CN104216879A - Video quality excavation system and method - Google Patents
Video quality excavation system and method Download PDFInfo
- Publication number
- CN104216879A CN104216879A CN201310205575.2A CN201310205575A CN104216879A CN 104216879 A CN104216879 A CN 104216879A CN 201310205575 A CN201310205575 A CN 201310205575A CN 104216879 A CN104216879 A CN 104216879A
- Authority
- CN
- China
- Prior art keywords
- video
- weight
- classification
- module
- attribute
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24317—Piecewise classification, i.e. whereby each classification requires several discriminant rules
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a video quality excavation system and method. The system comprises a video source module, a classification module, a weight calculation module and a ranking list candidate set generation module. The video source module contains multiple videos which are used for video excavation and come from multiple different video sources. The classification module classifies the videos of all the video sources in the video source module according to the video attributes, and combining the classification results of the different video sources. The weight calculation module obtains multiple weights for all the classified videos according to different video attributes, and obtaining the final weights of all the videos through the obtained weights according to the weight calculation formula. The ranking list candidate set generation module is used for generating various classifications of ranking list candidate sets through all the videos in each classification according to the weights of the videos. By means of the video quality excavation system and method, high-quality videos can be objectively excavated, and quality of the excavated videos can be ensured.
Description
Technical field
The present invention, about a kind of network video system and method, particularly relates to video quality digging system and the method for high-quality video in a kind of UGC video website.
Background technology
UGC full name is User Generated Content, the namely meaning of user-generated content.Different from conventional video website, in UGC video website, video has following features:
1) short-sighted frequency is main
2) video is uploaded by user, and video quality is uneven
3) quantity is large, video updating decision.
These features, make robotization find and excavate the high-quality video in website, becoming difficulty.
At present, the video quality for UGC video website excavates, and a kind of simple solution, is that the advantage of the method is by manually screening, but can ensures the quality of screened video to a certain extent. and shortcoming is also apparent:
1) labor intensive material resources, due to video updating decision, need editor constantly to go to hold current focus.
2) subjective degree is high, and this comprises: a, selects the quality of video, depends on the experience .b with editor, during selecting video, and the inevitably tendency of a guy.
For the another kind of method that UGC video website video quality excavates, the data placed one's entire reliance upon as statistical significances such as video playback number, comment numbers, common are as: one week play maximum ranking list, today comments on ranking list etc., these statisticss, to a certain extent, can reflect the popular degree of video really, playing the video that number is higher, is also generally the video that current hot topic is play really.But also there are the following problems:
1) video playback number affects by spread and demonstrate..Such as, certain video appears at Baidu's homepage, or is forwarded by certain famous person in community, then video is more easily exposed, thus has and play number in a large number. but these videos, quality might not be got well, and user may be by " deceive " click and enter viewing video, and non-user is really liked.
Citing:
User wants on website, look for the film just shown, but due to copyright reasons, community website does not have full version film, only has titbit or propaganda film. and user " is deceived " by search engine and is clicked and entered, and really is liked.
2) video quality is difficult to ensure.Some title party videos, often have considerable click, but these videos, also should not classify as high-quality video.
Summary of the invention
For overcoming the deficiency that above-mentioned prior art exists, the object of the present invention is to provide a kind of video quality digging system and method, by the video in different video source being carried out respectively classify and merge, according to video attribute, weight is obtained to each video, the ranking list candidate collection of each classification is obtained according to weight, not only can excavate high-quality video comparatively objectively, and the quality of excavated video can be ensured.
For reaching above-mentioned and other object, the present invention proposes a kind of video quality digging system, at least comprises:
Video source module, comprises the multiple videos from multiple different video source for Video Mining;
Classification die set, classifies respectively according to video attribute to the video of video source each in this video source module, and is merged by the classification results in different video source;
Weight calculation module, to sorted each video, obtains multiple weight according to different video attributes, and obtained multiple weights are obtained the final weight of each video according to weight calculation formula;
Ranking list candidate collection generates module, then for each video in each classification, the weight according to video generates the ranking list candidate collection of each classification.
Further, this system also comprises memory module, the ranking list candidate collection of each classification is stored in database.
Further, this video source module comprises the video source that the video source of real-time statistics broadcasting number acquisition, the video source of human-edited's recommendation and important VCU upload.
Further, this real-time statistics is play the time in the video source foundation video attribute of number acquisition and is play number by the visual classification in video source by this classification die set.
