PREDIKSI RATING FILM ANIMASI BERDASARKAN ELEMEN MISE EN SCENE MENGGUNAKAN NEURAL NETWORK

Abstract

A production house can make adjustments in releasing the film if knows the
possibility of success to the related film. The goal is to get maximum profit after the
film is released. It can even use a prediction to know how market developments
are. In this study, the authors used data from IMDb to predict the rating on a film.
Because IMDb is the largest database consisting of relevant and comprehensive
information related to films. It is necessary to consider the additional features of
visual, which basically refers to everything that appears on the screen. Everything
captured by the camera consisting of settings, lighting, the human figure, and
composition, these four elements are part of the mise en scene. To determine the
level of accuracy in film rating predictions, the author uses the science of soft
computing, namely the neural network, focusing on the prediction part. The
accuracy of the animated movie rating prediction is based on the mise en scene
element which is equal to 44%, the accuracy is based on the average of each test
option.