Multi-label Classification of Movie Posters into Genres with Rakel Ensemble Method

Movies can belong to more than one genre, so the problem of determining the genres of a movie from its poster is a multi-label classification problem. To solve the multi-label problem, we have used the RAKEL ensemble method along with three typical single-label base classification methods: Naive Bayes, C4.5 decision tree, and k-NN. The RAKEL method strives to overcome the problem of computational cost and power set label explosion by breaking the initial set of labels into several small-sized label sets. The classification performance of base classifiers on different feature sets is evaluated using multi-label evaluation measures on poster dataset containing 6000 posters classified into 18 and 11 genres. Keeping this in mind, we wanted to examine how different visual feature sets, extracted from poster images, are related to the performance of automatic detection of movie genres, as well as compare it to the performance obtained with the Classeme feature descriptors trained on the datasets of general images.