Granted Patent B2

US 9,432,477 · App. 14/231,285 · Granted Aug 30, 2016

Abstract

Claims (30)

1. A computer implemented method for matching video content comprising:

recording, by use of a processor, characteristics of a plurality of videos, the characteristics comprising keywords, views, comments, subscriptions to content channels, uploaded content, likes, user followings, ownership, content, characters, topics, tags, demographics, popularity, and user identities;

calculating content profile elements comprising keyword data, view data, comment data, subscription data, upload data, like data, upload data, like data, user data, following data, ownership data, content data, character data, topic data, tag data, demographic data, and popularity data from the keywords, the views, the comments, the subscriptions to content channels, the uploaded content, the likes, the user followings, the ownership, the content, the characters, the topics, the tags, the demographics, the popularity, and the user identities;

calculating a content profile for each video as a multidimensional coordinate of the content profile elements in a profile space;

calculating a content proximity of a target video content profile as a function of a weight vector, wherein the content proximity is a volume in the profile space;

identifying a matching content profile within the content proximity; and linking the video of the matching content profile to a target video corresponding to the target video content profile.

2. The method of claim 1 , wherein the matching content profile further has a number of likes that exceeds a like threshold.

3. The method of claim 1 , wherein the matching content profile further has a number of user followings that exceeds a following threshold.

4. The method of claim 1 , wherein the matching content profile is further identified with one of a linear programming algorithm, a blossom algorithm, an un-weighted bipartite algorithm, a weighted bipartite algorithm, and a maximum matching edge algorithm.

5. An apparatus for matching video content comprising:

a processor;

a memory storing program code executable by the processor to:

record characteristics of a plurality of videos, the characteristics comprising keywords, views, comments, subscriptions to content channels, uploaded content, likes, user followings, ownership, content, characters, topics, tags, demographics, popularity, and user identities;

calculate content profile elements comprising keyword data, view data, comment data, subscription data, upload data, like data, upload data, like data, user data, following data, ownership data, content data, character data, topic data, tag data, demographic data, and popularity data from the keywords, the views, the comments, the subscriptions to content channels, the uploaded content, the likes, the user followings, the ownership, the content, the characters, the topics, the tags, the demographics, the popularity, and the user identities;

calculate a content profile for each video as a multidimensional coordinate of the content profile elements in a profile space;

calculate a content proximity of a target video content profile as a function of a weight vector, wherein the content proximity is a volume in the profile space;

identifies the matching content profile within the content proximity; and

link the video of the matching content profile to a target video corresponding to the target video content profile.

6. The apparatus of claim 5 , wherein the matching content profile further has a number of likes that exceeds a like threshold.

7. The apparatus of claim 5 , wherein the matching content profile further has a number of user followings that exceeds a following threshold.

8. The apparatus of claim 5 , wherein the matching content profile is further identified with one of a linear programming algorithm, a blossom algorithm, an un-weighted bipartite algorithm, a weighted bipartite algorithm, and a maximum matching edge algorithm.

9. A program product for matching video content comprising a non-transitory computer readable storage medium storing program code executable by a processor to perform:

recording characteristics of a plurality of videos, the characteristics comprising keywords, views, comments, subscriptions to content channels, uploaded content, likes, user followings, ownership, content, characters, topics, tags, demographics, popularity, and user identities;

calculating content profile elements comprising keyword data, view data, comment data, subscription data, upload data, like data, upload data, like data, user data, following data, ownership data, content data, character data, topic data, tag data, demographic data, and popularity data from the keywords, the views, the comments, the subscriptions to content channels, the uploaded content, the likes, the user followings, the ownership, the content, the characters, the topics, the tags, the demographics, the popularity, and the user identities;

calculating a content profile for each video as a multidimensional coordinate of the content profile elements in a profile space;

calculating a content proximity of a target video content profile as a function of a weight vector, wherein the content proximity is a volume in the profile space;

identifying a matching content profile within the content proximity; and

linking the video of the matching content profile to a target video corresponding to the target video content profile.

10. The program product of claim 9 , wherein the matching content profile further has one or more of a number of likes that exceeds a like threshold and a number of user followings that exceeds a following threshold.

11. The program product of claim 9 , wherein the matching content profile is further identified with one of a linear programming algorithm, a blossom algorithm, an un-weighted bipartite algorithm, a weighted bipartite algorithm, and a maximum matching edge algorithm.

Assignments (16)