Further, this classification die set video source that human-edited is recommended according to the video location in video attribute by the visual classification in video source.
Further, this classification die set video source that this important VCU is uploaded according to the uplink time in video attribute by the visual classification in video source.
Further, this weight calculation formula is:
Rank＝w1*R1+w2*R2+w3*R3
Wherein, Rank is that this weight calculation module calculates gained weight, w1, w2, w3 is empirical parameter, according to the broadcasting number of statistics video, R1 show that the weight of video, R2 are that the position calculation appearing at page breakage according to video draws video weight R2, R3 is the significance level according to this VCU, gives the weight of this VCU uploaded videos.
Further, this video quality digging system also comprises the module that reorders, the attribute that this module that reorders newly obtains according to video for the video in each classification ranking list candidate collection and primary attribute recalculate weight, and reorder according to the weight after recalculating, form the ranking list candidate collection of new each classification.
Further, this module that reorders comprises click feedback module, duplicate removal module, blacklist filtering module, play number total amount acquisition module and weight re-computation module, this click feedback module for excavate the previous day user to broadcast page recommend digital video click situation, this duplicate removal module is for judging whether the video with repetition, this blacklist filtering module is used for filtering black list and is stored in the illegal video of state in list of videos, this broadcasting number total amount acquisition module in conjunction with video novel degrees and play number total amount, adjustment video weight, this weight re-computation module is according to the weight clicking the video click feedback that feedback module obtains, according to the penalty term that duplicate removal module calculates, and the attribute weight of video is calculated according to the primary attribute of video, re-computation formula is utilized to obtain the new weighted value of each video, and according to new weighted value, the video in each classification ranking list candidate collection is reordered, form the ranking list candidate collection of new each classification.
Further, this re-computation formula is:
V＝R+v1+v2+v3
Wherein, V is new weighted value, R is the weighted value that this power kind calculates module acquisition, the weight that the V1 number of times clicked according to video that be this click feedback module obtains, V2 is this duplicate removal module is the penalty term that palinopsia frequency meter calculates to video, and V3 is the attribute weight of the video gone out according to video elementary property calculation.
For reaching above-mentioned and other object, the present invention also provides a kind of video quality method for digging, comprises the steps:
Step one, classifies respectively according to video attribute to the video in different video source, and is merged by the classification results in different video source;
Step 2, to sorted each video, obtains multiple weight according to different video attributes, and obtains the final weight of each video according to weight calculation formula;
Step 3, for each video in each classification, the weight according to video generates the ranking list candidate collection of each classification.
Further, after step 3, also comprise the step ranking list candidate collection of each classification be stored in database.
Further, this video source comprises the video source that the video source of real-time statistics broadcasting number acquisition, the video source of human-edited's recommendation and important VCU upload.
Further, real-time statistics is play to the video source of number acquisition, the video attribute of classification foundation is time and broadcasting number; For the video source that human-edited recommends, the video attribute of classification foundation is video location; For the video source that important VCU uploads, the video attribute of classification foundation is uplink time.
Further, in step one, the result that classification merges comprises the fastest video of rising, popular video and selected video.
Further, step 2 also comprises the steps:
According to the broadcasting number of the video of statistics, obtain the weight R1 of video;
According to the position of the appearance of video, calculate video weight R2;
According to the significance level of VCU, give the weight R3 of this VCU uploaded videos;
Exploitation right re-computation formula Rank=w1*R1+w2*R2+w3*R3 obtains the final weight of each video, and wherein w1, w2, w3 are empirical parameter.
Further, in step 2, according to clicking rate situation on line, the weight of each video is regulated.
Further, also comprise the steps: after this step 3
The attribute newly obtained for the video foundation video in each classification ranking list candidate collection and primitive attribute recalculate weight, and reorder according to the weight after recalculating, and form the ranking list candidate collection of new each classification.
Further, the step that reorders comprises:
Obtain the clicked number of times of this video according to click feedback module, show that video clicks the weight v1 of feedback;
According to the judged result of duplicate removal module, calculate penalty term v2;
According to the primary attribute of video, calculate the attribute weight v3 of video;
Obtain the new weighted value of each video according to re-computation formula V=R+v1+v2+v3, wherein R is the weighted value obtained in step 2;
Reorder according to the ranking list candidate collection of weight to each classification after recalculating, form the ranking list candidate collection of new each classification.
