CPC H04N 21/4756 (2013.01) [H04N 21/4524 (2013.01); H04N 21/4665 (2013.01)] | 21 Claims |
1. A method comprising:
accessing, by a hardware processor, contextual data for a current media asset having a reference rating associated with an initial geographic region, the contextual data describing a set of events within content data of the current media asset;
generating, based on the contextual data and by the hardware processor, a plurality of predicted ratings for a set of given event types present in the set of events, the generating of the plurality of predicted ratings comprising:
accessing a set of cultural inference rules for the set of given event types, the set of cultural inference rules being associated with a first geographic region;
generating a set of impact scores corresponding to the set of given event types based on one or more first cultural inference rules from the set of cultural inference rules;
using a machine learning algorithm to generate a set of rating point values corresponding to the set of given event types for the first geographic region based on the set of impact scores; and
determining the plurality of predicted ratings associated with the first geographic region by transforming the reference rating to the plurality of predicted ratings based on the set of rating point values; and
generating, based on the plurality of predicted ratings and by the hardware processor, an overall rating for the current media asset and associated with the first geographic region.
|