US 12,223,001 B2
Ranking candidate search results by activeness
Rui Ma, Santa Clara, CA (US); Keheng Zhang, Belmont, CA (US); Yongqing Yuan, Mountain View, CA (US); Baolu Shen, Santa Clara, CA (US); and Kefang Ning, Sunnyvale, CA (US)
Assigned to Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed by Microsoft Technology Licensing, LLC, Redmond, WA (US)
Filed on Mar. 28, 2019, as Appl. No. 16/367,753.
Prior Publication US 2020/0311163 A1, Oct. 1, 2020
Int. Cl. G06F 15/16 (2006.01); G06F 16/903 (2019.01); G06F 16/9536 (2019.01); G06F 16/9538 (2019.01); G06N 20/00 (2019.01)
CPC G06F 16/9536 (2019.01) [G06F 16/90335 (2019.01); G06F 16/9538 (2019.01); G06N 20/00 (2019.01)] 13 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving, via a user interface, a search query comprising search parameters specified by a user of an online service;
identifying, by one or more computer systems, a plurality of candidates having profiles hosted by the online service with profile attributes that satisfy the search parameters of the search query;
for each candidate in the plurality of candidates, applying, by one or more computer systems, a first machine learned model to a first set of features to generate relevance scores indicating a measure of relevance for each candidate, the first set of features including features associated with the profile attributes of the candidate and features associated with the search query;
for each candidate in the plurality of candidates having a relevance score that exceeds a threshold, determining, by one or more computer systems, a second set of features platform comprising at least a feature indicating a number of messages communicated to the candidate from moderators of opportunities hosted by the online service and a feature indicating a percentage of messages communicated to the candidate that have been accepted by the candidate;
for each candidate in the plurality of candidates having a relevance score that exceeds the threshold, applying, by the one or more computer systems, a second machine learned model to the second set of features to produce an activeness score representing a level of activity of the candidate with respect to the online service;
generating a first ranking of the plurality of candidates according to the relevance scores; and
outputting via a user interface at least a portion of the first ranking of the plurality of candidates as a set of search results of the search, and including within the user interface a user interface element that, when selected, causes one or more computer systems to:
generate a second ranking of the plurality of candidates according to the activeness scores; and
output via the user interface at least a portion of the second ranking of the plurality of candidates as a set of search results of the search.