a second-stage index table is created by adopting the video clustering IDs and the corresponding UGC video recommendation list. Due to the fact that the description information of the clustered UGC video is included in the recommendation list, the description information is more comprehensive, the problem that certain episodes are lacked is solved, the description information is ordered, and the problem of the chaotic sequence is solved.

Description

The creation method of video index table and device, the recommend method of video and device
Technical field
The present invention relates to multimedia technology field, particularly relate to a kind of creation method and device of video index table, and a kind of recommend method of video and device.
Background technology
UGC full name is User Generated Content, the namely meaning of user-generated content.The concept of UGC originates from internet arena the earliest, and user shows own original content or offers other users and shares by internet platform.UGC is accompanied by advocate the Web2.0 concept that individual character turns to principal feature to rise.UGC is not a certain concrete business, but a kind of user uses the new mode of internet, by original take, is downloaded as the download of main transformer one-tenth and is uploaded and lay equal stress on.Good friend's social networks, video sharing network, Knowledge Sharing network, community, forum, microblogging etc. are all the main application forms of UGC.
UGC video refers to what user oneself generated, and by Internet technology, uploads to the video content of each large video sharing website.UGC video has the feature of magnanimity, and user can be by the UGC video that search watches other users to upload in video sharing website.When user watches some UGC videos, some videos relevant to the current UGC video of watching can be recommended to user in this video sharing website.
But the very flattening of UGC video that in each large video sharing website, a large number of users is uploaded at present, lacks rigorous message structure.Therefore, when recommending relevant UGC video to user, the video order of recommending is chaotic, and may lack some video, and the video even recommended and the current video of watching do not belong to same classification.For example, for the UGC video of some TV play classes, the collection of drama order that represents when recommending is chaotic, lack some collection number etc., thereby makes user need numerous and diverse by search engine, to carry out collection of drama location.
Therefore, the defect such as video that exist to recommend in current UGC video recommended technology is unordered, disappearance, video are inaccurate causes searching for position fixing process loaded down with trivial detailsly, virtually increases user's turnover rate.
Summary of the invention
The invention provides a kind of creation method and device of video index table, and a kind of recommend method of video and device, to solve in current UGC video recommended technology, exist the video of recommendation unordered, the defect such as disappearance, video are inaccurate, cause searching for position fixing process loaded down with trivial detailsly, virtually increase the problem of user's turnover rate.
In order to address the above problem, the invention discloses a kind of creation method of video index table, it is characterized in that, comprising:
Obtain a plurality of user generated content (UGC) videos of client upload and the video ID of each UGC video;
Described a plurality of UGC videos are carried out to cluster, and distribute Video clustering ID for each cluster;
For the UGC video in described each cluster, generate respectively descriptor, and described descriptor is sorted;
Adopt the descriptor after described sequence to generate UGC video recommendation list;
Adopt described video ID and corresponding Video clustering ID to create first order concordance list, and, adopt described Video clustering ID and corresponding UGC video recommendation list to create second level concordance list.
Preferably, after the described step of obtaining a plurality of UGC videos of client upload and the video ID of each UGC video, also comprise:
Extract respectively the video metadata of each UGC video.
Preferably, the described step that described a plurality of UGC videos are carried out to cluster comprises:
Respectively the video metadata of every two UGC videos is carried out to similarity calculating;
UGC video corresponding to video metadata that similarity is positioned within default similarity interval is defined as belonging to same cluster.
Preferably, described video metadata comprises video title; Describedly for the UGC video in described each cluster, generate respectively descriptor, and the step that described descriptor is sorted comprises:
For the UGC video in described each cluster, according to preset rules, from video title, extract respectively video set number;
The video set number of UGC video is defined as to the descriptor of this UGC video;
Video set number for the UGC video in described each cluster carries out respectively ascending sort or descending sort.
Preferably, described video metadata comprises video duration; Described described a plurality of UGC videos are carried out to the step of cluster after, also comprise:
For the UGC video in described each cluster, obtain respectively video duration;
When the video duration of certain UGC video is positioned at outside default duration interval, will the cluster of this UGC video under it, delete.
Preferably, described method also comprises:
Adopt described video ID and corresponding video metadata to create third level concordance list.
According to a further aspect in the invention, also disclose a kind of recommend method of video, it is characterized in that, having comprised:
Receive the user generated content (UGC) video playback request that client sends, described playing request comprises target UGC video ID;
In the first order concordance list being pre-created, search Video clustering ID corresponding to described target UGC video ID; Described first order concordance list comprises video ID and corresponding Video clustering ID;
In the second level concordance list being pre-created, search the UGC video recommendation list that described Video clustering ID is corresponding, as target UGC video recommendation list; Described second level concordance list comprises Video clustering ID and corresponding UGC video recommendation list; Wherein, described Video clustering ID is for characterizing the cluster of UGC video, and the descriptor after the UGC video sequence of described UGC video recommendation list in each cluster generates;
Described target UGC video recommendation list is issued to described client.
Preferably, the descriptor of described UGC video is video set number;
Described described target UGC video recommendation list is issued to the step of described client before, also comprise:
In the third level concordance list being pre-created, search video metadata corresponding to described target UGC video ID; Described third level concordance list comprises video ID and corresponding video metadata, and described video metadata comprises video title;
According to preset rules, from described video title, extract video set number;
From described target UGC video recommendation list, intercept the part UGC video recommendation list before or after described video set number;
The described step that described target UGC video recommendation list is issued to described client comprises: described part UGC video recommendation list is issued to described client.
Preferably, described sequence is ascending sort or descending sort; The described step that intercepts the part UGC video recommendation list before or after described video set number from described target UGC video recommendation list comprises:
If described sequence is ascending sort, from described target UGC video recommendation list, intercept described video set number part UGC video recommendation list afterwards;
If described sequence is descending sort, from described target UGC video recommendation list, intercept described video set number part UGC video recommendation list before.
According to a further aspect in the invention, also disclose a kind of creation apparatus of video index table, it is characterized in that, having comprised:
Video acquiring module, for obtaining a plurality of user generated content (UGC) videos of client upload and the video ID of each UGC video;
Cluster module, for described a plurality of UGC videos are carried out to cluster, and distributes Video clustering ID for each cluster;
Order module, generates respectively descriptor for the UGC video for described each cluster, and described descriptor is sorted;
Generation module, for adopting the descriptor after described sequence to generate UGC video recommendation list;
The first creation module, for adopting described video ID and corresponding Video clustering ID to create first order concordance list;
The second creation module, for adopting described Video clustering ID and corresponding UGC video recommendation list to create second level concordance list.
