Ranking Based on Facial Image Analysis

A system comprising:

• a social network comprising a plurality of users, each of said plurality of users having user metadata and an image collection;
• a ranking engine that:
• receives image metadata from a first image collection associated with a first user, said image metadata comprising processed face objects from said images;
• analyzes said image metadata to identify similar face objects, said similar face objects having a matching criteria with respect to a first threshold;
• groups said similar face objects into clusters;
• and
• determines a ranking of said clusters according to at least cluster size.

A method comprising:

• receiving a first image collection comprising images comprising faces;
• for each of said images in said first image collection, analyzing an image to identify said faces within said image, processing said faces to determine face vectors for each of said faces, and store said face vectors as image metadata;
• analyzing said image metadata to identify similar face objects, said similar face objects having a matching criteria with respect to a first threshold;
• grouping said similar face objects into clusters;
• determining a ranking for said clusters according to at least cluster size;
• identifying a person for a plurality of said clusters;
• receiving a set of results relating to people, at least some of said people being persons associated with said clusters;
• and
• presenting said set of results according to said ranking.