PREDICTING ANIMATED FILM OF BOX-OFFICE SUCCESS WITH NEURAL NETWORKS

Abstract

Animation industry involves huge funds in production process and its success will give  great income. Predicting the box-office of animated film has become an interesting topic to be discussed, because past studies are shown to be contradictory. Sharda and Delen conducted a similar study that used seven parameters, i.e. MPAA rating, competition, star value, genre, special effects, sequel and number of screens; and generated pinpoint accuracy (i.e. Bingo) with 36.9% and within one category (1-Away) with 75.2%. The authors proposed new and simple parameters that can be used to predict the success of animated films, i.e. the actors/actress, animation studio, genre, MPAA rating and the sequel of the film. These five parameters are relatively simple because it can be easily collected. In this study, the use of neural networks in predicting the financial performance of 120 animated films from 1995 until 2013 was explored. There are three categories of financial performance that become the class label of this study, they are: low, medium and high. Our prediction result in bingo is 58% and 1-away is 89,7%. By using the simple parameters, this study can reach a better accuracy. It is expected that this prediction can help animation film industry to predict the expected revenue range before its theatrical release.