Preferably, described device also comprises:
Abstraction module, for obtain a plurality of UGC videos of client upload and the video ID of each UGC video in described video acquiring module after, extracts respectively the video metadata of each UGC video.
Preferably, described cluster module comprises:
Calculating sub module, for carrying out similarity calculating to the video metadata of every two UGC videos respectively;
Cluster is determined submodule, for UGC video corresponding to video metadata similarity being positioned within default similarity interval, is defined as belonging to same cluster.
Preferably, described order module comprises:
Extract submodule, for the UGC video for described each cluster, according to preset rules, from video title, extract respectively video set number;
Information is determined submodule, for the video set number of UGC video being defined as to the descriptor of this UGC video;
Sequence submodule, carries out respectively ascending sort or descending sort for the video set number of the UGC video for described each cluster.
Preferably, described video metadata comprises video duration; Described device also comprises:
Duration acquisition module, for after described cluster module is carried out cluster to described a plurality of UGC videos, obtains respectively video duration for the UGC video in described each cluster;
Video removing module, while being positioned at outside default duration interval for the video duration when certain UGC video, will delete the cluster of this UGC video under it.
Preferably, described device also comprises:
The 3rd creation module, for adopting described video ID and corresponding video metadata to create third level concordance list.
According to a further aspect in the invention, also disclose a kind of recommendation apparatus of video, it is characterized in that, having comprised:
Receiver module, the user generated content (UGC) video playback request sending for receiving client, described playing request comprises target UGC video ID;
First searches module, for the first order concordance list being pre-created, searches Video clustering ID corresponding to described target UGC video ID; Described first order concordance list comprises video ID and corresponding Video clustering ID;
Second searches module, for the second level concordance list being pre-created, searches the UGC video recommendation list that described Video clustering ID is corresponding, as target UGC video recommendation list; Described second level concordance list comprises Video clustering ID and corresponding UGC video recommendation list; Wherein, described Video clustering ID is for characterizing the cluster of UGC video, and the descriptor after the UGC video sequence of described UGC video recommendation list in each cluster generates;
Issue module, for described target UGC video recommendation list is issued to described client.
Preferably, the descriptor of described UGC video is video set number; Described device also comprises:
The 3rd searches module, for issuing before module is issued to described client by described target UGC video recommendation list described, searches video metadata corresponding to described target UGC video ID in the third level concordance list being pre-created; Described third level concordance list comprises video ID and corresponding video metadata, and described video metadata comprises video title;
Extraction module, for extracting video set number according to preset rules from described video title;
Interception module, for intercepting the part UGC video recommendation list before or after described video set number from described target UGC video recommendation list;
The described module that issues, specifically for being issued to described client by described part UGC video recommendation list.
Preferably, described sequence is ascending sort or descending sort; Described interception module comprises:
The first intercepting submodule for when described sequence is ascending sort, intercepts described video set number part UGC video recommendation list afterwards from described target UGC video recommendation list;
The second intercepting submodule for when described sequence is descending sort, intercepts described video set number part UGC video recommendation list before from described target UGC video recommendation list.
Compared with prior art, the present invention includes following advantage:
In the present invention, first can carry out cluster to a plurality of UGC videos of client upload, and distribute Video clustering ID for each cluster; Then for the UGC video in each cluster, generate respectively descriptor, and descriptor is sorted, adopt the descriptor after sequence to generate UGC video recommendation list; Finally adopt the video ID of UGC video and corresponding Video clustering ID to create first order concordance list, and, adopt Video clustering ID and corresponding UGC video recommendation list to create second level concordance list.After receiving the request of UGC video playback, can in first order concordance list and second level concordance list, find the target UGC video recommendation list that target UGC video ID is corresponding, and be issued to client.Because video recommendation list has comprised the descriptor of carrying out cluster UGC video afterwards, so these descriptors are more comprehensive, solved the problem that lacks some collection number; And these descriptors are sorted, therefore solved the chaotic problem of order.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of creation method of a kind of video index table of the embodiment of the present invention one;
Fig. 2 is the process flow diagram of recommend method of a kind of video of the embodiment of the present invention one;
Fig. 3 is the process flow diagram of creation method of a kind of video index table of the embodiment of the present invention two;
Fig. 4 is the process flow diagram of recommend method of a kind of video of the embodiment of the present invention two;
Fig. 5 is the structured flowchart of creation apparatus of a kind of video index table of the embodiment of the present invention three;
Fig. 6 is the structured flowchart of recommendation apparatus of a kind of video of the embodiment of the present invention three.
Embodiment
For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, below in conjunction with the drawings and specific embodiments, the present invention is further detailed explanation.
When user watches certain UGC video in video sharing website, some videos relevant to the current UGC video of watching can be recommended to user in this website.But, the video order that current video recommended technology is recommended is chaotic, and may lack some video, and the video even recommended and the current video of watching do not belong to same classification, cause needs to search for position fixing process loaded down with trivial detailsly, virtually increase user's turnover rate.For the problems referred to above, the present invention proposes a kind of creation method and device of video index table, and a kind of recommend method of video and device, more comprehensive, orderly thereby user's video is recommended in assurance, improve user and experience.
Below, by following each embodiment, to the creation method of video index table and device, and the recommend method of video and device describe in detail respectively.
Embodiment mono-:
With reference to Fig. 1, show the process flow diagram of creation method of a kind of video index table of the embodiment of the present invention one, the method specifically can comprise the following steps:
Step 101, obtains a plurality of UGC videos of client upload and the video ID of each UGC video.
UGC video refers to what user oneself generated, and by Internet technology, uploads to the video content of each large video sharing website.UGC video has the feature of magnanimity, and user can watch UGC video own or that other users upload by search in video sharing website.User, by after client upload UGC video, server can be preserved these UGC videos, and each UGC video also has a unique video labeling (ID).In the embodiment of the present invention, can integrate the UGC video of preserving in server, thereby for video recommendation process provides condition, first can obtain a plurality of UGC videos of client upload and the video ID of each UGC video.
Step 102, carries out cluster to described a plurality of UGC videos, and distributes Video clustering ID for each cluster.
After getting a plurality of UGC videos of client upload, can carry out cluster to described a plurality of UGC videos, therefore can generate at least one cluster, in each cluster, include and belong to other at least one UGC video of same class, then can also distribute Video clustering ID for each cluster, to identify each cluster.
Step 103, generates respectively descriptor for the UGC video in described each cluster, and described descriptor is sorted.
