CPC H04L 51/52 (2022.05) [G06N 20/00 (2019.01); G06Q 50/01 (2013.01); H04L 51/10 (2013.01)] | 17 Claims |
1. A computer-implemented method comprising:
capturing interactions of content generators with notifications received by the content generators, the received notifications being associated with viewer responses to content items generated by the content generators;
training a machine-learning (ML) model based on the captured interactions;
detecting a plurality of first notifications generated for a first content generator, the plurality of first notifications being associated with interactions of a first plurality of viewers to content items from the first content generator;
calculating, by the ML model, an affinity score between the first content generator and each of the first plurality of viewers;
ranking the plurality of first notifications based on affinity scores of the first content generator and a viewer associated with each first notification, wherein ranking the plurality of the first notifications comprises calculating a multi-objective optimization (MOO) score based at least in part on the affinity score, and ranking the plurality of the first notifications according to the MOO score between the first content generator and the viewer associated with the notification;
selecting a plurality of second notifications from the ranked plurality of first notifications; and
generating notifications for the first content generator for the selected plurality of second notifications.
|