Target-video specific co-watched video clusters

• US 9,152,708 B1
• Filed: 12/14/2009
• Issued: 10/06/2015
• Est. Priority Date: 12/14/2009
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: He, Yu, Stoutamire, David P., Chen, Zhihui, He, Hao
• Primary Examiner(s): Allen, Brittany N
• Application Number:
  US12/637,635
• Time in Patent Office: 2,122 Days
• Field of Search: 707/738
• US Class Current: 1/1
• CPC Class Codes: G06F 16/70 of video data
• CPC Class Codes: G06F 16/7867 using information manually generated, e.g. tags, keywords, comments, title and artist information, manually generated time, location and usage information, user ratings

Abstract

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for clustering videos and using video clusters to present content. One method includes generating target video-specific clusters of co-watched videos for target videos, according to keywords for each target video, and storing data associating the target video with its clusters. The clusters can then be used in various applications, including identifying related videos, augmenting video search results, re-ordering video search results, and identifying content for a target video.

Claims

1. A system comprising:

one or more computers programmed to perform operations comprising;

for each of a plurality of target videos, wherein each target video was viewed by a user in response to a particular query submitted by the user, identifying a respective plurality of co-watched videos, wherein each respective co-watched video for the target video was viewed by a respective user within a time window of when the same user viewed the target video;

identifying a respective first query for each respective co-watched video wherein the respective co-watched video was selected by users more often when identified in a search result responsive to the respective first query in a plurality of queries submitted by users than when identified in search results responsive to other queries in the plurality of queries submitted by users, and wherein the selections responsive to the plurality of queries are counted only when the length of time the video was watched exceeds a threshold time, and wherein the first respective query for the respective co-watched video is different from the particular query of the target video for which the respective co-watched video was identified;

deriving, from each respective first query for each respective co-watched video, one or more respective keywords to be associated with the respective co-watched video;

for each target video in the plurality of target videos, using the one or more respective keywords for each co-watched video to generate one or more target video-specific clusters of co-watched videos for the target video, and wherein a first video from each target video-specific cluster is a co-watched video for the target video, the co-watched video for the target video having first keywords including the respective first query, the first keywords being in common with second keywords from at least one second video from the cluster, the first keywords being different from the particular query; and

storing data identifying the clusters for each target video, wherein the data indexes the clusters for each target video according to the corresponding target video.

2. The system of claim 1, further operable to perform operations comprising selecting content related to a first target video according to the clusters for the first target video.

3. The system of claim 2, wherein the content is a first co-watched video selected from one of the clusters for the first target video.

4. The system of claim 1, further operable to perform operations comprising selecting a plurality of related videos for a first target video, the selecting comprising:

identifying the clusters for the first target video from the stored data; and

selecting a co-watched video from each of the clusters.

5. The system of claim 1, further operable to perform operations comprising, for each target video in the plurality of target videos:

determining a score for each cluster generated for the target video; and

identifying one or more clusters as being off-topic for the target video according to the scores.

6. The system of claim 1, wherein each co-watched cluster is associated with one or more keywords of the respective co-watched videos that belong to the cluster, the system further operable to perform operations comprising:

receiving input requesting presentation of a first target video;

identifying content to present with the first target video according to one or more of the keywords associated with the clusters for the first target video; and

providing the first target video along with the content.

7. The system of claim 6, further operable to perform operations comprising:

receiving data identifying a set of initial videos responsive to a query;

selecting a cluster of co-watched videos for an initial target video selected from the initial videos, wherein the cluster of co-watched videos is selected according to how well the keywords associated with the cluster match the query; and

providing one or more of the initial videos and one or more of the co-watched videos from the selected cluster in response to the query.

8. The system of claim 1, further operable to perform operations comprising:

receiving data identifying a set of target videos responsive to a query, wherein the target videos are ordered according to a respective result score for each target video, and wherein the target videos include a first target video having a first rank and a second target video having a second rank below the first rank;

determining that the second target video is in a cluster of co-watched videos for the first target video; and

providing the target videos in a different order according to a score adjustment for the second target video.