After generating above-mentioned at least one cluster, can also further for the UGC video in each cluster, generate respectively descriptor, described descriptor, for describing the feature of corresponding UGC video, can be distinguished certain UGC video and other UGC video according to this descriptor.
Step 104, adopts the descriptor after described sequence to generate UGC video recommendation list.
After generating descriptor respectively for the UGC video in described each cluster, can sort to described descriptor, descriptor after sequence can characterize the order of corresponding UGC video, then can adopt the descriptor after described sequence to generate UGC video recommendation list.For a cluster, the descriptor of UGC video wherein can be sorted, and generate a corresponding UGC video recommendation list.
Described UGC recommends list of videos to be generated by the descriptor of carrying out cluster UGC video afterwards, and therefore included UGC video is more comprehensive, and descriptor is sorted, and the orderly UGC that is therefore generating recommends list of videos.
Step 105, adopts described video ID and corresponding Video clustering ID to create first order concordance list, and, adopt described Video clustering ID and corresponding UGC video recommendation list to create second level concordance list.
According to above-mentioned each step, UGC video is carried out to cluster, in each cluster, can comprise at least one UGC video, each UGC video has a video ID, and each cluster is assigned with a Video clustering ID.Therefore, can adopt described video ID and corresponding Video clustering ID to create first order concordance list, video ID can a corresponding Video clustering ID, and this Video clustering ID is the Video clustering ID of cluster under the UGC video that this video ID is corresponding; Video clustering ID can corresponding one or more video ID, and this video ID is the video ID of the UGC video in the cluster that this Video clustering ID is corresponding.
For each cluster, also the descriptor of UGC video is wherein sorted, and adopt the descriptor after sequence to generate UGC video recommendation list.Therefore, can adopt described Video clustering ID and corresponding UGC video recommendation list to create second level concordance list, Video clustering ID can a corresponding UGC video recommendation list, and this UGC video recommendation list is the UGC video recommendation list that the descriptor of the UGC video in the cluster that this Video clustering ID is corresponding generates.
Generating above-mentioned UGC video recommendation list, and after creating first order concordance list and second level concordance list, can, when carrying out the recommendation of UGC video, according to first order concordance list and second level concordance list, search corresponding UGC video recommendation list, and be shown to user.
Below, the method that the video index table that introduction creates according to the creation method of the video index table shown in above-mentioned Fig. 1 carries out video recommendation.
With reference to Fig. 2, show the process flow diagram of recommend method of a kind of video of the embodiment of the present invention one, the method specifically can comprise the following steps:
Step 201, receives the UGC video playback request that client sends, and described playing request comprises target UGC video ID.
User is when accessing video sharing website, can ask to play the UGC video of preserving in the server of this video sharing website, now can send by the user end to server of this video sharing website the request of UGC video playback, in this playing request, can comprise that user wants the video ID of the target UGC video of broadcasting, i.e. target UGC video ID.
Step 202 is searched Video clustering ID corresponding to described target UGC video ID in the first order concordance list being pre-created.
After receiving the request of above-mentioned UGC video playback, can in the first order concordance list being pre-created, search playing request in included Video clustering ID corresponding to target UGC video ID.Described first order concordance list can be thought the first order concordance list creating in the creation method of the video index table shown in above-mentioned Fig. 1, this first order concordance list can comprise video ID and corresponding Video clustering ID, therefore, according to described target UGC video ID, can find its corresponding Video clustering ID.
Step 203 is searched the UGC video recommendation list that described Video clustering ID is corresponding in the second level concordance list being pre-created, as target UGC video recommendation list.
After finding the Video clustering ID that described target UGC video ID is corresponding, can also in the second level concordance list being pre-created, search the UGC video recommendation list that described Video clustering ID is corresponding.Described first order concordance list can be thought the second level concordance list creating in the creation method of the video index table shown in above-mentioned Fig. 1, and this second level concordance list can comprise Video clustering ID and corresponding UGC video recommendation list; Wherein, described Video clustering ID is for characterizing the cluster of UGC video, and the descriptor after the UGC video sequence of described UGC video recommendation list in each cluster generates.Therefore, according to described Video clustering ID, can find its corresponding UGC video recommendation list, the UGC video recommendation list finding is the UGC video recommendation list that described target UGC video ID is corresponding, therefore as target UGC video recommendation list.
Step 204, is issued to described client by described target UGC video recommendation list.
Find after target UGC video recommendation list, this target UGC video recommendation list can be issued to described client, therefore on the display interface of client, can show this target UGC video recommendation list, user can play the corresponding video of recommending according to this target UGC video recommendation list.
Video recommendation list in the embodiment of the present invention has comprised the descriptor of carrying out cluster UGC video afterwards, so these descriptors are more comprehensive, has solved the problem that lacks some collection number; And these descriptors are sorted, therefore solved the chaotic problem of order.
Below, by embodiment bis-, specifically describe a kind of creation method of video index table and a kind of recommend method of video.
Embodiment bis-:
With reference to Fig. 3, show the process flow diagram of creation method of a kind of video index table of the embodiment of the present invention two, the method specifically can comprise the following steps:
Step 301, obtains a plurality of UGC videos of client upload and the video ID of each UGC video.
During user uploads onto the server UGC video by client, server can be preserved these UGC videos, and wherein, each UGC video is to there being a video ID.First, server can obtain the video ID of its UGC video of preserving and each UGC video, then it is carried out to relevant treatment, to generate corresponding video index table.
Step 302, extracts respectively the video metadata of each UGC video.
After getting above-mentioned a plurality of UGC video, extract respectively the video metadata of each UGC video, then according to these video metadatas, can carry out cluster to described a plurality of UGC videos.
Step 303, carries out cluster to described a plurality of UGC videos, and distributes Video clustering ID for each cluster.
In the embodiment of the present invention, described video metadata can comprise the information such as video title, video duration, video uploader, video type, can to described a plurality of UGC videos, carry out cluster according to these video metadatas, thereby will belong to other UGC video of same class, flock together.For example, if UGC video is the UGC video of TV play class, can respectively gathering and gathering together same portion TV play; If UGC video is the UGC video of variety show class, each phase program of same variety show can be flocked together, etc.
Due to for some UGC video, only by its video title, video duration, video uploader, these information of video type, cannot carry out exactly cluster.For example, UGC video for TV play class, the TV play of same title may comprise a plurality of versions, for this TV play that is " Hero Shooting Vulture " for title, it may comprise a plurality of versions such as 83 editions, 94 editions, 03 edition, 08 edition, if only each collection of " Hero Shooting Vulture " TV play of same version cannot be flocked together exactly by information such as above-mentioned video titles.Therefore, in the embodiment of the present invention, described video metadata can also comprise the information such as video director, video performer, the video director of the TV play of different editions, video performer are different, therefore according to these video informations, get final product the TV play of a certain version of unique identification, thereby improve the accuracy of Video clustering.