Further, this primary attribute comprises video title length, uplink time, video duration, readability.
Compared with prior art, a kind of video quality digging system of the present invention and method are by carrying out respectively classifying and merging by the video in different video source, according to video attribute, weight is obtained to each video, the ranking list candidate collection of each classification is obtained according to weight, not only can excavate high-quality video comparatively objectively, and the quality of excavated video can be ensured.
Accompanying drawing explanation
Fig. 1 is the system architecture diagram of a kind of video quality digging system of the present invention;
Fig. 2 is the flow chart of steps of a kind of video quality method for digging of the present invention;
Fig. 3, Fig. 4 and Fig. 5 are the video quality Result schematic diagram of the preferred embodiment of the present invention.
Embodiment
Below by way of specific instantiation and accompanying drawings embodiments of the present invention, those skilled in the art can understand other advantage of the present invention and effect easily by content disclosed in the present specification.The present invention is also implemented by other different instantiation or is applied, and the every details in this instructions also can based on different viewpoints and application, carries out various modification and change not deviating under spirit of the present invention.
Fig. 1 is the system architecture diagram of a kind of video quality digging system of the present invention.As shown in Figure 1, a kind of video quality digging system of the present invention, excavate for the high-quality video in UGC video website, at least comprise: video source module 101, classification die set 102, weight calculation module 103, ranking list candidate collection generate module 104 and memory module 105.
Wherein, video source module 101 comprises the multiple videos from multiple different video source for Video Mining, in present pre-ferred embodiments, video source module 101 has following three kinds of video source: 1, the video playback number of each video of real-time statistics, and playing number more than the video of a preset value is the video source that real-time statistics obtains; 2, the video recommended of human-edited, comprises page breakage and column, i.e. each classification of arranging out of web editor, the video of each column, and the quality of this kind of video source is general higher; 3, important VCU (Value Creating User/Unit, user/the mechanism of the creation of value) video uploaded, to for a long time, the video that the VCU that website and community are approved uploads, run-of-the-mill is all guaranteed. and these videos are sources for high-quality video.
In classification die set 102 pairs of video source module 101, the video of each video source is classified respectively according to video attribute, and is merged by the classification results in different video source.Specifically, real-time statistics is play to the video source of the acquisition of number, classification die set 102 is according to the time in video attribute and play number by the visual classification in this video source, in present pre-ferred embodiments, be divided three classes: rise the fastest, popular video and selected video, for example, classification die set 102 was by nearest 5 hours, the video that hour level plays number > 20 is classified as the fastest video of rising, by nearest 24 hours, the video that hour level plays number > 50 classifies as popular video, by nearest 10 days. the video playing number > 200 every day is classified as selected video, for the video that human-edited recommends, classification die set 102 is classified according to the video location in video attribute, such as, current page fragment video is classified as the fastest video of rising by classification die set 102, current column video is classified as popular video, history column video is classified as selected video, for the video that important VCU uploads, classification die set 102 is classified according to the uplink time in video attribute, such as, nearest 1 day uploaded videos is classified as the fastest video of rising by classification die set 102, the nearest video uploaded for 2 days is classified as popular video, the nearest video uploaded for 10 days is classified as selected video, can be as shown in table 1 to the classification results of three kinds of video source, the classification results in different video source merges by classification results 102, forms the classification results total to video in video source module 101.
Table 1
Weight calculation module 103, to sorted each video, obtains multiple weight according to different video attributes, and obtains the final weight of each video according to weight calculation formula.Specifically, weight calculation module 103 is according to the broadcasting number of the video of statistics, obtain the weight R1 of video, according to the position of the appearance of video, as the position of page breakage, calculate video weight R2, according to the significance level of VCU, give the weight R3 of this VCU uploaded videos, then by weight calculation formula Rank=w1*R1+w2*R2+w3*R3 wherein w1, w2, w3 is empirical parameter, obtains the final weight of each video, more preferably, weight calculation module 103 also can click situation according on line, does suitable adjustment to the final weight of each video.Ranking list candidate collection generates module 104 for each video in each classification, and the weight according to video generates the ranking list candidate collection of each classification; The ranking list candidate collection of each classification is stored in database by memory module 105, in present pre-ferred embodiments, the ranking list candidate collection of each classification can be stored in the Key-Value database Redis of the non-relational database (NoSQL) of increasing income.