9. The system of claim 1, wherein each of one or more co-watched videos for the respective target video was viewed by a user after the user submitted a first search query, and the respective target video was viewed by the user after the user submitted a second search query, and the first search query is similar to the second search query.

10. The system of claim 1, wherein one or more of the co-watched videos for the respective target video were selected by a user from a watch page on which the user viewed the respective target video.

11. The system of claim 1, wherein the time window is defined by an amount of time.

12. The system of claim 1, wherein the time window is defined by a number of videos viewed by the user.

13. A method implemented by data processing apparatus, the method comprising:

for each of a plurality of target videos, wherein each target video was viewed by a user in response to a particular query submitted by the user, identifying a respective plurality of co-watched videos, wherein each respective co-watched video for the target video was viewed by a respective user within a time window of when the same user viewed the target video;

identifying a respective first query for each respective co-watched video wherein the respective co-watched video was selected by users more often when identified in a search result responsive to the respective first query in a plurality of queries submitted by users than when identified in search results responsive to other queries in the plurality of queries submitted by users, and wherein the selections responsive to the plurality of queries are counted only when the length of time the video was watched exceeds a threshold time, and wherein the first respective query for the respective co-watched video is different from the particular query of the target video for which the respective co-watched video was identified;

deriving, from each respective first query for each respective co-watched video, one or more respective keywords to be associated with the respective co-watched video;

for each target video in the plurality of target videos, using the one or more respective keywords for each respective co-watched video to generate one or more target video-specific clusters of co-watched videos for the target video, wherein a first video from each target video-specific cluster is a co-watched video for the target video, the co-watched video for the target video having first keywords including the respective first query, the first keywords being in common with second keywords from at least one second video from the cluster, the first keywords being different from the particular query; and

storing data identifying the clusters for each target video, wherein the data indexes the clusters for each target video according to the corresponding target video.

14. The method of claim 13, further comprising selecting content related to a first target video according to the clusters for the first target video.

15. The method of claim 14 wherein the content is a first co-watched video selected from one of the clusters for the first target video.

16. The method of claim 13, further comprising selecting a plurality of related videos for a first target video, the selecting comprising:

identifying the clusters for the first target video from the stored data; and

selecting a co-watched video from each of the clusters.

17. The method of claim 13, further comprising, for each target video in the plurality of target videos:

determining a score for each cluster generated for the target video; and

identifying one or more clusters as being off-topic for the target video according to the scores.

18. The method of claim 13 wherein each co-watched cluster is associated with one or more keywords of the respective co-watched videos that belong to the cluster, the method further comprising:

receiving input requesting presentation of a first target video;

identifying content to present with the first target video according to one or more of the keywords associated with the clusters for the first target video; and

providing the first target video along with the content.

19. The method of claim 18, further comprising:

receiving data identifying a set of initial videos responsive to a query;

selecting a cluster of co-watched videos for an initial target video selected from the initial videos, wherein the cluster of co-watched videos is selected according to how well the keywords associated with the cluster match the query; and

providing one or more of the initial videos and one or more of the co-watched videos from the selected cluster in response to the query.

20. The method of claim 13, further comprising:

receiving data identifying a set of target videos responsive to a query, wherein the target videos are ordered according to a respective result score for each target video, and wherein the target videos include a first target video having a first rank and a second target video having a second rank below the first rank;

determining that the second target video is in a cluster of co-watched videos for the first target video; and

providing the target videos in a different order according to a score adjustment for the second target video.

21. The method of claim 13, wherein each of one or more co-watched videos for the respective target video was viewed by a user after the user submitted a first search query, and the respective target video was viewed by the user after the user submitted a second search query, and the first search query is similar to the second search query.

22. The method of claim 13, wherein one or more of the co-watched videos for the respective target video were selected by a user from a watch page on which the user viewed the respective target video.