In a preferred embodiment of the present invention, this step 303 specifically can comprise following sub-step:
Sub-step a1, carries out similarity calculating to the video metadata of every two UGC videos respectively;
Sub-step a2, UGC video corresponding to video metadata that similarity is positioned within default similarity interval is defined as belonging to same cluster.
Respectively the video metadata of every two UGC videos is carried out to similarity calculating, if within similarity is positioned at default similarity interval, can illustrate that two corresponding UGC videos belong to same cluster.Setting for described similarity interval, the similarity of video metadata that guarantees to belong to the UGC video of same cluster can not be too little, can not be too large, if too large, the relevant information difference that two UGC videos are described is very large, does not belong to same classification, if too little, the relevant information difference that two UGC videos are described is very little, and it may belong to identical UGC video.Those skilled in the art can arrange the corresponding numerical value in similarity interval according to practical experience, the embodiment of the present invention is not limited concrete numerical value.
In a preferred embodiment of the present invention, in the similarity of judging the video metadata of two UGC videos, be positioned at outside default similarity interval, and while being less than the less boundary value in this interval, for example similarity interval is [a, b], when judging similarity and be less than boundary value a, because these two UGC videos in this kind of situation can be thought identical UGC video, therefore can also further delete in the UGC video that belongs to identical, thereby avoid video to repeat, take the problem of video resource.
In the embodiment of the present invention, above-mentioned sub-step a1 can comprise following sub-step:
Sub-step a11, becomes a corresponding vector by the metadata conversion of UGC video described in each;
Sub-step a12, carries out similarity calculating according to corresponding vector to the video metadata of every two UGC videos respectively.
In the embodiment of the present invention, sub-step a12 can carry out similarity calculating to the video metadata of every two UGC videos by calculating the methods such as two vectorial Euclidean distances, manhatton distance, included angle cosine value.For example, can calculate the Euclidean distance between every two vectors, in the time of within described Euclidean distance is positioned between the first default distance regions, within the similarity of determining the video metadatas that these two vectors are corresponding is positioned at default similarity interval; Can also calculate the manhatton distance between every two vectors, in the time of within described manhatton distance is positioned at default second distance interval, within the similarity of determining the video metadatas that these two vectors are corresponding is positioned at default similarity interval; Can also calculate the cosine value of every two vectorial angles, when the cosine value of described angle is positioned within default cosine value interval, within the similarity of determining the video metadatas that these two vectors are corresponding is positioned at default similarity interval.For wherein between the first distance regions, the concrete numerical value in second distance interval and cosine value interval, be utilized technician and carry out related setting according to practical experience, the embodiment of the present invention is not limited concrete numerical value.
Euclidean distance is the most understandable a kind of distance calculating method, is derived from the range formula of point-to-point transmission in Euclidean space.Two n-dimensional vector a (x 11, x 12..., x 1n) and b (x 21, x 22..., x 2n) between Euclidean distance be: also can be by the form that is expressed as vector operation: d 12 = ( a - b ) ( a - b ) T .
Manhatton distance is also referred to as city block distance.Two n-dimensional vector a (x 11, x 12..., x 1n) and b (x 21, x 22..., x 2n) between manhatton distance be:
In how much, included angle cosine can be used to weigh the difference of two vector direction, uses this concept and weigh the difference between sample vector in machine learning.For two n-dimensional vector a (x 11, x 12..., x 1n) and b (x 21, x 22..., x 2n), can weigh the similarity degree between them by the concept that is similar to included angle cosine: cos ( θ ) = a · b | a | | b | , ? cos ( θ ) = Σ k = 1 n x 1 k x 2 k Σ k = 1 n x 1 k 2 Σ k = 1 n x 2 k 2 . Included angle cosine span is [1,1].Two vectorial angles of the larger expression of included angle cosine are less, and the angle of less expression two vectors of included angle cosine is larger, and when two vectorial directions overlap, included angle cosine is got maximal value 1, when the completely contrary included angle cosine of two vectorial directions is got minimum value-1.
Through said process, can flock together belonging to other UGC video of same class, after therefore processing, can generate a plurality of clusters, then for each cluster, can distribute a Video clustering ID, with cluster of unique identification.
Step 304, obtains respectively video duration for the UGC video in described each cluster, when the video duration of certain UGC video is positioned at outside default duration interval, will the cluster of this UGC video under it, delete.
In a preferred embodiment of the present invention, after getting each cluster, can also carry out Denoising disposal for the UGC video in described each cluster.Obtain video duration first respectively, and within judging whether video duration is positioned at default duration interval, for example, cluster for a certain portion TV play " expert detective Di Ren is outstanding ", the duration of each collection (being each UGC video) is wherein about 40 minutes, therefore duration now can be set interval for (35, 45), if the duration of some UGC videos is positioned at outside this interval, this UGC video may not be a certain collection in this TV play, for example the duration of certain UGC video is 10 minutes, this UGC video may be film clips, now this UGC video can be deleted.Therefore, when the video duration of certain UGC video is positioned at outside default duration interval, can will the cluster of this UGC video under it, delete, thereby further improve the accuracy of UGC video in each cluster.
Step 305, generates respectively descriptor for the UGC video in described each cluster, and described descriptor is sorted.
After each cluster is carried out to above-mentioned processing, can generate respectively descriptor for the UGC video in described each cluster, and described descriptor is sorted, to generate UGC video recommendation list corresponding to this cluster.
In a preferred embodiment of the present invention, this step 305 can comprise following sub-step:
Sub-step b1 extracts respectively video set number for the UGC video in described each cluster from video title according to preset rules;
Sub-step b2, is defined as the video set number of UGC video the descriptor of this UGC video;
Sub-step b3, carries out respectively ascending sort or descending sort for the video set number of the UGC video in described each cluster.
In a preferred embodiment of the present invention, in described video title, can comprise video name and video set number, can from video title, extract video set number according to the orderly pattern base of setting up in advance.Pattern refers to a set of criterion that video title can be cut, generally according to data analysis, forms, and constantly adds and safeguard.According to data analysis, the video council of most collection of drama class is just like the naming method of " xxx yy collection " or " xxxyy ", and " xxx " is word, and " yy " is numeral or date, therefore " xxx " wherein can be regarded as to video name, " yy " wherein regarded as to video set number.Therefore, in the embodiment of the present invention, can according to which, from video title, extract numeral or date " yy ", as video set number.
