Can AI predict a movie's MPAA rating with just its description?
This repo demonstrates how to build a natural language processing system in order to predict, with reliable accuracy, a film's rating as assigned by the Motion Picture Association of America, based entirely off a studio's summary/plot/description/synopsis. The system learns patterns from the vocabulary and language used to promote movies to audiences, using verbiage and vernacular that attempts to influence people to go to the theater, rent, purchase, or stream a film.
Obviously, this practice started back when studios had limited space on a movie poster, marquee or box cover - and in the more modern day have to craftily appeal to search engine indexing and social media buzziness. It's a creative endeavor that produces wildly inconsistent data. So therein lies the challenge.
It's taking a machine learning approach to assessing advertising effectiveness.
Instructions
- Download the Rotten Tomatoes Movie Database on Kaggle
- Download Stanford's GloVe 6B word embedding - make sure you grab the 6B version!