23. The method of claim 13, wherein the time window is defined by an amount of time.

24. The method of claim 13, wherein the time window is defined by a number of videos viewed by the user.

25. A non-transitory computer storage medium having instructions stored thereon that, when executed by data processing apparatus, cause the data processing apparatus to perform operations comprising:

for each of a plurality of target videos, wherein each target video was viewed by a user in response to a particular query submitted by the user, identifying a respective plurality of co-watched videos, wherein each respective co-watched video for the target video was viewed by a respective user within a time window of when the same user viewed the target video;

identifying a respective first query for each respective co-watched video wherein the respective co-watched video was selected by users more often when identified in a search result responsive to the respective first query in a plurality of queries submitted by users than when identified in search results responsive to other queries in the plurality of queries submitted by users, and wherein the selections responsive to the plurality of queries are counted only when the length of time the video was watched exceeds a threshold time, and wherein the respective first query for the respective co-watched video is different from the particular query of the target video for which the respective co-watched video was identified;

deriving, from each respective first query for each respective co-watched video, one or more respective keywords to be associated with the respective co-watched video,

for each target video in the plurality of target videos, using the one or more respective keywords for each respective co-watched video to generate one or more target video-specific clusters of co-watched videos for the target video, wherein a first video from each target video-specific cluster is a co-watched video for the target video, the co-watched video for the target video having first keywords including the respective first query, the first keywords being in common with second keywords from at least one second video from the cluster, the first keywords being different from the particular query; and

storing data identifying the clusters for each target video, wherein the data indexes the clusters for each target video according to the corresponding target video.

26. The computer storage medium of claim 25, further operable to perform operations comprising selecting content related to a first target video according to the clusters for the first target video.

27. The computer storage medium of claim 26, wherein the content is a first co-watched video selected from one of the clusters for the first target video.

28. The computer storage medium of claim 25, further operable to perform operations comprising selecting a plurality of related videos for a first target video, the selecting comprising:

identifying the clusters for the first target video from the stored data; and

selecting a co-watched video from each of the clusters.

29. The computer storage medium of claim 25, further operable to perform operations comprising, for each target video in the plurality of target videos:

determining a score for each cluster generated for the target video; and

identifying one or more clusters as being off-topic for the target video according to the scores.

30. The computer storage medium of claim 25, wherein each co-watched cluster is associated with one or more keywords of the respective co-watched videos that belong to the cluster, the computer storage medium further operable to perform operations comprising:

receiving input requesting presentation of a first target video;

identifying content to present with the first target video according to one or more of the keywords associated with the clusters for the first target video; and

providing the first target video along with the content.

31. The computer storage medium of claim 30, further operable to perform operations comprising:

receiving data identifying a set of initial videos responsive to a query;

selecting a cluster of co-watched videos for an initial target video selected from the initial videos, wherein the cluster of co-watched videos is selected according to how well the keywords associated with the cluster match the query; and

providing one or more of the initial videos and one or more of the co-watched videos from the selected cluster in response to the query.

32. The computer storage medium of claim 25, further operable to perform operations comprising:

receiving data identifying a set of target videos responsive to a query, wherein the target videos are ordered according to a respective result score for each target video, and wherein the target videos include a first target video having a first rank and a second target video having a second rank below the first rank;

determining that the second target video is in a cluster of co-watched videos for the first target video; and

providing the target videos in a different order according to a score adjustment for the second target video.

33. The computer storage medium of claim 25, wherein each of one or more co-watched videos for the respective target video was viewed by a user after the user submitted a first search query, and the respective target video was viewed by the user after the user submitted a second search query, and the first search query is similar to the second search query.

34. The computer storage medium of claim 25, wherein one or more of the co-watched videos for the respective target video were selected by a user from a watch page on which the user viewed the respective target video.

35. The computer storage medium of claim 25, wherein the time window is defined by an amount of time.

36. The computer storage medium of claim 25, wherein the time window is defined by a number of videos viewed by the user.