Extract video set number in video title after, this video set number can be defined as to the descriptor of corresponding UGC video, this descriptor can either by this UGC video and under it UGC of other in cluster video distinguish, can also characterize the corresponding order of UGC video.Therefore, can also carry out respectively ascending sort or descending sort for the video set number of the UGC video in described each cluster.
In the embodiment of the present invention, can determine and carry out ascending sort or descending sort according to the video type of UGC video in cluster.For example, if the UGC video in certain cluster is TV play class, the video set number of the UGC video in this cluster is carried out to ascending sort, for example the descriptor after sequence is the 1st collection, the 2nd collection, the 3rd collection ...If the UGC video in certain cluster is variety show class, the video set number of the UGC video in this cluster is carried out to descending sort, for example the descriptor after sequence is in October, 2013, in September, 2013, in August, 2013 ...Certainly, for the video set number of the UGC video of TV play class, can carry out descending sort yet, for the video set number of variety show class UGC video, can carry out ascending sort yet, the embodiment of the present invention is not limited this.
Step 306, adopts the descriptor after described sequence to generate UGC video recommendation list.
After the descriptor of the UGC video in each cluster sorts, can adopt the descriptor after described sequence to generate UGC video recommendation list corresponding to this play class.UGC video recommendation list can be demonstrated to user, and user can select video wherein to play according to this UGC video recommendation list.
Step 307, adopt described video ID and corresponding Video clustering ID to create first order concordance list, adopt described Video clustering ID and corresponding UGC video recommendation list to create second level concordance list, adopt described video ID and corresponding video metadata to create third level concordance list.
UGC video is carried out to cluster, in each cluster, can comprise at least one UGC video, each UGC video has a video ID, and each cluster is assigned with a Video clustering ID, therefore, can adopt described video ID and corresponding Video clustering ID to create first order concordance list.Video ID can a corresponding Video clustering ID, and a Video clustering ID can corresponding one or more video ID.
For each cluster, also the descriptor of UGC video is wherein sorted, and adopt the descriptor after sequence to generate UGC video recommendation list, therefore, can adopt described Video clustering ID and corresponding UGC video recommendation list to create second level concordance list.A Video clustering ID can a corresponding UGC video recommendation list.
Each UGC video has a video ID, and each UGC video also has corresponding video metadata, therefore, can adopt described video ID and corresponding video metadata to create third level concordance list.A video ID can a corresponding video metadata.
After creating first order concordance list, second level concordance list and third level concordance list, can, when carrying out the recommendation of UGC video, according to first order concordance list second level concordance list and third level concordance list, search corresponding UGC video recommendation list, and be shown to user.
Below, the method that the video index table that introduction creates according to the creation method of the video index table shown in above-mentioned Fig. 3 carries out video recommendation.
With reference to Fig. 4, show the process flow diagram of recommend method of a kind of video of the embodiment of the present invention two, first order concordance list wherein, second level concordance list and third level concordance list can be the concordance list of setting up by the method shown in above-mentioned Fig. 1 or Fig. 3, and UGC video recommendation list can be the video recommendation list generating by the method shown in above-mentioned Fig. 1 or Fig. 3.The method specifically can comprise the following steps:
Step 401, receives the UGC video playback request that client sends, and described playing request comprises target UGC video ID.
User, when accessing video sharing website, can ask to play the UGC video of preserving in the server of this video sharing website, the video that the UGC video of asking can be uploaded for active user oneself, the video that also can upload for other users.For example, user can input the title of target UGC video in the search box in video sharing web station interface, and clicks search, now can send the request of UGC video playback to server, can comprise target UGC video ID in this playing request.
Step 402 is searched Video clustering ID corresponding to described target UGC video ID in the first order concordance list being pre-created.
Because first order concordance list comprises video ID and corresponding Video clustering ID, therefore after receiving the request of above-mentioned UGC video playback, can in the first order concordance list being pre-created, search playing request in included Video clustering ID corresponding to target UGC video ID.
Step 403 is searched the UGC video recommendation list that described Video clustering ID is corresponding in the second level concordance list being pre-created, as target UGC video recommendation list.
Because second level concordance list comprises Video clustering ID and corresponding UGC video recommendation list, therefore after finding the Video clustering ID that described target UGC video ID is corresponding, can also in the second level concordance list being pre-created, search the UGC video recommendation list that described Video clustering ID is corresponding.
Step 404 is searched video metadata corresponding to described target UGC video ID in the third level concordance list being pre-created.
In a preferred embodiment of the present invention, can also further in the third level concordance list being pre-created, search video metadata corresponding to described target UGC video ID.Described third level concordance list comprises video ID and corresponding video metadata.
Step 405 is extracted video set number from described video title according to preset rules; From described target UGC video recommendation list, intercept the part UGC video recommendation list before or after described video set number.
In described video metadata, can comprise video title, after finding the video metadata that target UGC video ID is corresponding, according to preset rules, from this video title, extract video set number, this video set number is the video set number of the target UGC video that user asks.In the embodiment of the present invention, in order further to reduce for the occupancy of recommending position, to save, recommend a position resource, can also from described target UGC video recommendation list, intercept the part UGC video recommendation list before or after described video set number.
The described step that intercepts the part UGC video recommendation list before or after described video set number from described target UGC video recommendation list can comprise: if described sequence is ascending sort, from described target UGC video recommendation list, intercept the part UGC video recommendation list after described video set number; If described sequence is descending sort, from described target UGC video recommendation list, intercept described video set number part UGC video recommendation list before.
Above-mentioned interception way is just according to user's the interception way of watching custom to arrange, for example, if user's current request is the 4th collection of some TV play, this user is follow-up may watch the 4th collection collection of drama afterwards, and conventionally can carry out ascending sort for UGC video recommendation list corresponding to the video of TV play class, therefore can intercept the 4th collection part UGC video recommendation list afterwards; If user's current request is this first phase of in April, 2013 of some variety shows; the follow-up collection of drama that may watch after in April, 2013 this first phase of this user; and conventionally can carry out descending sort for UGC video recommendation list corresponding to the video of variety show class, therefore can intercept in April, 2013 part UGC video recommendation list before.
Certainly, in the embodiment of the present invention, when sequence is ascending sort, also can from described target UGC video recommendation list, intercept described video set number part UGC video recommendation list before; When sequence is descending sort, also can from described target UGC video recommendation list, intercept described video set number part UGC video recommendation list afterwards, those skilled in the art carry out relevant treatment according to practical experience, and the embodiment of the present invention is not limited this.
