US 9,811,569 B2
Suggesting candidate profiles similar to a reference profile
Christian Posse, Forster City, CA (US); Abhishek Gupta, San Francisco, CA (US); Anmol Bhasin, Los Altos, CA (US); and Monica Rogati, Sunnyvale, CA (US)
Assigned to LinkedIn Corporation, Sunnyvale, CA (US)
Filed by LinkedIn Corporation, Sunnyvale, CA (US)
Filed on Nov. 30, 2016, as Appl. No. 15/364,621.
Application 15/364,621 is a continuation of application No. 14/602,661, filed on Jan. 22, 2015, granted, now 9,544,392.
Application 14/602,661 is a continuation of application No. 13/194,883, filed on Jul. 29, 2011, granted, now 8,972,414.
Prior Publication US 2017/0124089 A1, May 4, 2017
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 17/30 (2006.01); H04L 29/08 (2006.01)
CPC G06F 17/3053 (2013.01) [G06F 17/30554 (2013.01); G06F 17/30867 (2013.01); H04L 67/22 (2013.01); H04L 67/306 (2013.01)] 20 Claims
OG exemplary drawing
1. A method comprising:
detecting, by one or more processors, a request for suggestion of one or more candidate entities based on a reference entity represented in a social database, the reference entity being represented by a reference profile in the social database, candidate profiles in the social database each representing a different candidate entity;
generating, by the one or more processors, scores that each correspond to a different one of the candidate profiles, the generating including, for each score:
accessing parts of a different candidate profile among the candidate profiles for comparison to corresponding parts of the reference profile; and
determining sub-scores that each quantify similarity of a different part of the candidate profile to a corresponding part of the reference profile, the generating of the score for the candidate profile being based on the determined sub-scores; and
causing, by the one or more processors, display of one or more profile identifiers that collectively reference a subset of the candidate profiles, the displayed profile identifiers being selected based on comparisons of the generated scores of candidate profiles in the subset to an extremum score among the generated scores.