Deep Learning Approach to Predicting the Success of Content

Movie Industry across the globe is a multi-billion dollar in revenue. There are several factors and somewhat complicated, which makes a movie a box-office success. The financial cost of making a movie [1] can range from modest cost to mammoth several 100 million dollars in spending. The audience rules in terms of liking/disliking of a movie. Financiers and producers reel through several financial losses if the movie does not do well in the box-office. Can Artificial Intelligence techniques help from historical data, be it movies, audience reviews/sentiments, cast/crews to predict the success of a movie? Our paper explores the proposal for a hybrid RNN/LSTM/CNN based neural network architecture that can predict the success of the content. It leads to a useful foundation that enables content creators to manage the pre-production part of the journey better by making the right decisions that enhance the probability of success. The foundation of our proposal can help the pre-production state with a certain degree of predictability to the success of the content, explainability causing the success, and the next best action.

Published

2020-11

Content type

Original Research

Keywords

CNN, LSTM, RNN, Movies, Posters, Videos, Genres, Explainability

DOI

10.5594/M001914