Step 406, is issued to described client by described part UGC video recommendation list.
After being truncated to above-mentioned part UGC video recommendation list, can only this part UGC video recommendation list be issued to described client, thereby reduce the occupancy of recommending position in client, save and recommend position resource.Certainly, for comprehensive UGC video is recommended to user, found whole target UGC video recommendation list can be issued to described client, the embodiment of the present invention is not limited this yet.
Client is after receiving UGC video recommendation list, can select the UGC video in UGC video recommendation list to play, each UGC video is a corresponding broadcast address also, client is play according to this broadcast address, for concrete playing process, the embodiment of the present invention is discussed no longer in detail.
The embodiment of the present invention has changed originally Information Organization form unordered in internet, by magnanimity information structuring as far as possible, ordering, finally naturally represent to user, reduce user interactions, strengthening user experiences, for video website, can effectively reduce the customer loss causing because of content disappearance, the defect such as unordered.
For aforesaid each embodiment of the method, for simple description, therefore it is all expressed as to a series of combination of actions, but those skilled in the art should know, the present invention is not subject to the restriction of described sequence of movement, because according to the present invention, some step can adopt other orders or carry out simultaneously.Secondly, those skilled in the art also should know, the embodiment described in instructions all belongs to preferred embodiment, and related action and module might not be that the present invention is necessary.
Embodiment tri-:
With reference to Fig. 5, show the structured flowchart of creation apparatus of a kind of video index table of the embodiment of the present invention three, this device specifically can comprise with lower module:
Video acquiring module 501, for obtaining a plurality of user generated content (UGC) videos of client upload and the video ID of each UGC video;
Cluster module 502, for described a plurality of UGC videos are carried out to cluster, and distributes Video clustering ID for each cluster;
Order module 503, generates respectively descriptor for the UGC video for described each cluster, and described descriptor is sorted;
Generation module 504, for adopting the descriptor after described sequence to generate UGC video recommendation list;
The first creation module 505, for adopting described video ID and corresponding Video clustering ID to create first order concordance list;
The second creation module 506, for adopting described Video clustering ID and corresponding UGC video recommendation list to create second level concordance list.
In a preferred embodiment of the present invention, described device can also comprise with lower module:
Abstraction module, for obtain a plurality of UGC videos of client upload and the video ID of each UGC video in described video acquiring module after, extracts respectively the video metadata of each UGC video;
Described video metadata can comprise the information such as video title, video duration, video uploader, video type, video director, video performer.
Duration acquisition module, for after described cluster module is carried out cluster to described a plurality of UGC videos, obtains respectively video duration for the UGC video in described each cluster;
Video removing module, while being positioned at outside default duration interval for the video duration when certain UGC video, will delete the cluster of this UGC video under it;
The 3rd creation module, for adopting described video ID and corresponding video metadata to create third level concordance list.
In a preferred embodiment of the present invention, described cluster module can comprise following submodule:
Calculating sub module, for carrying out similarity calculating to the video metadata of every two UGC videos respectively;
Cluster is determined submodule, for UGC video corresponding to video metadata similarity being positioned within default similarity interval, is defined as belonging to same cluster.
Described order module can comprise following submodule:
Extract submodule, for the UGC video for described each cluster, according to preset rules, from video title, extract respectively video set number;
Information is determined submodule, for the video set number of UGC video being defined as to the descriptor of this UGC video;
Sequence submodule, carries out respectively ascending sort or descending sort for the video set number of the UGC video for described each cluster.
In the embodiment of the present invention, can also, according to the first order concordance list of above-mentioned establishment, second level concordance list and third level concordance list, carry out the recommendation of UGC video.With reference to Fig. 6, show the structured flowchart of recommendation apparatus of a kind of video of the embodiment of the present invention three, this device specifically can comprise with lower module:
Receiver module 601, the user generated content (UGC) video playback request sending for receiving client, described playing request comprises target UGC video ID;
First searches module 602, for the first order concordance list being pre-created, searches Video clustering ID corresponding to described target UGC video ID; Described first order concordance list comprises video ID and corresponding Video clustering ID;
Second searches module 603, for the second level concordance list being pre-created, searches the UGC video recommendation list that described Video clustering ID is corresponding, as target UGC video recommendation list; Described second level concordance list comprises Video clustering ID and corresponding UGC video recommendation list; Wherein, described Video clustering ID is for characterizing the cluster of UGC video, and the descriptor after the UGC video sequence of described UGC video recommendation list in each cluster generates; The descriptor of described UGC video is video set number;
Issue module 604, for described target UGC video recommendation list is issued to described client.
In a preferred embodiment of the present invention, described device can also comprise with lower module:
The 3rd searches module, for issuing before module is issued to described client by described target UGC video recommendation list described, searches video metadata corresponding to described target UGC video ID in the third level concordance list being pre-created; Described third level concordance list comprises video ID and corresponding video metadata, and described video metadata comprises video title;
Extraction module, for extracting video set number according to preset rules from described video title;
Interception module, for intercepting the part UGC video recommendation list before or after described video set number from described target UGC video recommendation list;
The described module that issues, specifically for being issued to described client by described part UGC video recommendation list.
In the embodiment of the present invention, described sequence can be ascending sort or descending sort; Described interception module can comprise following submodule:
The first intercepting submodule for when described sequence is ascending sort, intercepts described video set number part UGC video recommendation list afterwards from described target UGC video recommendation list;
The second intercepting submodule for when described sequence is descending sort, intercepts described video set number part UGC video recommendation list before from described target UGC video recommendation list.
In the embodiment of the present invention, first can carry out cluster to a plurality of UGC videos of client upload, and distribute Video clustering ID for each cluster; Then for the UGC video in each cluster, generate respectively descriptor, and descriptor is sorted, adopt the descriptor after sequence to generate UGC video recommendation list; Finally adopt the video ID of UGC video and corresponding Video clustering ID to create first order concordance list, and, adopt Video clustering ID and corresponding UGC video recommendation list to create second level concordance list.After receiving the request of UGC video playback, can in first order concordance list and second level concordance list, find the target UGC video recommendation list that target UGC video ID is corresponding, and be issued to client.Because video recommendation list has comprised the descriptor of carrying out cluster UGC video afterwards, so these descriptors are more comprehensive, solved the problem that lacks some collection number; And these descriptors are sorted, therefore solved the chaotic problem of order.
