What can Hollywood Learn from YouTube?
Florida Southern College
Though most of a film’s marketing budget is directed toward television advertisements, the rise in video-sharing platforms provides an additional outlet for promotion. Arguably, the most notable of these video-sharing platforms is YouTube. This paper addresses how the data of a film trailer released on YouTube – like number of views, comments, likes, and dislikes – help predict the generated revenue for a film, additionally exploring how these features expand on the effect of other features in the filmmaking process. The data used to evaluate these features is generated from films released in the 2010s in The Movie Database (TMDb) and video trailer statistics from the YouTube Data API. Specifically, these features are used to train machine learning models to evaluate which models are most successful in revealing connections between these features and revenue. This work closely examines the impact of the statistics of a film trailer from YouTube to help producers better understand the impact of their investments.
Honors Thesis Spring 2021
Motion pictures, Machine learning, Decision trees, Neural networks (Computer science), Regression analysis, Motion picture industry, Motion pictures -- Reviews, YouTube