US 11,989,746 B2
Methods and apparatus to project ratings for future broadcasts of media
Jingsong Cui, Pennington, NJ (US); Peter Campbell Doe, Ridgewood, NJ (US); and Scott Sereday, Rochelle Park, NJ (US)
Assigned to The Nielsen Company (US), LLC, New York, NY (US)
Filed by The Nielsen Company (US), LLC, New York, NY (US)
Filed on Apr. 14, 2023, as Appl. No. 18/301,183.
Application 18/301,183 is a continuation of application No. 17/121,323, filed on Dec. 14, 2020, granted, now 11,657,413.
Application 17/121,323 is a continuation of application No. 16/036,614, filed on Jul. 16, 2018, granted, now 10,867,308, issued on Dec. 15, 2020.
Application 16/036,614 is a continuation of application No. 14/951,465, filed on Nov. 24, 2015, abandoned.
Claims priority of provisional application 62/083,716, filed on Nov. 24, 2014.
Prior Publication US 2023/0325858 A1, Oct. 12, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06Q 30/0202 (2023.01); G06Q 30/0201 (2023.01); G06Q 50/00 (2012.01); H04N 21/25 (2011.01); H04N 21/258 (2011.01); H04N 21/442 (2011.01); H04N 21/45 (2011.01); H04N 21/466 (2011.01); H04N 21/658 (2011.01)
CPC G06Q 30/0202 (2013.01) [G06Q 30/0201 (2013.01); G06Q 50/01 (2013.01); H04N 21/251 (2013.01); H04N 21/252 (2013.01); H04N 21/25891 (2013.01); H04N 21/44226 (2020.08); H04N 21/4532 (2013.01); H04N 21/4665 (2013.01); H04N 21/6582 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An apparatus comprising:
a data transformer to transform audience measurement data to determine normalized training data;
a model builder to:
build and train a machine learning model for predicting ratings of a media asset for a future quarter of programming,
wherein a gap between a current quarter and the future quarter is used as a basis to (i) select the machine learning model from a plurality of models, and (ii) determine whether or not to exclude a portion of historical data used to build the machine learning model,
wherein a classification of the media asset is used as a basis to select to a predictive feature schema from a plurality of schemas,
and wherein a subset of predictive features selected from the normalized training data according to the predictive feature schema is used to train the machine learning model; and
a ratings projector to:
obtain predictive features data based on the selected predictive features schema, and
apply the obtained predictive features data to the trained machine learning model to predict ratings for the media asset for the future quarter.