For device embodiment, because it is substantially similar to embodiment of the method, so description is fairly simple, relevant part is referring to the part explanation of embodiment of the method.
Each embodiment in this instructions all adopts the mode of going forward one by one to describe, and each embodiment stresses is the difference with other embodiment, between each embodiment identical similar part mutually referring to.
The present invention can describe in the general context of the computer executable instructions of being carried out by computing machine, for example program module.Usually, program module comprises the routine carrying out particular task or realize particular abstract data type, program, object, assembly, data structure etc.Also can in distributed computing environment, put into practice the present invention, in these distributed computing environment, by the teleprocessing equipment being connected by communication network, be executed the task.In distributed computing environment, program module can be arranged in the local and remote computer-readable storage medium that comprises memory device.
Finally, also it should be noted that, in this article, relational terms such as the first and second grades is only used for an entity or operation to separate with another entity or operational zone, and not necessarily requires or imply and between these entities or operation, have the relation of any this reality or sequentially.And, term " comprises ", " comprising " or its any other variant are intended to contain comprising of nonexcludability, thereby the process, method, commodity or the equipment that make to comprise a series of key elements not only comprise those key elements, but also comprise other key elements of clearly not listing, or be also included as the intrinsic key element of this process, method, commodity or equipment.The in the situation that of more restrictions not, the key element being limited by statement " comprising ... ", and be not precluded within process, method, commodity or the equipment that comprises described key element and also have other identical element.
Above to the creation method of a kind of video index table provided by the present invention and device, and a kind of recommend method of video and device, be described in detail, applied specific case herein principle of the present invention and embodiment are set forth, the explanation of above embodiment is just for helping to understand method of the present invention and core concept thereof; , for one of ordinary skill in the art, according to thought of the present invention, all will change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention meanwhile.

Claims (18)

1. a creation method for video index table, is characterized in that, comprising:
Obtain a plurality of user generated content (UGC) videos of client upload and the video ID of each UGC video;
Described a plurality of UGC videos are carried out to cluster, and distribute Video clustering ID for each cluster;
For the UGC video in described each cluster, generate respectively descriptor, and described descriptor is sorted;
Adopt the descriptor after described sequence to generate UGC video recommendation list;
Adopt described video ID and corresponding Video clustering ID to create first order concordance list, and, adopt described Video clustering ID and corresponding UGC video recommendation list to create second level concordance list.
2. method according to claim 1, is characterized in that, after the described step of obtaining a plurality of UGC videos of client upload and the video ID of each UGC video, also comprises:
Extract respectively the video metadata of each UGC video.
3. method according to claim 2, is characterized in that, the described step that described a plurality of UGC videos are carried out to cluster comprises:
Respectively the video metadata of every two UGC videos is carried out to similarity calculating;
UGC video corresponding to video metadata that similarity is positioned within default similarity interval is defined as belonging to same cluster.
4. method according to claim 2, is characterized in that, described video metadata comprises video title; Describedly for the UGC video in described each cluster, generate respectively descriptor, and the step that described descriptor is sorted comprises:
For the UGC video in described each cluster, according to preset rules, from video title, extract respectively video set number;
The video set number of UGC video is defined as to the descriptor of this UGC video;
Video set number for the UGC video in described each cluster carries out respectively ascending sort or descending sort.
5. method according to claim 2, is characterized in that, described video metadata comprises video duration; Described described a plurality of UGC videos are carried out to the step of cluster after, also comprise:
For the UGC video in described each cluster, obtain respectively video duration;
When the video duration of certain UGC video is positioned at outside default duration interval, will the cluster of this UGC video under it, delete.
6. method according to claim 2, is characterized in that, also comprises:
Adopt described video ID and corresponding video metadata to create third level concordance list.
7. a recommend method for video, is characterized in that, comprising:
Receive the user generated content (UGC) video playback request that client sends, described playing request comprises target UGC video ID;
In the first order concordance list being pre-created, search Video clustering ID corresponding to described target UGC video ID; Described first order concordance list comprises video ID and corresponding Video clustering ID;
In the second level concordance list being pre-created, search the UGC video recommendation list that described Video clustering ID is corresponding, as target UGC video recommendation list; Described second level concordance list comprises Video clustering ID and corresponding UGC video recommendation list; Wherein, described Video clustering ID is for characterizing the cluster of UGC video, and the descriptor after the UGC video sequence of described UGC video recommendation list in each cluster generates;
Described target UGC video recommendation list is issued to described client.
8. method according to claim 7, is characterized in that, the descriptor of described UGC video is video set number;
Described described target UGC video recommendation list is issued to the step of described client before, also comprise:
In the third level concordance list being pre-created, search video metadata corresponding to described target UGC video ID; Described third level concordance list comprises video ID and corresponding video metadata, and described video metadata comprises video title;
According to preset rules, from described video title, extract video set number;
From described target UGC video recommendation list, intercept the part UGC video recommendation list before or after described video set number;
The described step that described target UGC video recommendation list is issued to described client comprises: described part UGC video recommendation list is issued to described client.
9. method according to claim 8, is characterized in that, described sequence is ascending sort or descending sort; The described step that intercepts the part UGC video recommendation list before or after described video set number from described target UGC video recommendation list comprises:
If described sequence is ascending sort, from described target UGC video recommendation list, intercept described video set number part UGC video recommendation list afterwards;
If described sequence is descending sort, from described target UGC video recommendation list, intercept described video set number part UGC video recommendation list before.
10. a creation apparatus for video index table, is characterized in that, comprising:
Video acquiring module, for obtaining a plurality of user generated content (UGC) videos of client upload and the video ID of each UGC video;
Cluster module, for described a plurality of UGC videos are carried out to cluster, and distributes Video clustering ID for each cluster;
Order module, generates respectively descriptor for the UGC video for described each cluster, and described descriptor is sorted;
Generation module, for adopting the descriptor after described sequence to generate UGC video recommendation list;
The first creation module, for adopting described video ID and corresponding Video clustering ID to create first order concordance list;
The second creation module, for adopting described Video clustering ID and corresponding UGC video recommendation list to create second level concordance list.
11. devices according to claim 10, is characterized in that, described device also comprises:
Abstraction module, for obtain a plurality of UGC videos of client upload and the video ID of each UGC video in described video acquiring module after, extracts respectively the video metadata of each UGC video.
12. devices according to claim 11, is characterized in that, described cluster module comprises:
Calculating sub module, for carrying out similarity calculating to the video metadata of every two UGC videos respectively;
Cluster is determined submodule, for UGC video corresponding to video metadata similarity being positioned within default similarity interval, is defined as belonging to same cluster.
