US 12,008,051 B1
Generating digital content clusters based on web activity
Devin Mancuso, San Francisco, CA (US); Benjamin Joseph Calabrese, Portland, OR (US); Austin Sung, San Francisco, CA (US); Jason Stakelon, Sausalito, CA (US); and Bonita Leung, Brooklyn, NY (US)
Assigned to Dropbox, Inc., San Francisco, CA (US)
Filed by Dropbox, Inc., San Francisco, CA (US)
Filed on Dec. 19, 2022, as Appl. No. 18/068,132.
Int. Cl. G06F 16/90 (2019.01); G06F 11/14 (2006.01); G06F 16/906 (2019.01); G06F 16/9535 (2019.01)
CPC G06F 16/906 (2019.01) [G06F 11/1469 (2013.01); G06F 16/9535 (2019.01); G06F 2201/84 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
identifying topic data associated with a plurality of digital content items from an access history of a user account of a content management system;
determining focus data indicating activity patterns associated with the plurality of digital content items from the access history of the user account;
generating, from the topic data and the focus data, a content cluster by grouping two or more digital content items of the plurality of digital content items together according to the topic data and the focus data of the plurality of digital content items;
determining return visit likelihood scores for the two or more digital content items within the content cluster based at least on the focus data;
based on the return visit likelihood scores, generating a content restore element to visually indicate one or more content items from the content cluster that satisfy a threshold return visit likelihood score; and
providing, for display on a client device associated with the user account, the content restore element representing the content cluster, wherein the content restore element is selectable to restore the two or more digital content items of the content cluster by accessing the two or more digital content items via the client device.