Selecting content objects for recommendation based on content object collections

• US 10,417,294 B2
• Filed: 12/13/2016
• Issued: 09/17/2019
• Est. Priority Date: 08/30/2012
• Status: Active Grant

• Current Assignee: Wix.com Ltd.
• Original Assignee: Wix.com Ltd.
• Inventors: McCann, Andrew Simz Arneson, Donaldson, Roger David, Nell, Christopher Warren
• Primary Examiner(s): Owyang, Michelle N
• Application Number:
  US15/376,666
• Publication Number:
  US 20170220689A1
• Time in Patent Office: 1,008 Days
• Field of Search: 707705-780
• CPC Class Codes: G06F 16/2462 Approximate or statistical queries
• CPC Class Codes: G06F 16/51 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/532 Query formulation, e.g. graphical querying
• CPC Class Codes: G06F 16/583 using metadata automatically derived from the content
• CPC Class Codes: G06F 16/5866 using information manually generated, e.g. tags, keywords, comments, manually generated location and time information
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9538 Presentation of query results

Abstract

Collections identifying associated content objects, such as user-defined collections, are stored. A request to display or play a seed content object is received. A plurality of seed collections are identified based on the seed object. A plurality of candidate content objects are identified based on the identified seed collections, for instance by identifying the content objects that are identified by the seed collections. Properties of the seed object are identified, such as the categorization of the seed object, or the tags associated with the seed object. The candidate objects are ranked based on the identified seed collections and the identified properties of the seed object, and a candidate object is selected for recommendation to the user based on the ranking. The ranking of candidate objects can also be determined based on the coherence of the seed collections and the centrality of the seed object to the seed collections.

Claims

1. A method of selecting an image for recommendation to a user of a content hosting service, the method operating on the content hosting service having a content storage module, a collection storage module and a processor and operative with a client module, the method comprising:

storing a plurality of user-defined collections of images in said collection storage module from a plurality of said users of said content hosting service;

for each image collection of a plurality of said user-defined image collections stored in said collection storage module, determining, via said processor, a coherence score representative of a similarity of a set of images within the image collection to each other;

via said processor, using a seed image, received from a user of said content hosting service via said client module, to identify seed collections from among said user-defined collections of images which have said seed image as one of the images of said seed collection;

determining, via said processor, a centrality score for each said seed collection representing a representativeness of the seed image to the seed collection;

selecting candidate seed collections with a centrality score above a pre-determined threshold;

identifying a plurality of candidate images for recommendation from said selected seed collections;

determining, via said processor, for each candidate image, a per seed score as a function of the coherence scores for said selected candidate seed collections and the coherence scores of the plurality of image collections;

ranking, via said processor, each candidate image based on the per seed collection scores and on properties of the seed image; and

recommending, via said processor, at least one candidate image for recommendation to the user via said client module based on the ranking of the candidate images.

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

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

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

5. The method of claim 1, and also comprising further ranking the candidate images based on a number of seed collections to which each candidate image belongs.

6. The method of claim 1, and also comprising further ranking the candidate images 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.

7. The method of claim 1, and also comprising further ranking the candidate images based on common properties between each candidate image and the seed image.

8. The method of claim 1, 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.

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

store a plurality of user-defined collections of images in said collection storage module from a plurality of said users of said content hosting service;

for each image collection of a plurality of said user-defined image collections, determine a coherence score representative of a similarity of a set of images within the image collection to each other;

use a seed image, received from a user of said content hosting service, identify seed collections from among said user-defined collections of images which have said seed image as one of the images of said seed collection;

determine a centrality score for each said seed collection representing a representativeness of the seed image to the seed collection;

select candidate seed collections with a centrality score above a pre-determined threshold;

determine, for each candidate image, a per seed score as a function of of the coherence scores for said selected candidate seed collections the coherence scores of the plurality of image collections;

rank each candidate image based on the per seed collection scores and on properties of the seed image; and

recommend at least one candidate image for recommendation to the user based on the ranking of the candidate images.

10. A method of selecting an image for recommendation to a user of a content hosting service, the method operating on the content hosting service having a content storage module, a collection storage module and a processor and operative with a client module, the method comprising:

for each image collection of a plurality of user-defined image collections stored in said collection storage module, determining, via said processor, a coherence score representative of a similarity of a set of images within the image collection to each other;

given a seed image indicated by said user via said client module, identifying, from among a plurality of stored image collections, a plurality of seed collections of images having said seed image as one of their images;

determining, via said processor, a centrality score for each said seed collection representing a representativeness of the seed image to the seed collection and selecting candidate seed collections with a centrality score above a pre-determined threshold;

determining, for each of a plurality of candidate images from said selected candidate seed collections, a per seed score as a function of the coherence scores for said selected candidate seed collections and the coherence scores of the plurality of image collections; and

recommending, via said processor, at least one candidate image for recommendation to the user via said client module based on the scores of the candidate images.