Preferably, the video quality digging system of the present invention also comprises the module that reorders, the attribute that the module that reorders newly obtains according to video for the video in each classification ranking list candidate collection and primary attribute recalculate weight, and reorder according to the weight after recalculating, form the ranking list candidate collection of new each classification.The module that reorders comprises clicks feedback module 107, duplicate removal module 108, blacklist filtering module 109, broadcasting number total amount acquisition module 110 and weight re-computation module 111, click feedback module 107 for excavate the previous day user to broadcast page recommend digital video click situation, due to the click in recommendation position, representative of consumer that can be clearer and more definite is to the interest level of video, therefore, click more videos, better weight should be able to be obtained, duplicate removal module 108 is for judging whether the video with repetition, and duplicate removal is completed by following two steps: title duplicate removal, by video interface (video_cluster interface), obtain belonging to video and gather together (cluster), if two videos belong to one gather together (cluster), then only retain the video blacklist filtering module 109 of higher weights value, for filtering black list and be stored in state (status) illegal video in list of videos (the video table in mysql database), play number total amount acquisition module 110, in conjunction with novel degrees and the broadcasting number total amount of video, adjustment video weight, such as, rising in the fastest video, not only needing the temperature calculating current hour level video, also should consider video novel degrees. to hotter in the recent period, but the video that total playback volume is very high, should fall power, to find more new videos, weight re-computation module 111 obtains according to clicking feedback module 107 the weight v1 that video clicks feedback, penalty term v2 is calculated according to duplicate removal module 108, and according to the primary attribute of video, as uplink time, video duration, readability, calculate the attribute weight v3 of video, the new weighted value of each video is obtained here according to re-computation formula V=R+v1+v2+v3 (R is that weight calculation module 103 calculates acquisition weight), and according to new weighted value, the video in each classification ranking list candidate collection is reordered, form the ranking list candidate collection of new each classification.
Fig. 2 is the flow chart of steps of a kind of video quality method for digging of the present invention.As shown in Figure 2, a kind of video quality method for digging of the present invention, comprises the steps:
Step 201, classifies respectively according to video attribute to the video in different video source, and is merged by the classification results in different video source.In present pre-ferred embodiments, video source has three kinds: 1, the video playback number of each video of real-time statistics, and playing number more than the video of a preset value is the video source that real-time statistics obtains; 2, the video recommended of human-edited, comprises page breakage and column, i.e. each classification of arranging out of web editor, the video of each column, and the quality of this kind of video source is general higher; 3, the video uploaded of important VCU (Value Creating User/Unit, the user/mechanism of the creation of value).Different video source is different for the video attribute carrying out classifying, such as, for the first video source, the video attribute of classification foundation can be time and broadcasting number, by nearest 5 hours, the video that hour level plays number > 20 is classified as the fastest video of rising, by nearest 24 hours, the video that hour level plays number > 50 classifies as popular video, by nearest 10 days, the video playing number > 200 every day is classified as selected video, for the second video source, the video attribute of classification foundation can be video location, that is: by nearest 5 hours, the video that hour level plays number > 20 is classified as the fastest video of rising, by nearest 24 hours, the video that hour level plays number > 50 classifies as popular video, by nearest 10 days. the video playing number > 200 every day is classified as selected video, for the third video source, the video attribute of classification foundation can be then uplink time, such as, nearest 1 day uploaded videos is classified as the fastest video of rising, the nearest video uploaded for 2 days is classified as popular video, the nearest video uploaded for 10 days is classified as selected video.
Step 202, to sorted each video, obtains multiple weight according to different video attributes, and obtains the final weight of each video according to weight calculation formula.Specifically, step 202 can comprise the steps: further
According to the broadcasting number of the video of statistics, obtain the weight R1 of video;
According to the position of the appearance of video, as the position of page breakage, calculate video weight R2;
According to the significance level of VCU, give the weight R3 of this VCU uploaded videos;
Finally, exploitation right re-computation formula Rank=w1*R1+w2*R2+w3*R3, wherein w1, w2, w3 are empirical parameter, obtain the final weight of each video.
Certainly, step 202 according to clicking rate situation on line, can also do suitable adjustment to the weight of each video.
Step 203, for each video in each classification, the weight according to video generates the ranking list candidate collection of each classification.
