Selecting content objects for recommendation based on content object collections

1. A computer implemented method of selecting an image for recommendation to a user, the method comprising:

storing a plurality of image collections in a non-transitory computer-readable storage medium, wherein each image collection comprises a set of associated images;

receiving a request for a seed image from a user;

identifying a plurality of seed collections, wherein each seed collection comprises a stored image collection identifying the seed image;

identifying a plurality of candidate images based on the identified seed collections;

identifying properties of the seed image;

determining, for each image collection, a coherence score representative of a similarity of the set of images within the image collection, the coherence score comprising a measure of entropy of object vectors associated with each of the set of images within the image collection, each entry of an object vector representative of a property of an image associated with the object vector;

determining, for each candidate image in the plurality of candidate images, a corresponding result by dividing a first sum of the determined coherence scores for the image collections comprising seed collections to which the candidate image belongs by a second sum of the determined coherence scores for all of the plurality of image collections;

ranking each candidate image in the plurality of candidate images based on the determined corresponding results and based on the identified properties of the seed image; and

selecting a candidate image for recommendation to the user based on the ranking of candidate images.

2. The method of claim 1 , wherein at least one of the plurality of image collections comprises a user-defined collection.

3. The method of claim 1 , wherein at least one of the plurality of image collections comprises a machine-generated collection.

4. The method of claim 1 , wherein identifying a plurality of seed collections comprises identifying all of the stored image collections that identify the seed image.

5. The method of claim 1 , wherein identifying a plurality of candidate images comprises identifying all of the images identified by the identified seed collections except for the seed image.

6. The method of claim 1 , wherein the identifying properties of the seed image comprises identifying a categorization of the seed image.

7. The method of claim 1 , wherein the identifying properties of the seed image comprises identifying tags associated with the seed image.

8. The method of claim 1 , wherein the identifying properties of the seed image comprises identifying characteristics of the seed image.

9. The method of claim 1 , wherein the identified candidate images are ranked further based on a number of seed collections that identify each candidate image.

10. The method of claim 1 , wherein the identified candidate images are ranked further based on a proportion of seed collections that identify each candidate image relative to a proportion of stored image collections that identify each candidate image.

11. The method of claim 1 , wherein the identified candidate images are ranked further based on common properties between each candidate image and the seed image.

12. The method of claim 1 , wherein the selected candidate image for recommendation to the user comprises a top-ranked candidate image.

13. The method of claim 1 , further comprising:

displaying the selected candidate image to the user within a portion of an interface dedicated to displaying recommended images.

14. The method of claim 1 , further comprising:

determining a centrality score for each seed collection representing a representativeness of the seed image to the seed collection; and

further ranking the identified candidate images based on the determined centrality scores.

15. The method of claim 14 , wherein identifying a plurality of seed collections further comprises identifying only seed collections with a centrality score above a pre-determined threshold.

16. The method of claim 14 , wherein the determined centrality score is based on a distance from the seed image in a distribution of the properties of the images identified by the seed collection to a mean of the distribution.

17. A non-transitory computer-readable storage medium having executable computer program instructions embodied therein for selecting an image for recommendation to a user, the computer program instructions configured to, when execute, cause a computer to:

store a plurality of image collections in a non-transitory computer-readable storage medium, wherein each image collection comprises a set of associated images;

receive a request for a seed image from a user;

identify a plurality of seed collections, wherein each seed collection comprises a stored image collection identifying the seed image;

identify a plurality of candidate images based on the identified seed collections;

identify properties of the seed image;

determine, for each image collection, a coherence score representative of a similarity of the set of images within the image collection, the coherence score comprising a measure of entropy of object vectors associated with each of the set of images within the image collection, each entry of an object vector representative of a property of an image associated with the object vector;

determine, for each candidate image in the plurality of candidate images, a corresponding result by dividing a first sum of the determined coherence scores for the image collections comprising seed collections to which the candidate image belongs by a second sum of the determined coherence scores for all of the plurality of image collections;

rank each candidate image in the plurality of candidate images based on the determined corresponding results and based on the identified properties of the seed image; and

select a candidate image for recommendation to the user based on the ranking of candidate images.

18. A computer implemented method of selecting an image for recommendation to a user, the method comprising:

storing a plurality of image collections in a non-transitory computer-readable storage medium, wherein each image collection comprises a set of associated images;

displaying a seed image to a user;

identifying a plurality of seed collections, wherein each seed collection comprises a stored image collection identifying the seed image;

identifying a plurality of candidate images from the plurality of seed collections;

determining, for each image collection, a coherence score representative of a similarity of the set of images within the image collection, the coherence score comprising a measure of entropy of object vectors associated with each of the set of images within the image collection, each entry of an object vector representative of a property of an image associated with the object vector;

determining, for each candidate image in the plurality of candidate images, a corresponding result by dividing a first sum of coherence scores for the image collections comprising seed collections to which the candidate image belongs by a second sum of coherence scores for all of the plurality of image collections; and

selecting an image from the plurality of candidate images for recommendation to the user based on the determined corresponding results.