Data mining and data clustering

Abstract

US movie industry, which is renowned as Hollywood, is the oldest and largest film industry in the world. Being the top amongst the world, United State has the largest market share in terms of revenue. Predicting the success of a movie, would be the primary task for marketers and theatres, which helps to plan marketing strategy after releasing a movie. Given that the online data is enormous and constantly updated nowadays, information would be available from online database. IMDB, which stands for "Internet Movie Database", is one of the popular database that gathered information of film and TV programs. In order to evaluate the approach to predict how successful a movie is, data of over 1000 US movies released between 2011 and 2016 would be collected from IMDB by data mining techniques. In this project, the movie gross would be the target data that can indicate the success of a movie. The extracted data, which includes MPAA rating, genre and user rating, would be used to predict movie gross by three classification methods - Naïve Bayes Classification, logistic regression and support vector machine.