Step 204, the ranking list candidate collection of each classification is stored in database, in present pre-ferred embodiments, the ranking list candidate collection of each classification can be stored in the Key-Value database Redis of the non-relational database (No SQL) of increasing income.
Preferably, after step 203, the video quality method for digging of the present invention can also comprise the steps:
The attribute newly obtained for the video foundation video in each classification ranking list candidate collection and primitive attribute recalculate weight, and reorder according to the weight after recalculating, and form the ranking list candidate collection of new each classification.Specifically, the step that reorders then comprises the steps:
Obtain the clicked number of times of this video according to click feedback module, show that video clicks the weight v1 of feedback;
According to the judged result of duplicate removal module, detecting that as crossed this video is palinopsia frequency, calculating penalty term v2 (v2 is negative value);
According to the primary attribute of video, as video title length, uplink time, video duration, readability etc., calculate the attribute weight v3 of video;
The new weighted value of each video is obtained here according to re-computation formula V=R+v1+v2+v3 (R is that step 202 calculates the weight obtained);
Reorder according to the ranking list candidate collection of weight to each classification after recalculating, form the ranking list candidate collection of new each classification.
Fig. 3, Fig. 4 and Fig. 5 are the video quality Result schematic diagram of the preferred embodiment of the present invention.Visible, by the present invention, high-quality video can be excavated more objectively, user is provided high-quality video.
In sum, a kind of video quality digging system of the present invention and method are by carrying out respectively classifying and merging by the video in different video source, according to video attribute, weight is obtained to each video, the ranking list candidate collection of each classification is obtained according to weight, not only can excavate high-quality video comparatively objectively, and the quality of excavated video can be ensured.
Above-described embodiment is illustrative principle of the present invention and effect thereof only, but not for limiting the present invention.Any those skilled in the art all without prejudice under spirit of the present invention and category, can carry out modifying to above-described embodiment and change.Therefore, the scope of the present invention, should listed by claims.
Claims (20)
1. a video quality digging system, at least comprises:
Video source module, comprises the multiple videos from multiple different video source for Video Mining;
Classification die set, classifies respectively according to video attribute to the video of video source each in this video source module, and is merged by the classification results in different video source;
Weight calculation module, to sorted each video, obtains multiple weight according to different video attributes, and obtained multiple weights are obtained the final weight of each video according to weight calculation formula;
Ranking list candidate collection generates module, then for each video in each classification, the weight according to video generates the ranking list candidate collection of each classification.
2. video quality digging system as claimed in claim 1, is characterized in that: this system also comprises memory module, the ranking list candidate collection of each classification is stored in database.
3. video quality digging system as claimed in claim 1, is characterized in that: this video source module comprises the video source that the video source of real-time statistics broadcasting number acquisition, the video source of human-edited's recommendation and important VCU upload.
4. video quality digging system as claimed in claim 3, is characterized in that: this real-time statistics is play the time in the video source foundation video attribute of number acquisition and play number by the visual classification in video source by this classification die set.
5. video quality digging system as claimed in claim 4, is characterized in that: the video location in the video source foundation video attribute that human-edited recommends by this classification die set is by the visual classification in video source.
6. video quality digging system as claimed in claim 5, is characterized in that: the uplink time in the video source foundation video attribute that this important VCU uploads by this classification die set is by the visual classification in video source.
7. video quality digging system as claimed in claim 6, it is characterized in that, this weight calculation formula is:
Rank＝w1*R1+w2*R2+w3*R3
Wherein, Rank is that this weight calculation module calculates gained weight, w1, w2, w3 is empirical parameter, according to the broadcasting number of statistics video, R1 show that the weight of video, R2 are that the position calculation appearing at page breakage according to video draws video weight R2, R3 is the significance level according to this VCU, gives the weight of this VCU uploaded videos.
8. video quality digging system as claimed in claim 1, it is characterized in that: this video quality digging system also comprises the module that reorders, the attribute that this module that reorders newly obtains according to video for the video in each classification ranking list candidate collection and primary attribute recalculate weight, and reorder according to the weight after recalculating, form the ranking list candidate collection of new each classification.
