Predicting Ratings for New Movie Releases from Twitter Content

With microblogging platforms such as Twitter generating huge amounts of textual data every day, the possibilities of knowledge discovery through Twitter data becomes increasingly relevant. Similar to the public voting mechanism on websites such as the Internet Movie Database (IMDb) that aggregates movies ratings, Twitter content contains reflections of public opinion about movies. This study aims to explore the use of Twitter content as textual data for predictive text mining. In this study, a corpus of tweets was compiled to predict the rating scores of newly released movies on IMDb. Predictions were done with several different machine learning algorithms, exploring both regression and classification methods. In addition, this study explores the use of several different kinds of textual features in the machine learning tasks. Results show that prediction performance based on textual features derived from our corpus of tweets improved on the baseline for both regression and classification tasks.