US 12,265,897 B2
Systems and methods for automated content curation using signature analysis
Christopher Ambrozic, Chapel Hill, NC (US); and Michael Dean Hoffman, Durham, NC (US)
Assigned to Adeia Guides Inc., San Jose, CA (US)
Filed by Adeia Guides Inc., San Jose, CA (US)
Filed on Jan. 10, 2023, as Appl. No. 18/095,169.
Application 18/095,169 is a continuation of application No. 16/838,236, filed on Apr. 2, 2020, granted, now 11,574,248.
Prior Publication US 2023/0274187 A1, Aug. 31, 2023
Int. Cl. H04N 21/44 (2011.01); G06N 20/00 (2019.01); H04N 21/466 (2011.01)
CPC G06N 20/00 (2019.01) [H04N 21/44008 (2013.01); H04N 21/4663 (2013.01); H04N 21/4666 (2013.01); H04N 21/4668 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A method for curating content, the method comprising:
determining, for each content item of a plurality of content items, a corresponding plurality of machine learning model-generated signature vectors;
receiving a request for generating a playlist, the request associated with a narrative;
determining a first machine learning model-generated signature vector for a first portion of the narrative and a second machine learning model-generated signature vector for a second portion of the narrative;
selecting a first content item from the plurality of content items based at least in part on a match of the first machine learning model-generated signature vector of the first portion of the narrative and a machine learning model-generated signature vector of the plurality of signature vectors corresponding to the first content item;
adding the first content item to the playlist;
selecting a second content item from the plurality of content items based at least in part on a match of the second machine learning model-generated signature vector of the second portion of the narrative and a machine learning model-generated signature vector of the plurality of signature vectors corresponding to the second content item;
adding the second content item to the playlist; and
providing the playlist for consumption.