9. video quality digging system as claimed in claim 8, is characterized in that: this module that reorders comprises clicks feedback module, duplicate removal module, blacklist filtering module, broadcasting number total amount acquisition module and weight re-computation module, wherein,
This click feedback module for excavate the previous day user to broadcast page recommend digital video click situation;
This duplicate removal module is for judging whether the video with repetition;
This blacklist filtering module is used for filtering black list and is stored in the illegal video of state in list of videos;
This broadcasting number total amount acquisition module in conjunction with video novel degrees and play number total amount, adjustment video weight;
This weight re-computation module is according to the weight clicking the video click feedback that feedback module obtains, according to the penalty term that duplicate removal module calculates, and the attribute weight of video is calculated according to the primary attribute of video, re-computation formula is utilized to obtain the new weighted value of each video, and according to new weighted value, the video in each classification ranking list candidate collection is reordered, form the ranking list candidate collection of new each classification.
10. video quality digging system as claimed in claim 8, it is characterized in that, this re-computation formula is:
V＝R+v1+v2+v3
Wherein, V is new weighted value, R is the weighted value that this power kind calculates module acquisition, the weight that the V1 number of times clicked according to video that be this click feedback module obtains, V2 is this duplicate removal module is the penalty term that palinopsia frequency meter calculates to video, and V3 is the attribute weight of the video gone out according to video elementary property calculation.
11. 1 kinds of video quality method for digging, comprise the steps:
Step one, classifies respectively according to video attribute to the video in different video source, and is merged by the classification results in different video source;
Step 2, to sorted each video, obtains multiple weight according to different video attributes, and obtains the final weight of each video according to weight calculation formula;
Step 3, for each video in each classification, the weight according to video generates the ranking list candidate collection of each classification.
12. a kind of video quality method for digging as claimed in claim 11, is characterized in that: after step 3, also comprise the step ranking list candidate collection of each classification be stored in database.
13. a kind of video quality method for digging as claimed in claim 11, is characterized in that: this video source comprises the video source that the video source of real-time statistics broadcasting number acquisition, the video source of human-edited's recommendation and important VCU upload.
14. a kind of video quality method for digging as claimed in claim 13, is characterized in that: video source real-time statistics being play to number acquisition, and the video attribute of classification foundation is time and broadcasting number; For the video source that human-edited recommends, the video attribute of classification foundation is video location; For the video source that important VCU uploads, the video attribute of classification foundation is uplink time.
15. a kind of video quality method for digging as claimed in claim 14, is characterized in that: in step one, and the result that classification merges comprises the fastest video of rising, popular video and selected video.
16. a kind of video quality method for digging as claimed in claim 14, it is characterized in that, step 2 also comprises the steps:
According to the broadcasting number of the video of statistics, obtain the weight R1 of video;
According to the position of the appearance of video, calculate video weight R2;
According to the significance level of VCU, give the weight R3 of this VCU uploaded videos;
Exploitation right re-computation formula Rank=w1*R1+w2*R2+w3*R3 obtains the final weight of each video, and wherein w1, w2, w3 are empirical parameter.
17. a kind of video quality method for digging as claimed in claim 16, is characterized in that: in step 2, according to clicking rate situation on line, regulate the weight of each video.
18. a kind of video quality method for digging as claimed in claim 11, is characterized in that, also comprise the steps: after this step 3
The attribute newly obtained for the video foundation video in each classification ranking list candidate collection and primitive attribute recalculate weight, and reorder according to the weight after recalculating, and form the ranking list candidate collection of new each classification.
19. a kind of video quality method for digging as claimed in claim 18, it is characterized in that, the step that reorders comprises:
Obtain the clicked number of times of this video according to click feedback module, show that video clicks the weight v1 of feedback;
According to the judged result of duplicate removal module, calculate penalty term v2;
According to the primary attribute of video, calculate the attribute weight v3 of video;
Obtain the new weighted value of each video according to re-computation formula V=R+v1+v2+v3, wherein R is the weighted value obtained in step 2;
Reorder according to the ranking list candidate collection of weight to each classification after recalculating, form the ranking list candidate collection of new each classification.