13. devices according to claim 11, is characterized in that, described order module comprises:
Extract submodule, for the UGC video for described each cluster, according to preset rules, from video title, extract respectively video set number;
Information is determined submodule, for the video set number of UGC video being defined as to the descriptor of this UGC video;
Sequence submodule, carries out respectively ascending sort or descending sort for the video set number of the UGC video for described each cluster.
14. devices according to claim 11, is characterized in that, described video metadata comprises video duration; Described device also comprises:
Duration acquisition module, for after described cluster module is carried out cluster to described a plurality of UGC videos, obtains respectively video duration for the UGC video in described each cluster;
Video removing module, while being positioned at outside default duration interval for the video duration when certain UGC video, will delete the cluster of this UGC video under it.
15. devices according to claim 11, is characterized in that, described device also comprises:
The 3rd creation module, for adopting described video ID and corresponding video metadata to create third level concordance list.
The recommendation apparatus of 16. 1 kinds of videos, is characterized in that, comprising:
Receiver module, the user generated content (UGC) video playback request sending for receiving client, described playing request comprises target UGC video ID;
First searches module, for the first order concordance list being pre-created, searches Video clustering ID corresponding to described target UGC video ID; Described first order concordance list comprises video ID and corresponding Video clustering ID;
Second searches module, for the second level concordance list being pre-created, searches the UGC video recommendation list that described Video clustering ID is corresponding, as target UGC video recommendation list; Described second level concordance list comprises Video clustering ID and corresponding UGC video recommendation list; Wherein, described Video clustering ID is for characterizing the cluster of UGC video, and the descriptor after the UGC video sequence of described UGC video recommendation list in each cluster generates;
Issue module, for described target UGC video recommendation list is issued to described client.
17. devices according to claim 16, is characterized in that, the descriptor of described UGC video is video set number; Described device also comprises:
The 3rd searches module, for issuing before module is issued to described client by described target UGC video recommendation list described, searches video metadata corresponding to described target UGC video ID in the third level concordance list being pre-created; Described third level concordance list comprises video ID and corresponding video metadata, and described video metadata comprises video title;
Extraction module, for extracting video set number according to preset rules from described video title;
Interception module, for intercepting the part UGC video recommendation list before or after described video set number from described target UGC video recommendation list;
The described module that issues, specifically for being issued to described client by described part UGC video recommendation list.
18. devices according to claim 17, is characterized in that, described sequence is ascending sort or descending sort; Described interception module comprises:
The first intercepting submodule for when described sequence is ascending sort, intercepts described video set number part UGC video recommendation list afterwards from described target UGC video recommendation list;
The second intercepting submodule for when described sequence is descending sort, intercepts described video set number part UGC video recommendation list before from described target UGC video recommendation list.
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 true CN104008139A (en) 2014-08-27
CN104008139B 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)

Cited By (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
CN105373600A (en) * 2015-10-30 2016-03-02 北京奇艺世纪科技有限公司 Method and device for sorting video playlists
CN105786857A (en) * 2014-12-24 2016-07-20 Tcl集团股份有限公司 Method and system for improving video aggregation efficiency
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
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
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
CN110633379A (en) * 2019-08-29 2019-12-31 北京睿企信息科技有限公司 System and method for searching images by images based on GPU parallel operation
CN111294636A (en) * 2020-01-21 2020-06-16 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
WO2021175040A1 (en) * 2020-03-02 2021-09-10 Oppo广东移动通信有限公司 Video processing method and related device

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080244681A1 (en) * 2007-03-30 2008-10-02 Gossweiler Richard C Conversion of Portable Program Modules for Constrained Displays
US20090019149A1 (en) * 2005-08-02 2009-01-15 Mobixell Networks Content distribution and tracking
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

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090019149A1 (en) * 2005-08-02 2009-01-15 Mobixell Networks Content distribution and tracking
US20080244681A1 (en) * 2007-03-30 2008-10-02 Gossweiler Richard C Conversion of Portable Program Modules for Constrained Displays
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

Cited By (14)

* 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
CN105786857A (en) * 2014-12-24 2016-07-20 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
CN105373600A (en) * 2015-10-30 2016-03-02 北京奇艺世纪科技有限公司 Method and device for sorting video playlists
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
CN110633379A (en) * 2019-08-29 2019-12-31 北京睿企信息科技有限公司 System and method for searching images by images based on GPU parallel operation
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
CN111294636A (en) * 2020-01-21 2020-06-16 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
WO2021175040A1 (en) * 2020-03-02 2021-09-10 Oppo广东移动通信有限公司 Video processing method and related device

Also Published As

Publication number Publication date
CN104008139B (en) 2017-09-22

Similar Documents

Publication Publication Date Title
CN104008139A (en) Method and device for creating video index table and method and device for recommending video
JP6855595B2 (en) Using machine learning to recommend live stream content
US10311478B2 (en) Recommending content based on user profiles clustered by subscription data
CN101981574B (en) Distributed media fingerprint repositories
CN1716255B (en) Dispersing search engine results by using page category information
CN106326391B (en) Multimedia resource recommendation method and device
US10671687B2 (en) Managing access rights to content using social media
KR20210040891A (en) Method and Apparatus of Recommending Information, Electronic Device, Computer-Readable Recording Medium, and Computer Program
US20170193075A1 (en) System and method for aggregating, classifying and enriching social media posts made by monitored author sources
CN104462573A (en) Method and device for displaying video retrieval results
CN104160712A (en) Computing similarity between media programs
CN104751354B (en) A kind of advertisement crowd screening technique
CN105900117A (en) Method and system for acquisition, normalization, matching, and enrichment of data
US20110320442A1 (en) Systems and Methods for Semantics Based Domain Independent Faceted Navigation Over Documents
CN103207917A (en) Method for marking multimedia content and method and system for generating recommended content
CN108765051B (en) Item recommendation method and device
CN107977678B (en) Method and apparatus for outputting information
CN110046278B (en) Video classification method and device, terminal equipment and storage medium
CN105916032A (en) Video recommendation method and video recommendation terminal equipment
CN107105349A (en) A kind of video recommendation method
CN104021140A (en) Network video processing method and device
Fletcher A quality-based web api selection for mashup development using affinity propagation
Hauger et al. Exploring geospatial music listening patterns in microblog data
Woo et al. Predicting the ratings of Amazon products using Big Data
Hashemzadeh et al. Using a data mining tool and fp-growth algorithm application for extraction of the rules in two different dataset

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