20. a kind of video quality method for digging as claimed in claim 19, is characterized in that: this primary attribute comprises video title length, uplink time, video duration, readability.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310205575.2A CN104216879A (en) | 2013-05-29 | 2013-05-29 | Video quality excavation system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310205575.2A CN104216879A (en) | 2013-05-29 | 2013-05-29 | Video quality excavation system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104216879A true CN104216879A (en) | 2014-12-17 |
Family
ID=52098385
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310205575.2A Pending CN104216879A (en) | 2013-05-29 | 2013-05-29 | Video quality excavation system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104216879A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| WO2016095628A1 (en) * | 2014-12-18 | 2016-06-23 | 网宿科技股份有限公司 | Video terminal and video play restricting method and system thereof |
| CN106339493A (en) * | 2016-08-31 | 2017-01-18 | 腾讯科技（深圳）有限公司 | Data processing method and related device |
| CN106682126A (en) * | 2016-12-14 | 2017-05-17 | 河海大学 | Subject data set filtering and ordering method and system based on total data quality |
| WO2017101299A1 (en) * | 2015-12-15 | 2017-06-22 | 乐视控股（北京）有限公司 | Method, device, and equipment for video recommendation |
| CN106919651A (en) * | 2017-01-22 | 2017-07-04 | 北京奇艺世纪科技有限公司 | The search ordering method and device of external website video |
| CN107493467A (en) * | 2017-07-06 | 2017-12-19 | 北京奇艺世纪科技有限公司 | A kind of video quality evaluation method and device |
| CN107786898A (en) * | 2017-09-28 | 2018-03-09 | 南京林洋电力科技有限公司 | A kind of electric power wisdom business hall media intelligent player method with compensation mechanism |
| CN108734491A (en) * | 2017-04-20 | 2018-11-02 | 合信息技术（北京）有限公司 | Assess the method and device of the copyright value of multi-medium data |
| CN109829165A (en) * | 2019-02-11 | 2019-05-31 | 杭州乾博科技有限公司 | One kind is from media article Valuation Method and system |
| CN109889891A (en) * | 2019-03-05 | 2019-06-14 | 腾讯科技（深圳）有限公司 | Obtain the method, apparatus and storage medium of target media file |
| CN109922334A (en) * | 2017-12-13 | 2019-06-21 | 优酷信息技术（北京）有限公司 | A kind of recognition methods and system of video quality |
| CN110121110A (en) * | 2019-05-07 | 2019-08-13 | 北京奇艺世纪科技有限公司 | Video quality evaluation method, equipment, video processing equipment and medium |
| CN111597384A (en) * | 2020-04-28 | 2020-08-28 | 深圳市商汤科技有限公司 | Video source management method and related device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2829484A1 (en) * | 2011-03-08 | 2012-09-13 | Tivo Inc. | Multi source and destination media discovery and management platform |
| CN102780916A (en) * | 2012-04-12 | 2012-11-14 | 天脉聚源(北京)传媒科技有限公司 | Method for aggregating and distributing live video streams |
-
2013
- 2013-05-29 CN CN201310205575.2A patent/CN104216879A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2829484A1 (en) * | 2011-03-08 | 2012-09-13 | Tivo Inc. | Multi source and destination media discovery and management platform |
| CN102780916A (en) * | 2012-04-12 | 2012-11-14 | 天脉聚源(北京)传媒科技有限公司 | Method for aggregating and distributing live video streams |
Non-Patent Citations (1)
| Title |
|---|
| 苏青章: "社会网络多策略视频推荐系统研究与设计", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016095628A1 (en) * | 2014-12-18 | 2016-06-23 | 网宿科技股份有限公司 | Video terminal and video play restricting method and system thereof |
| 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 |
| 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 |
| WO2017101299A1 (en) * | 2015-12-15 | 2017-06-22 | 乐视控股（北京）有限公司 | Method, device, and equipment for video recommendation |
| CN106339493A (en) * | 2016-08-31 | 2017-01-18 | 腾讯科技（深圳）有限公司 | Data processing method and related device |
| CN106682126A (en) * | 2016-12-14 | 2017-05-17 | 河海大学 | Subject data set filtering and ordering method and system based on total data quality |
| CN106682126B (en) * | 2016-12-14 | 2020-09-25 | 河海大学 | Topic data set filtering and sorting method and system based on overall data quality |
| CN106919651A (en) * | 2017-01-22 | 2017-07-04 | 北京奇艺世纪科技有限公司 | The search ordering method and device of external website video |
| CN108734491A (en) * | 2017-04-20 | 2018-11-02 | 合信息技术（北京）有限公司 | Assess the method and device of the copyright value of multi-medium data |
| CN108734491B (en) * | 2017-04-20 | 2023-04-07 | 阿里巴巴（中国）有限公司 | Method and device for evaluating copyright value of multimedia data |
| CN107493467A (en) * | 2017-07-06 | 2017-12-19 | 北京奇艺世纪科技有限公司 | A kind of video quality evaluation method and device |
| CN107786898A (en) * | 2017-09-28 | 2018-03-09 | 南京林洋电力科技有限公司 | A kind of electric power wisdom business hall media intelligent player method with compensation mechanism |
| CN109922334A (en) * | 2017-12-13 | 2019-06-21 | 优酷信息技术（北京）有限公司 | A kind of recognition methods and system of video quality |
| CN109922334B (en) * | 2017-12-13 | 2021-11-19 | 阿里巴巴（中国）有限公司 | Video quality identification method and system |
| CN109829165A (en) * | 2019-02-11 | 2019-05-31 | 杭州乾博科技有限公司 | One kind is from media article Valuation Method and system |
| CN109889891A (en) * | 2019-03-05 | 2019-06-14 | 腾讯科技（深圳）有限公司 | Obtain the method, apparatus and storage medium of target media file |
| CN110121110A (en) * | 2019-05-07 | 2019-08-13 | 北京奇艺世纪科技有限公司 | Video quality evaluation method, equipment, video processing equipment and medium |
| CN111597384A (en) * | 2020-04-28 | 2020-08-28 | 深圳市商汤科技有限公司 | Video source management method and related device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104216879A (en) | Video quality excavation system and method | |
| CN108509534B (en) | Personalized music recommendation system based on deep learning and implementation method thereof | |
| CN104216883A (en) | Video recommendation reason generating system and method | |
| US9280565B1 (en) | Systems, methods, and computer program products for displaying images | |
| CN102780920A (en) | Television program recommending method and system | |
| CN103747343B (en) | The method and apparatus that resource is recommended at times | |
| CN105915949A (en) | Video content recommending method, device and system | |
| TW201342088A (en) | Digital content reordering method and digital content aggregator | |
| US20180189668A1 (en) | System and Method for Measuring and Predicting Content Dissemination in Social Networks | |
| CN105224576A (en) | A kind of video display intelligent recommendation method | |
| CN104053023B (en) | A kind of method and device of determining video similarity | |
| JP2013033375A (en) | Information processing apparatus, information processing method, and program | |
| CN103686382A (en) | Program recommendation method | |
| WO2021238084A1 (en) | Voice packet recommendation method, apparatus and device, and storage medium | |
| CN104408077A (en) | Method and system for displaying pictures and terminal | |
| Kosterich | Reconfiguring the “hits”: The new portrait of television program success in an era of big data | |
| CN104699696A (en) | File recommendation method and device | |
| CN103699694A (en) | Data processing method and device | |
| CN104102733A (en) | Search content providing method and search engine | |
| CN107493467A (en) | A kind of video quality evaluation method and device | |
| CN102890950A (en) | Media automatic editing device and method, and media broadcasting method and media broadcasting system | |
| CN102957949A (en) | Device and method for recommending video to user | |
| CN105635749A (en) | Method and device for generating video frame set | |
| Matthews | 15 Journalists and their sources | |
| Zhou et al. | Interpreting video recommendation mechanisms by mining view count traces |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20180605 Address after: 201203 7, 1 Lane 666 lane, Zhang Heng Road, Pudong New Area, Shanghai. Applicant after: SHANGHAI ZHANGMEN TECHNOLOGY CO., LTD. Address before: 300467 Tianjin Binhai New Area Tianjin eco city animation road 126 anime building B1 area two layer 201-243 Applicant before: Cool Sheng (Tianjin) Technology Co., Ltd. |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20180801 Address after: 300450 Tianjin seaside new area Tianjin eco city animation road 126 anime building B1 area two layer 201-243 Applicant after: Cool Sheng (Tianjin) Technology Co., Ltd. Address before: 201203 7, 1 Lane 666 lane, Zhang Heng Road, Pudong New Area, Shanghai. Applicant before: SHANGHAI ZHANGMEN TECHNOLOGY CO., LTD. |
|
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20181219 Address after: 201306 N2025 room 24, 2 New Town Road, mud town, Pudong New Area, Shanghai Applicant after: Shanghai Lian Shang network technology Co., Ltd Address before: 300450 Tianjin Binhai New Area Tianjin eco city animation road 126 anime building B1 area two layer 201-243 Applicant before: Cool Sheng (Tianjin) Technology Co., Ltd. |
|
| TA01 | Transfer of patent application right | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20141217 |
|
| RJ01 | Rejection of invention patent application after publication |